Re: Review Request 13223: (CLOUDSTACK-2729) use file lock to prevent concurrent refreshPool/deleteVolume on KVM shared storage pool
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13223/#review24778 --- Commit 9ec8d48945f465b74284034b94f12efdd319bd0f in branch refs/heads/4.2 from Wei Zhou [ https://git-wip-us.apache.org/repos/asf?p=cloudstack.git;h=9ec8d48 ] Revert CLOUDSTACK-2729: don't refresh storage pool when accessing libvirt storage pool This reverts commit b50060be0cfb5e8fb4334ff4dbcbab3d2ee3b417. Signed-off-by: Wei Zhou w.z...@leaseweb.com - ASF Subversion and Git Services On Aug. 5, 2013, 6:07 p.m., Wei Zhou wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13223/ --- (Updated Aug. 5, 2013, 6:07 p.m.) Review request for cloudstack, edison su and Wido den Hollander. Bugs: CLOUDSTACK-2729 Repository: cloudstack-git Description --- The storage pool issue (CLOUDSTACK-2729) is because of a bug in libvirt (https://bugzilla.redhat.com/show_bug.cgi?id=977706) We need to prevent deleting a volume when refreshing the pool. This patch use a simple file lock to implement it. PS: I have tested another file lock similar to Read/Write file lock, but it was very unstable. Diffs - plugins/hypervisors/kvm/src/com/cloud/hypervisor/kvm/storage/LibvirtStorageAdaptor.java b8a9f0f Diff: https://reviews.apache.org/r/13223/diff/ Testing --- Applied on 4.0.2 and 4.0.1 Testing On 4.0.1 From 20,June 3 nodes, create a VM on each node every 15 minutes. Destroy the VMs 5 minutes later. expunge.inteval = 600 (10 minutes), expunge.worker = 2 Testing On 4.0.2 From 01,July 2 nodes, create two VMs on each node every 5 minutes. Destroy the VMs 4 minutes later. expunge.inteval = 600 (10 minutes), expunge.worker = 2 Thanks, Wei Zhou
Re: Review Request 13323: if a failure occurs while adding VM to another network (this should be the first vm in the subnet). The ip alias created as a part of this process is not removed.
On Aug. 7, 2013, 12:03 a.m., Sheng Yang wrote: server/src/com/cloud/network/NetworkManagerImpl.java, line 2433 https://reviews.apache.org/r/13323/diff/1/?file=337776#file337776line2433 It's better to show the semantic of if this is the last nic in the subnet in code. Say, if (vm.getType() == Type.User it's the last nic) bharat kumar wrote: Hi sheng, removeDhcpServiceInsubnet functon calls the function listLastNicsInSubnet to get the last nics in subnet and then it removes the DhcpService based on the networkId. So I mean we do not know if it is last nic in the subnet until we call the listLastNicsInsubnet so we cannot put this check before hand. You can wrap it in more meaningful way, e.g. get the check in network manager. The code should be the best comment for itself. In this case, only remove subnet after last nic unplugged is not in the if statement, but wrapped in the function. Semantic can be improved. On Aug. 7, 2013, 12:03 a.m., Sheng Yang wrote: server/src/com/cloud/vm/VirtualMachineManagerImpl.java, line 479 https://reviews.apache.org/r/13323/diff/1/?file=33#file33line479 Why we still need this if networkid is null then we remove all? removeNic() for each network in network manager is not enough? bharat kumar wrote: Hi sheng, In cases when a VM is getting expunged we need to remove the Dhcpservice in all the networks in which this vm has a last nic. so we pass a null as networkId to do this. I meant, when vm is expunged, the unplugged function above in network manager wouldn't be called? If it's called, then why you need to check the nics of vm again? - Sheng --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13323/#review24762 --- On Aug. 6, 2013, 12:48 p.m., bharat kumar wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13323/ --- (Updated Aug. 6, 2013, 12:48 p.m.) Review request for cloudstack, Alena Prokharchyk and Sheng Yang. Bugs: Cloudstack-4083 Repository: cloudstack-git Description --- https://issues.apache.org/jira/browse/CLOUDSTACK-4083 if a failure occurs while adding VM to another network (this should be the first vm in the subnet). The ip alias created as a part of this process is not removed. This occurred because we were not cleaning the alias ips in the event of a failure. As a part of the fix. 1.) moved the function removeDhcpServiceInsubnet and listLastNicsInSubnet to NetworkManager (These were in VirtualMachineManagerImpl earlier.) 2.) add the call to clean ipAlias in the remove nic function of the networkManager. 3.) Modified the removeDhcpServiceInsubnet to take network ad an argument. This will help in removing only the ipAlias which belong to a particular network. Diffs - server/src/com/cloud/network/NetworkManager.java dab7a13 server/src/com/cloud/network/NetworkManagerImpl.java effee96 server/src/com/cloud/vm/VirtualMachineManagerImpl.java b33ee49 server/test/com/cloud/network/MockNetworkManagerImpl.java 7f34e55 server/test/com/cloud/vpc/MockNetworkManagerImpl.java 178ebba Diff: https://reviews.apache.org/r/13323/diff/ Testing --- Tested on 4.2. created two guest networks guest1 and guest2. created VMS in both the networks. Added a VM(with no PV drivers) from guest1 to guest2. this failed and on failure ipAlias configured as part of nic creation was removed. Deleting the vm causes all the removal of all ipAliases from all the subnets in which this is the lastvm. Thanks, bharat kumar
Re: Review Request 12849: added backwards compatibility code to Networks enums
Hi Dave, I see what you mean by colons wouldn't be allowed. Nothing changed but I went for this solution for consistency (over backward compatibility). I will change again. Let me get to work first. nice working around the globe;) Daan On Wed, Aug 7, 2013 at 3:57 AM, Sheng Yang sh...@yasker.org wrote: This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/12849/ On August 6th, 2013, 11:40 p.m. UTC, *Sheng Yang* wrote: api/src/com/cloud/network/Networks.javahttps://reviews.apache.org/r/12849/diff/5/?file=337785#file337785line135 (Diff revision 5) public String scheme() { 95 // do we need to check that value does not contain a scheme 135 return new URI(scheme,value.toString(),null,null); I think by following the previous case, it should be URI(scheme,value.toString()) rather than null, null. It possibly doesn't make difference, but it's better not to use different constructor. The same with several URI constructor above. And I doubt if it's really need to discard this check. At least they would working for other broadcast domain type. On August 7th, 2013, 1:45 a.m. UTC, *Sheng Yang* wrote: Dave explained this. Please drop this issue. On August 7th, 2013, 1:54 a.m. UTC, *Dave Cahill* wrote: Thanks for the quick reply Sheng! Actually, I dug deeper and realized that if we use the 4-parameter URI constructor, the value parameter is interpreted as host, and colons wouldn't be allowed! It looks like the only way to preserve the behavior as it was is to stick with the old string concatenation behavior of scheme + :// + value. In fact, this is what Daan and I agreed on in IRC yesterday, but it looks like he went with the 4 param constructor - I'll discuss with him later to see what changed. Thank you Dave! - Sheng On August 6th, 2013, 2:25 p.m. UTC, daan Hoogland wrote: Review request for cloudstack, Chiradeep Vittal, Dave Cahill, Koushik Das, and Sheng Yang. By daan Hoogland. *Updated Aug. 6, 2013, 2:25 p.m.* *Repository: * cloudstack-git Description Both BroadcastDomainType and IsolationType needed some extra code for backwards compatibility Diffs - api/src/com/cloud/network/Networks.java (c76c3d4) - api/test/com/cloud/network/NetworksTest.java (31114e8) View Diff https://reviews.apache.org/r/12849/diff/
Re: Review Request 12849: added backwards compatibility code to Networks enums
Sheng, Some of this code will be used in a feature allowing vpc gateways to be plugged to sdn networks and to allow multiple gateways to share that network. I am using that in a hack on 4.1.1 and porting the code to master now, bit by bit so to speak. This is why you see some functions that seem not to be used. regards, Daan On Wed, Aug 7, 2013 at 8:52 AM, Daan Hoogland daan.hoogl...@gmail.comwrote: Hi Dave, I see what you mean by colons wouldn't be allowed. Nothing changed but I went for this solution for consistency (over backward compatibility). I will change again. Let me get to work first. nice working around the globe;) Daan On Wed, Aug 7, 2013 at 3:57 AM, Sheng Yang sh...@yasker.org wrote: This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/12849/ On August 6th, 2013, 11:40 p.m. UTC, *Sheng Yang* wrote: api/src/com/cloud/network/Networks.javahttps://reviews.apache.org/r/12849/diff/5/?file=337785#file337785line135 (Diff revision 5) public String scheme() { 95 // do we need to check that value does not contain a scheme 135 return new URI(scheme,value.toString(),null,null); I think by following the previous case, it should be URI(scheme,value.toString()) rather than null, null. It possibly doesn't make difference, but it's better not to use different constructor. The same with several URI constructor above. And I doubt if it's really need to discard this check. At least they would working for other broadcast domain type. On August 7th, 2013, 1:45 a.m. UTC, *Sheng Yang* wrote: Dave explained this. Please drop this issue. On August 7th, 2013, 1:54 a.m. UTC, *Dave Cahill* wrote: Thanks for the quick reply Sheng! Actually, I dug deeper and realized that if we use the 4-parameter URI constructor, the value parameter is interpreted as host, and colons wouldn't be allowed! It looks like the only way to preserve the behavior as it was is to stick with the old string concatenation behavior of scheme + :// + value. In fact, this is what Daan and I agreed on in IRC yesterday, but it looks like he went with the 4 param constructor - I'll discuss with him later to see what changed. Thank you Dave! - Sheng On August 6th, 2013, 2:25 p.m. UTC, daan Hoogland wrote: Review request for cloudstack, Chiradeep Vittal, Dave Cahill, Koushik Das, and Sheng Yang. By daan Hoogland. *Updated Aug. 6, 2013, 2:25 p.m.* *Repository: * cloudstack-git Description Both BroadcastDomainType and IsolationType needed some extra code for backwards compatibility Diffs - api/src/com/cloud/network/Networks.java (c76c3d4) - api/test/com/cloud/network/NetworksTest.java (31114e8) View Diff https://reviews.apache.org/r/12849/diff/
Re: CONSOLE access via API
Hi, my goal is obtain an url (API generated) to open directly the Console from my custom Cloud interface... Thanks. Regards. Il 06/08/2013 21:28, Musayev, Ilya ha scritto: Antonio, Would you explain what the end goal is? Console access is token based and works directly with your browser and java plugin, I'm not certain what you would achieve by executing console access command in command line. Regards ilya -Original Message- From: Antonio Petrocelli [mailto:a.petroce...@netsons.com] Sent: Tuesday, August 06, 2013 3:32 AM To: dev@cloudstack.apache.org Subject: CONSOLE access via API Hello, Under CloudStack UI interface, the console link is: http://xx.xx.xx.xx:8080/client/console?cmd=accessvm=x I'm trying to use signiture under API, all other commands works, but I'm trying to get VM console, just not work: http://xx.xx.xx.xx:8080/client/console?cmd=accessvm=xsignatu re=XX or http://xx.xx.xx.xx:8080/client/api?cmd=accessvm=xsignature=X X Could any friend point me correct command for console access through API? Thanks! smime.p7s Description: Firma crittografica S/MIME
Re: Breaking docs out
ok, +0 then, I'm off for a coffee. On Tue, Aug 6, 2013 at 10:21 PM, Jessica Tomechak jessica.tomec...@citrix.com wrote: I'm +0 on it. Don't really mind either way. Jessica T. From: Radhika Puthiyetath [radhika.puthiyet...@citrix.com] Sent: Monday, August 05, 2013 9:25 PM To: dev@cloudstack.apache.org Subject: RE: Breaking docs out Make perfect sense. + 1 -Radhika -Original Message- From: David Nalley [mailto:da...@gnsa.us] Sent: Tuesday, August 06, 2013 3:13 AM To: dev@cloudstack.apache.org Subject: Breaking docs out Hi folks: I'd like to propose breaking out a nuymber of our documents into their own repos. My thinking is that specifically; the release notes, midonet, and niciranvp documentation shares very little with the rest of the documentation, and should be broken out akin to how the QIG is currently broken out. The particular problem I am trying to solve is to deal with publishing. For instance, even though the release notes are contained in just a few xml documents, it copies content from every single xml file in thd directory - over 400 - and it also copies those up to the website. Splitting things up also allows us to prioritize l10n. Right now, we just dump 400 xml files worth of content into transifex and people translate away - they can't put a priority on release notes, or de-emphasize more esoteric documentation like Nicira or Midonet. Eventually I'd like to break out each of the individual guides into their own document - separate from the other. Right now they carry a ton of similar content so that isn't very practical; but it's what I am thinking, perhaps for 4.4 or 4.5. In the meantime, I'd like to make this change as soon as we think we have documentation pretty close to done for 4.2 to minimize the disruptive effects. Thoughts, comments, flames? --David
Re: Help needed with Cloudstack Setup
Hi Marcus, I wonder could we take the most recent Jenkins-built System VM template and get it hosted somewhere, then let people know the URL? We could then replace the old acton systemvm URL [1] with the new URL wherever it appears in docs etc. I'd imagine the template doesn't change all that much from day to day, and anything would be better than people using the old acton System VM template. I know people can just navigate to Jenkins, find the most recent image, and download it. However, the URLs become invalid within days (maybe less?), so they have to go search for it every time, and can't build repeatable dev-environment setup scripts. What do you think? Thanks, Dave. [1] Old template URL: http://download.cloud .com/templates/acton/acton-systemvm-02062012.qcow2.bz2 On Wed, Aug 7, 2013 at 5:55 AM, Marcus Sorensen shadow...@gmail.com wrote: 4.2.0-SNAPSHOT -- this will require a newer system vm than the one you installed. see http://jenkins.cloudstack.org/view/master/job/build-systemvm-master/ On Tue, Aug 6, 2013 at 2:52 PM, Marcus Sorensen shadow...@gmail.com wrote: Also, what version are you deploying? 4.1.1? On Tue, Aug 6, 2013 at 2:48 PM, Marcus Sorensen shadow...@gmail.com wrote: Any reason why you aren't using the developer packages? I don't see any info regarding the agent install. Did you also set your KVM traffic labels for the various networks (I'm assuming cloubr0 and cloudbr1 are for management, public, guest bridges, but not sure which is which, management is 3200, maybe?) whats the 3100 addr for? On Tue, Aug 6, 2013 at 2:41 PM, Chiradeep Vittal chiradeep.vit...@citrix.com wrote: Why would your primary (local) storage have a different IP than host-corporate-ip? There is also a flag system.vm.use.local.storage which tells CloudStack to use local storage for system vms. From: Sarath Chandra sarath.cloudst...@gmail.commailto: sarath.cloudst...@gmail.com Reply-To: dev@cloudstack.apache.orgmailto:dev@cloudstack.apache.org dev@cloudstack.apache.orgmailto:dev@cloudstack.apache.org Date: Tuesday, August 6, 2013 4:33 AM To: dev@cloudstack.apache.orgmailto:dev@cloudstack.apache.org dev@cloudstack.apache.orgmailto:dev@cloudstack.apache.org Subject: Help needed with Cloudstack Setup Hi All, I am doing a fresh cloudstack installation and am not able to instantiate any VMs (both system guest vms). I am seeing the following errors: ERROR [network.element.SspElement] (consoleproxy-1:ctx-6d22b00c) ReservationContext was null for NicProfile[2-1-null-null-null Ntwk[202|Control|3] ERROR [network.element.SspElement] (secstorage-1:ctx-3f99cd9c) ReservationContext was null for NicProfile[84-22-null-10.20.30.11-null Ntwk[200|Public|1] ERROR [network.element.SspElement] (consoleproxy-1:ctx-6d22b00c) ReservationContext was null for NicProfile[3-1-null-null-null Ntwk[201|Management|2] WARN [cloud.consoleproxy.ConsoleProxyManagerImpl] (consoleproxy-1:ctx-6d22b00c) Exception while trying to start console proxy com.cloud.exception.InsufficientServerCapacityException: Unable to create a deployment for VM[ConsoleProxy|v-1-VM]Scope=interface com.cloud.dc.DataCenter; id=1 at a later place com.cloud.utils.exception.CloudRuntimeException: Failed to find template routing-3 in VMTemplateStoragePool I suspect my configuration is wrong some where. Can anybody kindly check and confirm if my configuration is right. If its right, then please point me to the likely cause of this issue. I have attached a picture to this mail which shows my lab setup. To give a brief, the following is my setup: Server -- Switch 1 -- Switch 2 KVM Hypervisor The switches allow all Vlan traffic to pass through them (basically trunk all and have added vlans 3100,3200,3300 to both the switches) On my server I have done the following: * Ubuntu 13.04 OS * Started NFS and exported to /mnt/primary (not using this) and /mnt/secondary * Installed the System template by executing the command : cloud-install-sys-tmplt -m /mnt/secondary -u http://download.cloud.com/templates/acton/acton-systemvm-02062012.qcow2.bz2-h kvm -o localhost -r root -d * $ mvn clean install -P developer,systemvm * $ mvn -P developer -pl developer -Ddeploydb (Deployed the db) * $ mvn -pl :cloud-client-ui jetty:run (Started the management server) On my hypervisor I have the following: * Ubuntu 13.04 OS * KVM as hypervisor Server Interface configuration: eth0 : Corporate Lab network ip (lets call it as server-corporate-ip) eth1.3200 : 172.16.10.1 (assigned during CS setup) [eth1.3200 is a subinterface on eth1] Global Settings: system.vm.use.local.storage = true Zone Setup details: DNS 18.8.8.8 DNS 28.8.4.4 Internal DNS 1172.16.10.1 Guest CIDR10.1.1.0/24http://10.1.1.0/24 Local storage enabledYes Pod
CloudStack Storage Plugin for CloudByte
Hi, I will be working on implementing the CS storage plugin for CloudByte's Elastistor. This new plugin should be based on the newly refactored storage plugin changes which has gone into 4.2 release already. Which CS release can accommodate this new storage plugin feature ? Regards, Amit *CloudByte Inc.* http://www.cloudbyte.com/
Re: Review Request 13316: CLOUDSTACK-3723: System VM Template for XenServer has wrong guest OS type Information on the Database.
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13316/#review24780 --- Commit 23aaeafd21aa667a82dca813e7314ab60a6b2ac7 in branch refs/heads/4.2 from Sanjay Tripathi [ https://git-wip-us.apache.org/repos/asf?p=cloudstack.git;h=23aaeaf ] CLOUDSTACK-3723: System VM Template for XenServer has wrong guest OS type Information on the Database Also added the debian 7 to debian 6 mapping in CitrixHelper file to support the debian 7 template in XS 6.0, 6.0.2, 6,1 etc. - ASF Subversion and Git Services On Aug. 6, 2013, 9:14 a.m., Sanjay Tripathi wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13316/ --- (Updated Aug. 6, 2013, 9:14 a.m.) Review request for cloudstack, Abhinandan Prateek and Devdeep Singh. Bugs: CLOUDSTACK-3723 Repository: cloudstack-git Description --- CLOUDSTACK-3723: System VM Template for XenServer has wrong guest OS type Information on the Database. Made the correction in the DB entry to point to Debian 7.0 template instead of Debain 6.0. Diffs - plugins/hypervisors/xen/src/com/cloud/hypervisor/xen/resource/CitrixHelper.java 9cdbd84 setup/db/db/schema-410to420.sql f98bb50 setup/db/templates.sql ba13841 Diff: https://reviews.apache.org/r/13316/diff/ Testing --- Verifed the fix locally on cloudstack environment: 1. Created CS environment successfully. 2. SystemVMs are coming up fine. 3. UserVMs are getting deployed successfully. Thanks, Sanjay Tripathi
RE: CloudStack Storage Plugin for CloudByte
The next release will probably be 4.3. You can target the CloudByte plugin for that release. The can find more information on CloudStack release cycle here [1]. [1] https://cwiki.apache.org/confluence/display/CLOUDSTACK/Releases Regards, Devdeep -Original Message- From: Amit Das [mailto:amit@cloudbyte.com] Sent: Wednesday, August 07, 2013 12:51 PM To: dev@cloudstack.apache.org Subject: CloudStack Storage Plugin for CloudByte Hi, I will be working on implementing the CS storage plugin for CloudByte's Elastistor. This new plugin should be based on the newly refactored storage plugin changes which has gone into 4.2 release already. Which CS release can accommodate this new storage plugin feature ? Regards, Amit *CloudByte Inc.* http://www.cloudbyte.com/
Re: Help needed with Cloudstack Setup
I think Citrix has done this, at least that's why comes out of the db now: | http://download.cloud.com/templates/4.2/systemvmtemplate-2013-06-12-master-xen.vhd.bz2 | | http://download.cloud.com/templates/4.2/systemvmtemplate-2013-06-12-master-kvm.qcow2.bz2 | | http://download.cloud.com/templates/4.2/systemvmtemplate-4.2-vh7.ova | On Wed, Aug 07, 2013 at 04:17:26PM +0900, Dave Cahill wrote: Hi Marcus, I wonder could we take the most recent Jenkins-built System VM template and get it hosted somewhere, then let people know the URL? We could then replace the old acton systemvm URL [1] with the new URL wherever it appears in docs etc. I'd imagine the template doesn't change all that much from day to day, and anything would be better than people using the old acton System VM template. I know people can just navigate to Jenkins, find the most recent image, and download it. However, the URLs become invalid within days (maybe less?), so they have to go search for it every time, and can't build repeatable dev-environment setup scripts. What do you think? Thanks, Dave. [1] Old template URL: http://download.cloud .com/templates/acton/acton-systemvm-02062012.qcow2.bz2 On Wed, Aug 7, 2013 at 5:55 AM, Marcus Sorensen shadow...@gmail.com wrote: 4.2.0-SNAPSHOT -- this will require a newer system vm than the one you installed. see http://jenkins.cloudstack.org/view/master/job/build-systemvm-master/ On Tue, Aug 6, 2013 at 2:52 PM, Marcus Sorensen shadow...@gmail.com wrote: Also, what version are you deploying? 4.1.1? On Tue, Aug 6, 2013 at 2:48 PM, Marcus Sorensen shadow...@gmail.com wrote: Any reason why you aren't using the developer packages? I don't see any info regarding the agent install. Did you also set your KVM traffic labels for the various networks (I'm assuming cloubr0 and cloudbr1 are for management, public, guest bridges, but not sure which is which, management is 3200, maybe?) whats the 3100 addr for? On Tue, Aug 6, 2013 at 2:41 PM, Chiradeep Vittal chiradeep.vit...@citrix.com wrote: Why would your primary (local) storage have a different IP than host-corporate-ip? There is also a flag system.vm.use.local.storage which tells CloudStack to use local storage for system vms. From: Sarath Chandra sarath.cloudst...@gmail.commailto: sarath.cloudst...@gmail.com Reply-To: dev@cloudstack.apache.orgmailto:dev@cloudstack.apache.org dev@cloudstack.apache.orgmailto:dev@cloudstack.apache.org Date: Tuesday, August 6, 2013 4:33 AM To: dev@cloudstack.apache.orgmailto:dev@cloudstack.apache.org dev@cloudstack.apache.orgmailto:dev@cloudstack.apache.org Subject: Help needed with Cloudstack Setup Hi All, I am doing a fresh cloudstack installation and am not able to instantiate any VMs (both system guest vms). I am seeing the following errors: ERROR [network.element.SspElement] (consoleproxy-1:ctx-6d22b00c) ReservationContext was null for NicProfile[2-1-null-null-null Ntwk[202|Control|3] ERROR [network.element.SspElement] (secstorage-1:ctx-3f99cd9c) ReservationContext was null for NicProfile[84-22-null-10.20.30.11-null Ntwk[200|Public|1] ERROR [network.element.SspElement] (consoleproxy-1:ctx-6d22b00c) ReservationContext was null for NicProfile[3-1-null-null-null Ntwk[201|Management|2] WARN [cloud.consoleproxy.ConsoleProxyManagerImpl] (consoleproxy-1:ctx-6d22b00c) Exception while trying to start console proxy com.cloud.exception.InsufficientServerCapacityException: Unable to create a deployment for VM[ConsoleProxy|v-1-VM]Scope=interface com.cloud.dc.DataCenter; id=1 at a later place com.cloud.utils.exception.CloudRuntimeException: Failed to find template routing-3 in VMTemplateStoragePool I suspect my configuration is wrong some where. Can anybody kindly check and confirm if my configuration is right. If its right, then please point me to the likely cause of this issue. I have attached a picture to this mail which shows my lab setup. To give a brief, the following is my setup: Server -- Switch 1 -- Switch 2 KVM Hypervisor The switches allow all Vlan traffic to pass through them (basically trunk all and have added vlans 3100,3200,3300 to both the switches) On my server I have done the following: * Ubuntu 13.04 OS * Started NFS and exported to /mnt/primary (not using this) and /mnt/secondary * Installed the System template by executing the command : cloud-install-sys-tmplt -m /mnt/secondary -u http://download.cloud.com/templates/acton/acton-systemvm-02062012.qcow2.bz2-h kvm -o localhost -r root -d * $ mvn clean install -P developer,systemvm * $ mvn -P developer -pl developer -Ddeploydb (Deployed the db) * $ mvn -pl :cloud-client-ui jetty:run (Started the management server)
Re: Review Request 12849: added backwards compatibility code to Networks enums
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/12849/ --- (Updated Aug. 7, 2013, 8:50 a.m.) Review request for cloudstack, Chiradeep Vittal, Dave Cahill, Koushik Das, and Sheng Yang. Changes --- did IsolationType forgot BroadcastDomainType. fixed Repository: cloudstack-git Description --- Both BroadcastDomainType and IsolationType needed some extra code for backwards compatibility Diffs (updated) - api/src/com/cloud/network/Networks.java c76c3d4 api/test/com/cloud/network/NetworksTest.java 31114e8 Diff: https://reviews.apache.org/r/12849/diff/ Testing --- Thanks, daan Hoogland
Re: Review Request 12849: added backwards compatibility code to Networks enums
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/12849/#review24787 --- Three minor comments, but my main concerns are addressed by this diff. api/src/com/cloud/network/Networks.java https://reviews.apache.org/r/12849/#comment48907 Should this say isolation URI instead of broadcast URI? api/test/com/cloud/network/NetworksTest.java https://reviews.apache.org/r/12849/#comment48905 Not important, but I think this should be of broadcasttype lswitch. api/test/com/cloud/network/NetworksTest.java https://reviews.apache.org/r/12849/#comment48906 Not important, but I think this should be of broadcasttype mido. - Dave Cahill On Aug. 7, 2013, 8:50 a.m., daan Hoogland wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/12849/ --- (Updated Aug. 7, 2013, 8:50 a.m.) Review request for cloudstack, Chiradeep Vittal, Dave Cahill, Koushik Das, and Sheng Yang. Repository: cloudstack-git Description --- Both BroadcastDomainType and IsolationType needed some extra code for backwards compatibility Diffs - api/src/com/cloud/network/Networks.java c76c3d4 api/test/com/cloud/network/NetworksTest.java 31114e8 Diff: https://reviews.apache.org/r/12849/diff/ Testing --- Thanks, daan Hoogland
Re: weird test failure
I under-kinda-stand, Does this mean I should live with the whole build failing for a while once in a while? Meaning should I change this test and submit a patch or is there some windows or java setting I should look at? On Tue, Aug 6, 2013 at 8:19 PM, Chiradeep Vittal chiradeep.vit...@citrix.com wrote: Some insight here http://www.javatuning.com/why-is-thread-sleep-inherently-inaccurate/ On 8/6/13 8:16 AM, Daan Hoogland daan.hoogl...@gmail.com wrote: H, I just had a time skew that seems impossible: if in the folowing test the sleep is 1000 it reports a failure saying the duration was 999. now I see at least some clock ticks around it so whats heppening? An overeager optimizer, maybe? I had to set it to 1001 to make it work. This is not new code, it has been there since jan 24! So i ran it a few times. public class TestProfiler extends Log4jEnabledTestCase { protected final static Logger s_logger = Logger.getLogger(TestProfiler.class); @Test public void testProfiler() { s_logger.info(testProfiler() started); Profiler pf = new Profiler(); pf.start(); try { Thread.sleep(1001); } catch (InterruptedException e) { } pf.stop(); s_logger.info(Duration : + pf.getDuration()); Assert.assertTrue(pf.getDuration() = 1000); s_logger.info(testProfiler() stopped); } } any clue welcome, Daan
Re: IP Address Allocation
things are falling into place in slow-motion. Thanks Alex! And Soheil for bringing this up. Let's have a hack session on this at the next CCC. On Tue, Aug 6, 2013 at 10:10 PM, Alex Huang alex.hu...@citrix.com wrote: I'll say it's a bad design. He he I designed it. :( We had to make some compromises due to time constraints. I always intended to move the ip allocation into NetworkGurus and actually attempted to several times but, at the end for various reasons, we just couldn't complete that task. I generally believe that ip address allocation can actually be done even without CloudStack's database, for example, if it came from a master dhcp server or some IT central database. So an interface like IpAddressReserver to abstract that functionality would be great. I have to say the number of implementations for NetworkGuru grew faster than I expected. In one sense, it's a good sign that everyone was able to use it but then it exposes the work that was incomplete to begin with. IIRC, there's basically two problems I left on NetworkGuru to be better designed. One is this ip address allocation. The other problem is how to identify which set of NetworkGurus should manage the network. I believe that selection process should be much more transparent to the administrator who setup the CS deployment but because at the time only VLAN isolation was supported, I did not put that into CS. As more and more SDN technologies come online, it's best we design that process. I think Hugo made some changes there but I haven't had a chance to study how that all works now in 4.x yet. There are some artifacts in the code to show I was thinking about this. For example, Network and Nic carries the Isolation URI instead of just VLAN tag because it was intended to be cast by the NetworkGuru and then transported to the hypervisor. We expected anyone who plugs into CloudStack have two places to plug into: orchestration and actual provisioning. So the NetworkGuru forms the URI for isolation and then it has code at the hypervisor side to resolve the URI to something that the hypervisor can understand. --Alex -Original Message- From: Soheil Eizadi [mailto:seiz...@infoblox.com] Sent: Tuesday, August 6, 2013 11:54 AM To: dev@cloudstack.apache.org Subject: RE: IP Address Allocation Agree that the IP Address allocation is different than the DHCP and DNS service offered by the NetworkElements. The third party NetworkGurus just reference the standard NetworkManager IP Address Allocation routines, either directly (e.g. _networkMgr.assignPublicIpAddress()) or by calling out to the GuestNetworkGuru. I would say in the current design the Network Manager manages the IP Addresses not the Network Gurus. I don't think this is a bad design, in that the NetworkGurus are sharing the same ip address allocation and get new features (e.g. IP reservation or VM migration) without implementing new code to support them. -Soheil From: Alex Huang [alex.hu...@citrix.com] Sent: Tuesday, August 06, 2013 11:27 AM To: Chiradeep Vittal; dev@cloudstack.apache.org Subject: RE: IP Address Allocation We can either break NetworkManager down further and say NetworkGuru sets the actual parameters of the network and a new IpAddressReserver sets the ip address or we can let NetworkGurus identify the IpAddressReserver to use themselves. I generally think NetworkManager shouldn't care where the ip address came from because it really shouldn't know anything about the actual network we're laying out for the end user. That's always been left to the NetworkGurus. If you think about it from that perspective, then it's natural that ip address is allocated by the NetworkGuru. But I can see different NetworkGurus may want to share the same ip address allocation tables/schemes so an interface for them to share that code does make sense. Note there's a difference between allocating ip address and the act of actually issuing the ip address and associating the ip address to a fqdn. That act is performed by the network elements. Here we're specifically saying who is responsible for figuring out the ip address that can be allocated to a particular vm and works within a network. --Alex -Original Message- From: Chiradeep Vittal Sent: Tuesday, August 6, 2013 11:14 AM To: dev@cloudstack.apache.org Cc: Alex Huang Subject: Re: IP Address Allocation The way the gurus are consulted for network design and reservation makes it difficult to put this inside a guru. However it does seem odd to put this inside an element. Perhaps we need a 3rd type (IpAddressReserver?) On 8/6/13 10:42 AM, Soheil Eizadi seiz...@infoblox.com wrote: The NetworkGuru does a lot more than IP address allocation, e.g. handle the Layer2/3 networking unique to the particular vendor. There are a lot of NetworkElements
Re: api calls in 4.1.1
will do, to busy now but is bound to get priority without further pushing from my side:| On Wed, Aug 7, 2013 at 1:37 AM, Musayev, Ilya imusa...@webmd.net wrote: Daan, Take this back - the error was on my part it seems (I've used cloudmonkey with api/secret keys for another test user). Once I fixed the secret and api keys to a proper user that had vms, I would see the vms entitled to that user. If you do come around on specifics of this issue, please let me know. Thanks ilya -Original Message- From: Daan Hoogland [mailto:daan.hoogl...@gmail.com] Sent: Tuesday, August 06, 2013 5:31 PM To: dev Subject: Re: api calls in 4.1.1 H Ilya, I am not at the bottom of this yet. It seems the api no longer returns the full list hence a certain key can not be found in the result set. We are looking at it (between other work). If this is the case a test scenario would need a big set of VMs:( regards, On Tue, Aug 6, 2013 at 10:16 PM, Musayev, Ilya imusa...@webmd.net wrote: Daan, The commit CLOUDSTACK-3080 seems like the only change that was made - that would be related. I'm about to try this out and find it - if that's what broke it. Thanks ilya -Original Message- From: Daan Hoogland [mailto:daan.hoogl...@gmail.com] Sent: Monday, August 05, 2013 2:26 PM To: dev Subject: Re: api calls in 4.1.1 Ilya, listVirtualMachines (as user) for one, but I got a second report that wasn't very precise yet, so I will be working on reproducing next days and report back (wit test case or patch) regards, On Mon, Aug 5, 2013 at 6:56 PM, Chip Childers chip.child...@sungard.com wrote: On Mon, Aug 05, 2013 at 10:42:53AM +0200, Daan Hoogland wrote: H, What is the status of 4.1.1 at this moment. I am getting several reports of api calls not working as before. I saw the vote passing by but didn't give them much attention. In jira I don't see much mention of such issues. Is this noticed by anybosy else? regards, Daan 4.1.1 status is that we are working on pushing the docs, and then the release will be announced. -chip
Re: Unable to create image on Cloudcentral
Prabakaran, Can you make 'I got null pointer exception' more explicit? regards, On Thu, Jul 25, 2013 at 12:56 PM, Prabakaran R praba...@cusdelight.com wrote: Hi, I am trying to create image of virtual machine in cloudcentral using JClouds. Following is my code snippet CreateTemplateOptions templateOption = CreateTemplateOptions.NONE; templateOption.bits(64); templateOption.passwordEnabled(true); TemplateMetadata meta = TemplateMetadata.builder().name(Test Debian 6.0.6).osTypeId(26226f97-25c0-49e4-be29-03f13308ade8). displayText(Test Debian 6.0.6 64-bit).virtualMachineId(---).build(); template.createTemplate(meta,templateOption); But I got null pointer exception. VM is in stop state, kindly help me to this. Thanks, Prabakaran.R
Re: zone create wizard issue with create storagepool
Talluri, did you check the history of the files involved in creating a zone? regards, On Fri, Jul 26, 2013 at 4:57 AM, Srikanteswararao Talluri srikanteswararao.tall...@citrix.com wrote: Createstoragepool fails during zone creation while creating zone using zone create wizard. This is happening because add host is taking too long. Is there a recent change in this area? Filed a bug to track this issue https://issues.apache.org/jira/browse/CLOUDSTACK-3839 Thanks, ~Talluri
Re: Force delete domain through test case
On Fri, Aug 2, 2013 at 4:57 PM, Gaurav Aradhye gaurav.arad...@clogeny.com wrote: While forcefully deleting a domain through test case, it fails. The resources in the domain and the domain itself gets cleaned up, but the code fails with following error. After cleaning up the resources, it says can't find the specific domain. Gaurav, as you describe it it sounds as if you are trying to delete the domain twice. Does this make sense? regards, Daan
Re: Review Request 12098: Add PCI Passthrough to CloudStack on KVM Hypervisor
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/12098/#review24788 --- I will rebase it in a few days. This patch only includes support for KVM, but the concepts used can apply for all hypervisors supporting PCI Passthrough so additional patches can add support for XenServer. (Just to be clear, I'm not planning on adding support for other hypervisors myself.) When used with other hypervisors, any attempt to start any VM requesting PCI Passthrough will gracefully fail in planning step as no host will be able to provide the request PCI devices. - Pawit Pornkitprasan On July 2, 2013, 7:12 a.m., Pawit Pornkitprasan wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/12098/ --- (Updated July 2, 2013, 7:12 a.m.) Review request for cloudstack. Repository: cloudstack-git Description --- This patch adds PCI Passthrough support to CloudStack on the KVM Hypervisor as previously mentioned on the mailing list PCI Passthrough may be useful for high-performance networking devices or other hardware accelerators. This patch does not care what kind of the device it is or attempt to configure it, leaving it to the guest. Because the devices available at each agent may be different, the available devices for passthrough are configured at the agent configuration file (/etc/cloudstack/agent/agent.properties). Configuration is a comma separated list of available PCI devices and its given name. pci.devices=28:00.1|10GE,28:00.2|10GE,28:00.3|10GE,28:00.4|10GE,28:00.5|10GE,28:00.6|10GE,28:00.7|10GE,28:01.0|10GE At agent startup, the list of PCI devices is parsed and sent together with StartupRoutingCommand (in a new field, not in details). The management server then stores it in a new table “op_host_pci_devices”. If a device is added, removed, or renamed, the table is updated accordingly. The agent also probes libvirt which devices are currently in-use and send it with the command in attempt to synchronize any potential difference between the database and the actual status. The “name” of the PCI device is what is used to assign a device. In a compute offering, the user can specify the name of one or more PCI devices (as a comma-separated list) and CloudStack will find a host with the PCI device of the specified name available and assign it. A new manager, PciDeviceManager, is created to handle the allocation of PCI device. The manager implements StateListener and assigns PCI devices on state change to “starting” and also release the devices VM stop. First fit allocator and first fit planner are also modified to check for PCI device availability accordingly. For migration, intrusive migration is implementated. The PCI device is detached before migration and attached it again at the target host. This will interrupt whatever is using the device on the VM. However, it may be desirable for networking devices where the VM can use a bonding device to channel network traffic through the normal virtualized network device while the PCI Passthrough device is down. Diffs - api/src/com/cloud/agent/api/to/PciDeviceTO.java PRE-CREATION api/src/com/cloud/agent/api/to/VirtualMachineTO.java e6240ff api/src/com/cloud/offering/ServiceOffering.java 45d5f38 api/src/com/cloud/pci/PciDevice.java PRE-CREATION api/src/org/apache/cloudstack/api/ApiConstants.java 8ac7c3f api/src/org/apache/cloudstack/api/command/admin/offering/CreateServiceOfferingCmd.java decac29 api/src/org/apache/cloudstack/api/response/ServiceOfferingResponse.java 288f76b client/WEB-INF/classes/resources/messages.properties ad8d29d client/tomcatconf/applicationContext.xml.in ac1f3e4 core/src/com/cloud/agent/api/AttachPciDevicesCommand.java PRE-CREATION core/src/com/cloud/agent/api/StartupRoutingCommand.java 5961ab0 engine/schema/src/com/cloud/host/HostPciDeviceVO.java PRE-CREATION engine/schema/src/com/cloud/host/HostVO.java c814b94 engine/schema/src/com/cloud/host/dao/HostDaoImpl.java 810b973 engine/schema/src/com/cloud/host/dao/HostPciDevicesDao.java PRE-CREATION engine/schema/src/com/cloud/host/dao/HostPciDevicesDaoImpl.java PRE-CREATION engine/schema/src/com/cloud/migration/ServiceOffering21VO.java 7a49e63 engine/schema/src/com/cloud/service/ServiceOfferingVO.java 9a262c5 plugins/hypervisors/kvm/src/com/cloud/hypervisor/kvm/resource/LibvirtComputingResource.java 60cde8c plugins/hypervisors/kvm/src/com/cloud/hypervisor/kvm/resource/LibvirtDomainXMLParser.java a283768 plugins/hypervisors/kvm/src/com/cloud/hypervisor/kvm/resource/LibvirtVMDef.java 5120870
Review Request 13373: keep the existing changes in cloud-set-guest-sshkey.in to make sure we do not break the backward compatibility.
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13373/ --- Review request for cloudstack and Jayapal Reddy. Bugs: Cloudstack-4138 Repository: cloudstack-git Description --- keep the existing changes in cloud-set-guest-sshkey.in to make sure we do not break the backward compatibility. Diffs - setup/bindir/cloud-set-guest-sshkey.in 9436b03 Diff: https://reviews.apache.org/r/13373/diff/ Testing --- Tested on 4.2 Thanks, bharat kumar
Re: Review Request 13325: preserve data in iso_id column while upgrading form 3.0.7 to 4.2
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13325/#review24789 --- setup/db/db/schema-307to410-cleanup.sql https://reviews.apache.org/r/13325/#comment48908 this should be moved to Upgrade41-42.java. nay changes to iso_id in later releases will be lost otherwise - Kishan Kavala On Aug. 6, 2013, 6:41 p.m., bharat kumar wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13325/ --- (Updated Aug. 6, 2013, 6:41 p.m.) Review request for cloudstack and Kishan Kavala. Bugs: Cloudstack-4100 Repository: cloudstack-git Description --- preserve data in iso_id column while upgrading form 3.0.7 to 4.2 Earlier was dropping the iso_id column as a result we loose data from the iso_id column prior to upgrade. Now fixed this by. renaming the iso_id to iso_id1 in schema-307to401 file. and then after upgrade dropping the iso_id column and renaming the iso_id1 to iso_id in 307to401 cleanup schema. Diffs - setup/db/db/schema-307to410-cleanup.sql 6a9e2af setup/db/db/schema-307to410.sql af296ab Diff: https://reviews.apache.org/r/13325/diff/ Testing --- tested the upgrade from 3.0.7 to 4.2. Thanks, bharat kumar
Review Request 13374: Automation: Affinity Groups: Removing invalid test case
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13374/ --- Review request for cloudstack and Prasanna Santhanam. Repository: cloudstack-git Description --- Removing invalid test case. Affinity group can be deleted when vms are still using it. This can be reproduced from UI. Diffs - test/integration/component/test_affinity_groups.py 309adc6 Diff: https://reviews.apache.org/r/13374/diff/ Testing --- Thanks, Gaurav Aradhye
RE: [Doc] IP Address Reservation is Ready for Review
Thanks Saksham. They have been incorporated. The guide is refreshed and posted to https://issues.apache.org/jira/browse/CLOUDSTACK-850. From: Saksham Srivastava Sent: Wednesday, August 07, 2013 3:06 PM To: Radhika Puthiyetath; us...@cloudstack.apache.org; dev@cloudstack.apache.org; Abhinav Roy Subject: RE: [Doc] IP Address Reservation is Ready for Review Hi Radhika, two comments: * If you decide to reserve IP ranges for non-CloudStack purposes, you can specify a part of the IP address range or the CIDR that should only be allocated by... Remove IP address range as we only specify the CIDR, not the range. * Please mention that feature is not supported in shared networks. Thanks, Saksham From: Radhika Puthiyetath Sent: Friday, August 02, 2013 5:44 PM To: us...@cloudstack.apache.orgmailto:us...@cloudstack.apache.org; dev@cloudstack.apache.orgmailto:dev@cloudstack.apache.org; Saksham Srivastava; Abhinav Roy Subject: [Doc] IP Address Reservation is Ready for Review Hi, IP Reservation documentation is ready for review. The doc is attached at https://issues.apache.org/jira/browse/CLOUDSTACK-850. Please see 15.7. IP Reservation in Isolated Guest Networks, and provide your feedback. Regards -Radhika
Re: Review Request 13370: CLOUDSTACK-4101: Update the Xenserver template/checksum in template.sql to the one verified.
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13370/#review24791 --- Commit d3666b96cd6aec4a5ac8b968f89d9f60a638572c in branch refs/heads/4.2 from Sanjay Tripathi [ https://git-wip-us.apache.org/repos/asf?p=cloudstack.git;h=d3666b9 ] CLOUDSTACK-4101: Update the Xenserver template/checksum in template.sql to the one verified. - ASF Subversion and Git Services On Aug. 7, 2013, 8:12 a.m., Sanjay Tripathi wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13370/ --- (Updated Aug. 7, 2013, 8:12 a.m.) Review request for cloudstack, Abhinandan Prateek and Devdeep Singh. Bugs: CLOUDSTACK-4101 Repository: cloudstack-git Description --- CLOUDSTACK-4101: Update the Xenserver template/checksum in template.sql to the one verified. Update the Xenserver systemVM template details(URL and checksum) in template.sql file. Diffs - setup/db/templates.sql c243e0f Diff: https://reviews.apache.org/r/13370/diff/ Testing --- Tests: 1. Seeded the new XS systemVM template from the updated location. 2. Deployed the fresh CloudStack environment. 3. Verified that systemVMs are coming up successfully. Thanks, Sanjay Tripathi
Re: Review Request 13324: CLOUDSTACK-4104: VM Creation should fail if a zone is dedicated to a domain and user of the same domain deploys vm without using affinity group Explicit Dedication
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13324/#review24792 --- Commit a3ba198e5aba35297201bdf6b1ef127c0a5b01a5 in branch refs/heads/4.2 from Saksham Srivastava [ https://git-wip-us.apache.org/repos/asf?p=cloudstack.git;h=a3ba198 ] CLOUDSTACK-4104: VM Creation should fail if a zone is dedicated to a domain and user of the same domain deploys vm without using affinity group Explicit Dedication. - ASF Subversion and Git Services On Aug. 6, 2013, 12:53 p.m., Saksham Srivastava wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13324/ --- (Updated Aug. 6, 2013, 12:53 p.m.) Review request for cloudstack and Devdeep Singh. Bugs: 4104 Repository: cloudstack-git Description --- To use dedicated resources one must specify the affinity group explicit dedication. If the affinity groups is not specified dedicated resources should not accessible. Currently there are checks for pod/clusters and hosts but zone could still be used. Added check that throws exception whenever a dedicated zone is being used to deploy vm but no affinity group is being used. Diffs - server/src/com/cloud/deploy/DeploymentPlanningManagerImpl.java ebf2b0c Diff: https://reviews.apache.org/r/13324/diff/ Testing --- Dedicate zone to account/domain and do not use affinity group of type Explicit Dedication - VM deployment fails. Dedicate zone to appropriate account/domain and use affinity group of type Explicit Dedication - VM deployment passes. Build passes. Thanks, Saksham Srivastava
Re: Review Request 13372: CLOUDSTACK-4062: SLES 11 SP2, SLES 11 SP3, CENTOS 6.4, RHEL 6.3, RHEL 6.4 OS type missing from OS Type drop down list when registering template.
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13372/#review24790 --- Commit 5dd155d02b5e03d20faf6826cc86f570b95b6d37 in branch refs/heads/4.2 from Sanjay Tripathi [ https://git-wip-us.apache.org/repos/asf?p=cloudstack.git;h=5dd155d ] CLOUDSTACK-4062: SLES 11 SP2, SLES 11 SP3, CENTOS 6.4, RHEL 6.3, RHEL 6.4 OS type missing from OS Type drop down list when registering template. Also added the support of different versions of SLES, RHEL, Oracle Linux guest OS which was not there in the list. - ASF Subversion and Git Services On Aug. 7, 2013, 9:07 a.m., Sanjay Tripathi wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13372/ --- (Updated Aug. 7, 2013, 9:07 a.m.) Review request for cloudstack, Abhinandan Prateek and Devdeep Singh. Bugs: CLOUDSTACK-4062 Repository: cloudstack-git Description --- CLOUDSTACK-4062: SLES 11 SP2, SLES 11 SP3, CENTOS 6.4, RHEL 6.3, RHEL 6.4 OS type missing from OS Type drop down list when registering template. Also added the support of different versions of SLES, RHEL, Oracle Linux guest OS which was not there in the list. Diffs - setup/db/db/schema-410to420.sql cd9b0cf setup/db/templates.sql c243e0f Diff: https://reviews.apache.org/r/13372/diff/ Testing --- Deployed a fresh CloudStack environment and verified that register template is displaying the added list of guestOS in the list of OS Type. Thanks, Sanjay Tripathi
Re: Review Request 13324: CLOUDSTACK-4104: VM Creation should fail if a zone is dedicated to a domain and user of the same domain deploys vm without using affinity group Explicit Dedication
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13324/#review24795 --- Commit 61c7aa7afea078239146c8957f49e8ea07ebd5db in branch refs/heads/master from Saksham Srivastava [ https://git-wip-us.apache.org/repos/asf?p=cloudstack.git;h=61c7aa7 ] CLOUDSTACK-4104: VM Creation should fail if a zone is dedicated to a domain and user of the same domain deploys vm without using affinity group Explicit Dedication. - ASF Subversion and Git Services On Aug. 6, 2013, 12:53 p.m., Saksham Srivastava wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13324/ --- (Updated Aug. 6, 2013, 12:53 p.m.) Review request for cloudstack and Devdeep Singh. Bugs: 4104 Repository: cloudstack-git Description --- To use dedicated resources one must specify the affinity group explicit dedication. If the affinity groups is not specified dedicated resources should not accessible. Currently there are checks for pod/clusters and hosts but zone could still be used. Added check that throws exception whenever a dedicated zone is being used to deploy vm but no affinity group is being used. Diffs - server/src/com/cloud/deploy/DeploymentPlanningManagerImpl.java ebf2b0c Diff: https://reviews.apache.org/r/13324/diff/ Testing --- Dedicate zone to account/domain and do not use affinity group of type Explicit Dedication - VM deployment fails. Dedicate zone to appropriate account/domain and use affinity group of type Explicit Dedication - VM deployment passes. Build passes. Thanks, Saksham Srivastava
Re: Review Request 13370: CLOUDSTACK-4101: Update the Xenserver template/checksum in template.sql to the one verified.
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13370/#review24794 --- Commit d9568c89d7afea84387f4e265e3b400c33c4dda9 in branch refs/heads/master from Sanjay Tripathi [ https://git-wip-us.apache.org/repos/asf?p=cloudstack.git;h=d9568c8 ] CLOUDSTACK-4101: Update the Xenserver template/checksum in template.sql to the one verified. - ASF Subversion and Git Services On Aug. 7, 2013, 8:12 a.m., Sanjay Tripathi wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13370/ --- (Updated Aug. 7, 2013, 8:12 a.m.) Review request for cloudstack, Abhinandan Prateek and Devdeep Singh. Bugs: CLOUDSTACK-4101 Repository: cloudstack-git Description --- CLOUDSTACK-4101: Update the Xenserver template/checksum in template.sql to the one verified. Update the Xenserver systemVM template details(URL and checksum) in template.sql file. Diffs - setup/db/templates.sql c243e0f Diff: https://reviews.apache.org/r/13370/diff/ Testing --- Tests: 1. Seeded the new XS systemVM template from the updated location. 2. Deployed the fresh CloudStack environment. 3. Verified that systemVMs are coming up successfully. Thanks, Sanjay Tripathi
Re: Review Request 13372: CLOUDSTACK-4062: SLES 11 SP2, SLES 11 SP3, CENTOS 6.4, RHEL 6.3, RHEL 6.4 OS type missing from OS Type drop down list when registering template.
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13372/#review24793 --- Commit 3de6cdb4f9422264babf14728e3ffa939d72c0b1 in branch refs/heads/master from Sanjay Tripathi [ https://git-wip-us.apache.org/repos/asf?p=cloudstack.git;h=3de6cdb ] CLOUDSTACK-4062: SLES 11 SP2, SLES 11 SP3, CENTOS 6.4, RHEL 6.3, RHEL 6.4 OS type missing from OS Type drop down list when registering template. Also added the support of different versions of SLES, RHEL, Oracle Linux guest OS which was not there in the list. - ASF Subversion and Git Services On Aug. 7, 2013, 9:07 a.m., Sanjay Tripathi wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13372/ --- (Updated Aug. 7, 2013, 9:07 a.m.) Review request for cloudstack, Abhinandan Prateek and Devdeep Singh. Bugs: CLOUDSTACK-4062 Repository: cloudstack-git Description --- CLOUDSTACK-4062: SLES 11 SP2, SLES 11 SP3, CENTOS 6.4, RHEL 6.3, RHEL 6.4 OS type missing from OS Type drop down list when registering template. Also added the support of different versions of SLES, RHEL, Oracle Linux guest OS which was not there in the list. Diffs - setup/db/db/schema-410to420.sql cd9b0cf setup/db/templates.sql c243e0f Diff: https://reviews.apache.org/r/13372/diff/ Testing --- Deployed a fresh CloudStack environment and verified that register template is displaying the added list of guestOS in the list of OS Type. Thanks, Sanjay Tripathi
Re: Review Request 13324: CLOUDSTACK-4104: VM Creation should fail if a zone is dedicated to a domain and user of the same domain deploys vm without using affinity group Explicit Dedication
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13324/#review24796 --- Ship it! Ship It! - Devdeep Singh On Aug. 6, 2013, 12:53 p.m., Saksham Srivastava wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13324/ --- (Updated Aug. 6, 2013, 12:53 p.m.) Review request for cloudstack and Devdeep Singh. Bugs: 4104 Repository: cloudstack-git Description --- To use dedicated resources one must specify the affinity group explicit dedication. If the affinity groups is not specified dedicated resources should not accessible. Currently there are checks for pod/clusters and hosts but zone could still be used. Added check that throws exception whenever a dedicated zone is being used to deploy vm but no affinity group is being used. Diffs - server/src/com/cloud/deploy/DeploymentPlanningManagerImpl.java ebf2b0c Diff: https://reviews.apache.org/r/13324/diff/ Testing --- Dedicate zone to account/domain and do not use affinity group of type Explicit Dedication - VM deployment fails. Dedicate zone to appropriate account/domain and use affinity group of type Explicit Dedication - VM deployment passes. Build passes. Thanks, Saksham Srivastava
Re: Review Request 13370: CLOUDSTACK-4101: Update the Xenserver template/checksum in template.sql to the one verified.
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13370/#review24797 --- Ship it! Ship It! - Devdeep Singh On Aug. 7, 2013, 8:12 a.m., Sanjay Tripathi wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13370/ --- (Updated Aug. 7, 2013, 8:12 a.m.) Review request for cloudstack, Abhinandan Prateek and Devdeep Singh. Bugs: CLOUDSTACK-4101 Repository: cloudstack-git Description --- CLOUDSTACK-4101: Update the Xenserver template/checksum in template.sql to the one verified. Update the Xenserver systemVM template details(URL and checksum) in template.sql file. Diffs - setup/db/templates.sql c243e0f Diff: https://reviews.apache.org/r/13370/diff/ Testing --- Tests: 1. Seeded the new XS systemVM template from the updated location. 2. Deployed the fresh CloudStack environment. 3. Verified that systemVMs are coming up successfully. Thanks, Sanjay Tripathi
Re: Review Request 13372: CLOUDSTACK-4062: SLES 11 SP2, SLES 11 SP3, CENTOS 6.4, RHEL 6.3, RHEL 6.4 OS type missing from OS Type drop down list when registering template.
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13372/#review24798 --- Ship it! Ship It! - Devdeep Singh On Aug. 7, 2013, 9:07 a.m., Sanjay Tripathi wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13372/ --- (Updated Aug. 7, 2013, 9:07 a.m.) Review request for cloudstack, Abhinandan Prateek and Devdeep Singh. Bugs: CLOUDSTACK-4062 Repository: cloudstack-git Description --- CLOUDSTACK-4062: SLES 11 SP2, SLES 11 SP3, CENTOS 6.4, RHEL 6.3, RHEL 6.4 OS type missing from OS Type drop down list when registering template. Also added the support of different versions of SLES, RHEL, Oracle Linux guest OS which was not there in the list. Diffs - setup/db/db/schema-410to420.sql cd9b0cf setup/db/templates.sql c243e0f Diff: https://reviews.apache.org/r/13372/diff/ Testing --- Deployed a fresh CloudStack environment and verified that register template is displaying the added list of guestOS in the list of OS Type. Thanks, Sanjay Tripathi
Review Request 13375: CLOUDSTACK-4144 Make specifyVlan to false for shared network
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13375/ --- Review request for cloudstack and Prasanna Santhanam. Bugs: CLOUDSTACK-4144 Repository: cloudstack-git Description --- CLOUDSTACK-4144 Make specifyVlan to false for shared network Diffs - test/integration/component/test_vpc_vm_life_cycle.py 9b10133 Diff: https://reviews.apache.org/r/13375/diff/ Testing --- Thanks, Girish Shilamkar
RE: Unable to create image on Cloudcentral
Thanks, but it resolved. -Original Message- From: Daan Hoogland [mailto:daan.hoogl...@gmail.com] Sent: Wednesday, August 07, 2013 3:43 PM To: dev; Prabakaran R Subject: Re: Unable to create image on Cloudcentral Prabakaran, Can you make 'I got null pointer exception' more explicit? regards, On Thu, Jul 25, 2013 at 12:56 PM, Prabakaran R praba...@cusdelight.com wrote: Hi, I am trying to create image of virtual machine in cloudcentral using JClouds. Following is my code snippet CreateTemplateOptions templateOption = CreateTemplateOptions.NONE; templateOption.bits(64); templateOption.passwordEnabled(true); TemplateMetadata meta = TemplateMetadata.builder().name(Test Debian 6.0.6).osTypeId(26226f97-25c0-49e4-be29-03f13308ade8). displayText(Test Debian 6.0.6 64-bit).virtualMachineId(---).build(); template.createTemplate(meta,templateOption); But I got null pointer exception. VM is in stop state, kindly help me to this. Thanks, Prabakaran.R
Re: Review Request 13325: preserve data in iso_id column while upgrading form 3.0.7 to 4.2
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13325/ --- (Updated Aug. 7, 2013, 1:02 p.m.) Review request for cloudstack and Kishan Kavala. Bugs: Cloudstack-4100 Repository: cloudstack-git Description --- preserve data in iso_id column while upgrading form 3.0.7 to 4.2 Earlier was dropping the iso_id column as a result we loose data from the iso_id column prior to upgrade. Now fixed this by. renaming the iso_id to iso_id1 in schema-307to401 file. and then after upgrade dropping the iso_id column and renaming the iso_id1 to iso_id in 307to401 cleanup schema. Diffs (updated) - engine/schema/src/com/cloud/upgrade/dao/Upgrade410to420.java 2e5fbcb setup/db/db/schema-307to410.sql af296ab Diff: https://reviews.apache.org/r/13325/diff/ Testing --- tested the upgrade from 3.0.7 to 4.2. Thanks, bharat kumar
Re: Review Request 13325: preserve data in iso_id column while upgrading form 3.0.7 to 4.2
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13325/ --- (Updated Aug. 7, 2013, 1:02 p.m.) Review request for cloudstack and Kishan Kavala. Bugs: Cloudstack-4100 Repository: cloudstack-git Description --- preserve data in iso_id column while upgrading form 3.0.7 to 4.2 Earlier was dropping the iso_id column as a result we loose data from the iso_id column prior to upgrade. Now fixed this by. renaming the iso_id to iso_id1 in schema-307to401 file. and then after upgrade dropping the iso_id column and renaming the iso_id1 to iso_id in 307to401 cleanup schema. Diffs (updated) - engine/schema/src/com/cloud/upgrade/dao/Upgrade410to420.java 2e5fbcb setup/db/db/schema-307to410.sql af296ab Diff: https://reviews.apache.org/r/13325/diff/ Testing --- tested the upgrade from 3.0.7 to 4.2. Thanks, bharat kumar
Re: Review Request 13325: preserve data in iso_id column while upgrading form 3.0.7 to 4.2
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13325/#review24799 --- Ship it! commit 95f36798559c71d1de9f34fedad597d12f09766b - Kishan Kavala On Aug. 7, 2013, 6:32 p.m., bharat kumar wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13325/ --- (Updated Aug. 7, 2013, 6:32 p.m.) Review request for cloudstack and Kishan Kavala. Bugs: Cloudstack-4100 Repository: cloudstack-git Description --- preserve data in iso_id column while upgrading form 3.0.7 to 4.2 Earlier was dropping the iso_id column as a result we loose data from the iso_id column prior to upgrade. Now fixed this by. renaming the iso_id to iso_id1 in schema-307to401 file. and then after upgrade dropping the iso_id column and renaming the iso_id1 to iso_id in 307to401 cleanup schema. Diffs - engine/schema/src/com/cloud/upgrade/dao/Upgrade410to420.java 2e5fbcb setup/db/db/schema-307to410.sql af296ab Diff: https://reviews.apache.org/r/13325/diff/ Testing --- tested the upgrade from 3.0.7 to 4.2. Thanks, bharat kumar
Re: Review Request 13373: keep the existing changes in cloud-set-guest-sshkey.in to make sure we do not break the backward compatibility.
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13373/ --- (Updated Aug. 7, 2013, 1:25 p.m.) Review request for cloudstack, Jayapal Reddy and Sheng Yang. Bugs: Cloudstack-4138 Repository: cloudstack-git Description --- keep the existing changes in cloud-set-guest-sshkey.in to make sure we do not break the backward compatibility. Diffs - setup/bindir/cloud-set-guest-sshkey.in 9436b03 Diff: https://reviews.apache.org/r/13373/diff/ Testing --- Tested on 4.2 Thanks, bharat kumar
Review Request 13376: Fix QuickCloud secondary storage service ListTemplateCommand and ListVolumeCommand
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13376/ --- Review request for cloudstack, Chiradeep Vittal and Devdeep Singh. Repository: cloudstack-git Description --- Fix non-systemVM execution. ListTemplateCommand and ListVolumeCommand need to return the corresponding Answer object. No bug was logged. Diffs - services/secondary-storage/src/org/apache/cloudstack/storage/resource/NfsSecondaryStorageResource.java 89ce8e009a36df53f10efda8ef7c3d8f7ed6e2dc Diff: https://reviews.apache.org/r/13376/diff/ Testing --- Integration test: launched mgmt srvr with Quick Cloud configuration, launched QuickCloud secondary storage server, verified that no exceptions no longer occurred when secondary storage server connected to mgmt server. Thanks, Donal Lafferty
Re: Ruby dev wanted
Hi Chip, I just sent the pull request for the cloudstack_ruby_client on Github. Please have a look! I did some restructure works, let me know if you have any concerns, thanks. Regards Isaac On Thu, Aug 1, 2013 at 9:38 PM, Chip Childers chip.child...@sungard.comwrote: Happy to accept pull requests for cloudstack_ruby_client and to them push the gem. FWIW, you can always get the latest gem via: gem install cloudstack_ruby_client On Thu, Aug 01, 2013 at 06:03:35PM +0800, Isaac Chiang wrote: Hi Sebastien, I've read the ruby client project by Chips. It's pretty looks like I'm doing now for _https://github.com/isaacchiang/cloud_ruby. Please have a look. I believe there are some code segments that would be helpful to the rOCCI driver. I need more advice to the rOCCI project, thanks. Regard Isaac On Fri, Jul 26, 2013 at 11:16 PM, Sebastien Goasguen run...@gmail.com wrote: On Jul 25, 2013, at 8:04 PM, Isaac Chiang isaacchi...@gmail.com wrote: Hi Sebastien, I'm interested in the rOCCi project. I'd like to join. However, I know nothing about the OCCI and will spend some time on it in next few days. Please kindly let me know more information about the project(status, repo location, e.t.c., thanks :) Hi Isaac, Great, that's terrific of you to volunteer, should be fun. The first thing I think would be to check out Chip'c ruby client: https://github.com/chipchilders/cloudstack_ruby_client Test it and finish it up (I am sure Chip can chime in). Then, you would need to look at: https://github.com/gwdg/rOCCI Figure out how to write a driver, there is an example for opennebula and ec2 I believe. And then go for it :) Cheers, -Sebastien Regards Isaac Hi folks, I committed to do an interoperability workshop in Madrid, Spain in september: http://www.cloudplugfest.org/cloud-interoperability-week/detailed-program All four sisters of cloud are represented. The main goal is to show that we can talk to CloudStack with OCCI and even CIMI standards. The problem is that currently we cannot :) So I am looking for some help from Ruby developers who would be willing to work on: https://github.com/gwdg/rOCCI and also on delta cloud (even though the project may have stalled): http://deltacloud.apache.org A good start might be Chip's ruby client: https://github.com/chipchilders/cloudstack_ruby_client Help, any takers ? -Sebastien
Re: Review Request 13374: Automation: Affinity Groups: Removing invalid test case
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13374/#review24800 --- Ship it! - Prasanna Santhanam On Aug. 7, 2013, 11:18 a.m., Gaurav Aradhye wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13374/ --- (Updated Aug. 7, 2013, 11:18 a.m.) Review request for cloudstack and Prasanna Santhanam. Repository: cloudstack-git Description --- Removing invalid test case. Affinity group can be deleted when vms are still using it. This can be reproduced from UI. Diffs - test/integration/component/test_affinity_groups.py 309adc6 Diff: https://reviews.apache.org/r/13374/diff/ Testing --- Thanks, Gaurav Aradhye
Re: Ruby dev wanted
On Wed, Aug 07, 2013 at 10:05:13PM +0800, Isaac Chiang wrote: Hi Chip, I just sent the pull request for the cloudstack_ruby_client on Github. Please have a look! I did some restructure works, let me know if you have any concerns, thanks. Regards Isaac Looks great Isaac. I merged the changes and pushed a 0.1.0 gem.
RE: [Responsiveness report] 2013w31 dev
Daan, thanks a million for doing this! -Original Message- From: Daan Hoogland [mailto:dhoogl...@schubergphilis.com] Sent: Wednesday, August 07, 2013 6:24 AM To: dev@cloudstack.apache.org Subject: [Responsiveness report] 2013w31 dev the three from week 30 where still open. I asked for clarification on some of them. one is from Ian who has been present enough on this list to take care of himself. This week there is only one (disgarding Mins direct addressing of John :) I am not bothering you with this one as I responded with a witty question a minute ago. net result: no mails from week 31 where without response, Daan Hoogland
Re: IP Address Allocation
On 07/08/13 1:40 AM, Alex Huang alex.hu...@citrix.com wrote: I'll say it's a bad design. He he I designed it. :( We had to make some compromises due to time constraints. I always intended to move the ip allocation into NetworkGurus and actually attempted to several times but, at the end for various reasons, we just couldn't complete that task. I generally believe that ip address allocation can actually be done even without CloudStack's database, for example, if it came from a master dhcp server or some IT central database. So an interface like IpAddressReserver to abstract that functionality would be great. I have to say the number of implementations for NetworkGuru grew faster than I expected. In one sense, it's a good sign that everyone was able to use it but then it exposes the work that was incomplete to begin with. IIRC, there's basically two problems I left on NetworkGuru to be better designed. One is this ip address allocation. The other problem is how to identify which set of NetworkGurus should manage the network. I believe that selection process should be much more transparent to the administrator who setup the CS deployment but because at the time only VLAN isolation was supported, I did not put that into CS. As more and more SDN technologies come online, it's best we design that process. Can we follow the path of NetworkElement which was once a single base class implementing all service abstractions to set of derived interfaces for each service (DhcpServiceProvider, LoadBalancingServiceProvider etc). With this approach a network element can just implement the service it want to provide. Current guru implementation are tightly coupled with each other. For e.g Midonet, Nicira guru implement the network design for their isolation and for the rest of the functionality they fall back on Guest network guru. Perhaps we can split the abstraction of network guru into multiple derived interface specific to each function like isolation, network design, ip address management. So if Infoblox guru just want to provide IPAM, then it can just focus on it with out the knowledge of other functions of network guru. I think Hugo made some changes there but I haven't had a chance to study how that all works now in 4.x yet. There are some artifacts in the code to show I was thinking about this. For example, Network and Nic carries the Isolation URI instead of just VLAN tag because it was intended to be cast by the NetworkGuru and then transported to the hypervisor. We expected anyone who plugs into CloudStack have two places to plug into: orchestration and actual provisioning. So the NetworkGuru forms the URI for isolation and then it has code at the hypervisor side to resolve the URI to something that the hypervisor can understand. --Alex -Original Message- From: Soheil Eizadi [mailto:seiz...@infoblox.com] Sent: Tuesday, August 6, 2013 11:54 AM To: dev@cloudstack.apache.org Subject: RE: IP Address Allocation Agree that the IP Address allocation is different than the DHCP and DNS service offered by the NetworkElements. The third party NetworkGurus just reference the standard NetworkManager IP Address Allocation routines, either directly (e.g. _networkMgr.assignPublicIpAddress()) or by calling out to the GuestNetworkGuru. I would say in the current design the Network Manager manages the IP Addresses not the Network Gurus. I don't think this is a bad design, in that the NetworkGurus are sharing the same ip address allocation and get new features (e.g. IP reservation or VM migration) without implementing new code to support them. -Soheil From: Alex Huang [alex.hu...@citrix.com] Sent: Tuesday, August 06, 2013 11:27 AM To: Chiradeep Vittal; dev@cloudstack.apache.org Subject: RE: IP Address Allocation We can either break NetworkManager down further and say NetworkGuru sets the actual parameters of the network and a new IpAddressReserver sets the ip address or we can let NetworkGurus identify the IpAddressReserver to use themselves. I generally think NetworkManager shouldn't care where the ip address came from because it really shouldn't know anything about the actual network we're laying out for the end user. That's always been left to the NetworkGurus. If you think about it from that perspective, then it's natural that ip address is allocated by the NetworkGuru. But I can see different NetworkGurus may want to share the same ip address allocation tables/schemes so an interface for them to share that code does make sense. Note there's a difference between allocating ip address and the act of actually issuing the ip address and associating the ip address to a fqdn. That act is performed by the network elements. Here we're specifically saying who is responsible for figuring out the ip address that can be allocated to a particular vm and works within a network. --Alex -Original
Review Request 13378: Test assumes storagetype to be local therefore test_egress_fw_rules fail.
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13378/ --- Review request for cloudstack, Ashutosh Kelkar and Prasanna Santhanam. Repository: cloudstack-git Description --- Test assumes storagetype to be local therefore test_egress_fw_rules fail. Removed storagetype definition. Diffs - test/integration/component/test_egress_fw_rules.py 6f511e5 Diff: https://reviews.apache.org/r/13378/diff/ Testing --- Thanks, Girish Shilamkar
Re: Comment about Data Disk Offering when creating a VM from an ISO
That sounds reasonable to me. Do you know if a bug exists for this or perhaps I should create one? On Wed, Aug 7, 2013 at 2:03 AM, Ryan Lei ryan...@cht.com.tw wrote: I noticed this issue long ago, and I agree with you. In my understanding, while the tab in the VM wizard always says Data Disk Offering, if you create a VM using ISO, this step actually means Root Disk Offering. If this is the case, the tab name should depend on whether the user chooses Select Template (Data Disk Offering) or Select ISO (Root Disk Offering). --- Yu-Heng (Ryan) Lei, Associate Reasearcher Chunghwa Telecom Laboratories / Cloud Computing Laboratory ryan...@cht.com.tw https://email.cht.com.tw/owa/redir.aspx?C=-wE1FEC3G0SWYpVkiWo8SsDdf3ZqO9AIuAPTzRnFYCUi-z4YljtI_hyVKkNHfn9F1Bn-vUWJnQ4.URL=mailto%3aryanlei%40cht.com.tw or ryanlei750...@gmail.com On Wed, Aug 7, 2013 at 9:12 AM, Mike Tutkowski mike.tutkow...@solidfire.com wrote: Hi, I noticed while creating a VM from an ISO today that I am taken to a page in the wizard that talks about a Data Disk Offering. It seems a bit confusing to me because I don't actually get a data disk once the wizard has spun up the VM. This Data Disk Offering page is displayed to collect data regarding the size of the root disk the VM will have. Any thoughts on this? Is this a known issue or should I log a bug? Thanks! -- *Mike Tutkowski* *Senior CloudStack Developer, SolidFire Inc.* e: mike.tutkow...@solidfire.com o: 303.746.7302 Advancing the way the world uses the cloudhttp://solidfire.com/solution/overview/?video=play *™* -- *Mike Tutkowski* *Senior CloudStack Developer, SolidFire Inc.* e: mike.tutkow...@solidfire.com o: 303.746.7302 Advancing the way the world uses the cloudhttp://solidfire.com/solution/overview/?video=play *™*
Re: [Responsiveness report] 2013w31 dev
less and less a problem :P On Wed, Aug 7, 2013 at 4:34 PM, Musayev, Ilya imusa...@webmd.net wrote: Daan, thanks a million for doing this! -Original Message- From: Daan Hoogland [mailto:dhoogl...@schubergphilis.com] Sent: Wednesday, August 07, 2013 6:24 AM To: dev@cloudstack.apache.org Subject: [Responsiveness report] 2013w31 dev the three from week 30 where still open. I asked for clarification on some of them. one is from Ian who has been present enough on this list to take care of himself. This week there is only one (disgarding Mins direct addressing of John :) I am not bothering you with this one as I responded with a witty question a minute ago. net result: no mails from week 31 where without response, Daan Hoogland
Re: Review Request 13377: CLOUDSTACK-4021: [Automation] TestExplicitDedication.test_01_deploy_vm_with_explicit_dedication failed to deploy VM
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13377/ --- (Updated Aug. 7, 2013, 4:18 p.m.) Review request for cloudstack, Devdeep Singh and Prasanna Santhanam. Changes --- updated diff. Bugs: CLOUDSTACK-4021 Repository: cloudstack-git Description --- Test case integration.component.test_explicit_dedication.TestExplicitDedication.test_01_deploy_vm_with_explicit_dedication fails if it does not find an empty hosts. In case there is at least 1 empty host present, that host is dedicated and a vm is deployed using explicit dedication affinity group. Added check that skips the test in case there are no empty hosts found. Diffs (updated) - test/integration/component/test_explicit_dedication.py e1991ac Diff: https://reviews.apache.org/r/13377/diff/ Testing --- If there are no empty hosts present, the test returns without further dedication or deploying vms. If there is a single empty host present, the host is dedicated and vm is deployed. Thanks, Saksham Srivastava
Re: CloudStack Storage Plugin for CloudByte
Feel free to take a look at the SolidFire plug-in I developed for 4.2. I built it to support QoS (Min and Max IOPS guarantees), which may be of interest to you. In 4.2, it works for Disk Offerings only (hopefully Compute Offerings, as well, in 4.3). Also, since the hypervisor plug-ins needed to be changed for my purposes (to support a one-to-one mapping between a volume on the SAN and a CloudStack volume), it only supports XenServer and VMware in 4.2. I expect to extend that support to KVM in 4.3. You should probably work with me and John Burwell (CCed here) as you develop your plug-in since there may be additional changes to the storage framework for 4.3 that could impact your work. On Wed, Aug 7, 2013 at 1:50 AM, Amit Das amit@cloudbyte.com wrote: Thanks for the information. Has 4.3 release planning been done ? Regards, Amit *CloudByte Inc.* http://www.cloudbyte.com/ On Wed, Aug 7, 2013 at 1:06 PM, Prasanna Santhanam t...@apache.org wrote: On Wed, Aug 07, 2013 at 12:50:43PM +0530, Amit Das wrote: Hi, I will be working on implementing the CS storage plugin for CloudByte's Elastistor. This new plugin should be based on the newly refactored storage plugin changes which has gone into 4.2 release already. Which CS release can accommodate this new storage plugin feature ? You can discuss the design and work towards the next release. The feature will go into master and based on what the next release version would be (4.3 I think) it will automatically go into the appropriate branch provided it made it to master in time. 4.2 is frozen for now. -- Prasanna., Powered by BigRock.com -- *Mike Tutkowski* *Senior CloudStack Developer, SolidFire Inc.* e: mike.tutkow...@solidfire.com o: 303.746.7302 Advancing the way the world uses the cloudhttp://solidfire.com/solution/overview/?video=play *™*
Re: CloudStack Storage Plugin for CloudByte
We have time based releases, so as long as your feature is code complete (not necessarily bug-free) by 9/30, it will hit 4.3. But first you have to formally propose the feature. Please see https://cwiki.apache.org/confluence/x/0h-VAQ https://cwiki.apache.org/confluence/x/ni-VAQ https://cwiki.apache.org/confluence/x/XUfVAQ On 8/7/13 12:50 AM, Amit Das amit@cloudbyte.com wrote: Thanks for the information. Has 4.3 release planning been done ? Regards, Amit *CloudByte Inc.* http://www.cloudbyte.com/ On Wed, Aug 7, 2013 at 1:06 PM, Prasanna Santhanam t...@apache.org wrote: On Wed, Aug 07, 2013 at 12:50:43PM +0530, Amit Das wrote: Hi, I will be working on implementing the CS storage plugin for CloudByte's Elastistor. This new plugin should be based on the newly refactored storage plugin changes which has gone into 4.2 release already. Which CS release can accommodate this new storage plugin feature ? You can discuss the design and work towards the next release. The feature will go into master and based on what the next release version would be (4.3 I think) it will automatically go into the appropriate branch provided it made it to master in time. 4.2 is frozen for now. -- Prasanna., Powered by BigRock.com
[ACS42]Attach volumes to Instance which has VM snapshots
Hi, I have an instance which has VM snapshots. (VMWARE hypervisor) Now I am trying to attach DATA volume to this instance. It failed saying Unable to attach volume, please specify a VM that does not have VM snapshots Is there any limitation not to add additional volumes if it has VM snapshots? Thanks, Sailaja.M
Re: Review Request 12849: added backwards compatibility code to Networks enums
On Aug. 6, 2013, 11:40 p.m., Sheng Yang wrote: api/src/com/cloud/network/Networks.java, line 79 https://reviews.apache.org/r/12849/diff/5/?file=337785#file337785line79 I don't quite understand this part. Why // is missing in this case? Who would require // and who wouldn't? daan Hoogland wrote: Only the mido code needs the explicit // the rest can deal with URI constructors Are you sure? I'm seeing problems with CitrixResourceBase.getNetwork(CitrixResourceBase.java:1043): long vlan = Long.parseLong(broadcastUri.getHost()); - Donal --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/12849/#review24756 --- On Aug. 7, 2013, 9:40 a.m., daan Hoogland wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/12849/ --- (Updated Aug. 7, 2013, 9:40 a.m.) Review request for cloudstack, Chiradeep Vittal, Dave Cahill, Koushik Das, and Sheng Yang. Repository: cloudstack-git Description --- Both BroadcastDomainType and IsolationType needed some extra code for backwards compatibility Diffs - api/src/com/cloud/network/Networks.java c76c3d4 api/test/com/cloud/network/NetworksTest.java 31114e8 Diff: https://reviews.apache.org/r/12849/diff/ Testing --- Thanks, daan Hoogland
Re: CloudStack Storage Plugin for CloudByte
Thanks Mike. I have started looking into the storage related changes due to SolidFire storage plugin. I shall start the formal proposal process communicate my git feature branch soon. Regards, Amit *CloudByte Inc.* http://www.cloudbyte.com/ On Wed, Aug 7, 2013 at 9:50 PM, Mike Tutkowski mike.tutkow...@solidfire.com wrote: Feel free to take a look at the SolidFire plug-in I developed for 4.2. I built it to support QoS (Min and Max IOPS guarantees), which may be of interest to you. In 4.2, it works for Disk Offerings only (hopefully Compute Offerings, as well, in 4.3). Also, since the hypervisor plug-ins needed to be changed for my purposes (to support a one-to-one mapping between a volume on the SAN and a CloudStack volume), it only supports XenServer and VMware in 4.2. I expect to extend that support to KVM in 4.3. You should probably work with me and John Burwell (CCed here) as you develop your plug-in since there may be additional changes to the storage framework for 4.3 that could impact your work. On Wed, Aug 7, 2013 at 1:50 AM, Amit Das amit@cloudbyte.com wrote: Thanks for the information. Has 4.3 release planning been done ? Regards, Amit *CloudByte Inc.* http://www.cloudbyte.com/ On Wed, Aug 7, 2013 at 1:06 PM, Prasanna Santhanam t...@apache.org wrote: On Wed, Aug 07, 2013 at 12:50:43PM +0530, Amit Das wrote: Hi, I will be working on implementing the CS storage plugin for CloudByte's Elastistor. This new plugin should be based on the newly refactored storage plugin changes which has gone into 4.2 release already. Which CS release can accommodate this new storage plugin feature ? You can discuss the design and work towards the next release. The feature will go into master and based on what the next release version would be (4.3 I think) it will automatically go into the appropriate branch provided it made it to master in time. 4.2 is frozen for now. -- Prasanna., Powered by BigRock.com -- *Mike Tutkowski* *Senior CloudStack Developer, SolidFire Inc.* e: mike.tutkow...@solidfire.com o: 303.746.7302 Advancing the way the world uses the cloudhttp://solidfire.com/solution/overview/?video=play *™*
Re: Review Request 13008: Fix usage of vm.instancename.flag when generating VM names on ESX hypervisor
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13008/#review24812 --- Ship it! commit 059e4b2ae0a44205a8567a79bf9a154a54a83492 Author: Kelven Yang kelv...@gmail.com Date: Wed Aug 7 10:28:47 2013 -0700 CLOUDSTACK-3886: check in for review request https://reviews.apache.org/r/13008/ - Kelven Yang On Aug. 6, 2013, 9:10 p.m., Venkata Siva Vijayendra Bhamidipati wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13008/ --- (Updated Aug. 6, 2013, 9:10 p.m.) Review request for cloudstack, Alex Huang, Chip Childers, Kelven Yang, Sateesh Chodapuneedi, and William Chan. Bugs: CLOUDSTACK-3886 Repository: cloudstack-git Description --- The vminstancename flag has been incorrectly used to simply append the displayname to the internal VM name that shows up on vCenter in vmware deployments. It was intended to show the actual name supplied as hostname, on the hypervisor. This helps admins and deployers to quickly identify VMs and resolve issues related to those VMs. Its usage is very limited as it stands now. This fix corrects it to ensure that the name of the VM on the hypervisor matches the hostname if it is supplied, and if the vm.instancename.flag is set to true. Diffs - engine/orchestration/src/org/apache/cloudstack/platform/orchestration/CloudOrchestrator.java 96fb1d9 plugins/hypervisors/vmware/src/com/cloud/hypervisor/guru/VMwareGuru.java 292f7e9 plugins/hypervisors/vmware/src/com/cloud/hypervisor/vmware/manager/VmwareManagerImpl.java 8d6bdb8 plugins/hypervisors/vmware/src/com/cloud/hypervisor/vmware/manager/VmwareStorageManagerImpl.java 6b721de plugins/hypervisors/vmware/src/com/cloud/hypervisor/vmware/resource/VmwareResource.java d395425 plugins/hypervisors/vmware/src/com/cloud/storage/resource/VmwareStorageProcessor.java fe127e3 server/src/com/cloud/configuration/Config.java ad8e3aa server/src/com/cloud/ha/HighAvailabilityManagerImpl.java 25c5a04 server/src/com/cloud/hypervisor/HypervisorGuruBase.java e042eb8 server/src/com/cloud/vm/UserVmManagerImpl.java 4a222c4 server/src/com/cloud/vm/VirtualMachineManagerImpl.java b33ee49 setup/db/db/schema-410to420.sql b3206af vmware-base/src/com/cloud/hypervisor/vmware/mo/ClusterMO.java 04ef0f8 vmware-base/src/com/cloud/hypervisor/vmware/mo/CustomFieldConstants.java 11bc157 vmware-base/src/com/cloud/hypervisor/vmware/mo/HostMO.java 2735fb0 vmware-base/src/com/cloud/hypervisor/vmware/mo/HypervisorHostHelper.java d20eb6f vmware-base/src/com/cloud/hypervisor/vmware/mo/VirtualMachineMO.java cb97b26 vmware-base/src/com/cloud/hypervisor/vmware/mo/VmwareHypervisorHost.java ac14328 Diff: https://reviews.apache.org/r/13008/diff/ Testing --- Post this change, all major VM operations, namely creation/destruction/expunging/start/stop/reboot of the VM have been tested and observed to work correctly. Part of this patch also puts in a fix for VMSync operations where the CS mgmt server doesn't detect that a guest VM is down, if the guest VM has been shut down/powered off in vCenter. Other operations such as VM snapshot, volume snapshots of disks belonging to the VM, volume migration across primaries, volume attach/detach have also been tested and they are working as expected. This is a functional change, and completely transparent to any of cloudstack's existing functionalities and all the test cases that cover the above code paths and APIs - all existing tests should and do pass - no new tests are necessary. UPDATE for 08/05/2013 = 4.2 code has diverged - need to refactor the patch on top of latest 4.2. Will test out refactored patch and resubmit. Thanks, Venkata Siva Vijayendra Bhamidipati
Re: Review Request 12849: added backwards compatibility code to Networks enums
Donal, The mix of vlan://number and number is one of the issues i had to deal with doing this. The vlan not being handled by my first patch correctly is why this second one was made. for vlans it reverts what I did before. What is it you are seeing? And how does it relate to the BroadcastDomainType? The code you refer to expects something like 'scheme://number'. The reason mido needs the string concatination is because it can contain 'mido://somestring:someotherstring'. Daan On Wed, Aug 7, 2013 at 7:12 PM, Donal Lafferty donal.laffe...@citrix.comwrote: This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/12849/ On August 6th, 2013, 11:40 p.m. UTC, *Sheng Yang* wrote: api/src/com/cloud/network/Networks.javahttps://reviews.apache.org/r/12849/diff/5/?file=337785#file337785line79 (Diff revision 5) public class Networks { 79 return new URI(vlan, value.toString(), null, null); I don't quite understand this part. Why // is missing in this case? Who would require // and who wouldn't? On August 7th, 2013, 8:42 a.m. UTC, *daan Hoogland* wrote: Only the mido code needs the explicit // the rest can deal with URI constructors Are you sure? I'm seeing problems with CitrixResourceBase.getNetwork(CitrixResourceBase.java:1043): long vlan = Long.parseLong(broadcastUri.getHost()); - Donal On August 7th, 2013, 9:40 a.m. UTC, daan Hoogland wrote: Review request for cloudstack, Chiradeep Vittal, Dave Cahill, Koushik Das, and Sheng Yang. By daan Hoogland. *Updated Aug. 7, 2013, 9:40 a.m.* *Repository: * cloudstack-git Description Both BroadcastDomainType and IsolationType needed some extra code for backwards compatibility Diffs - api/src/com/cloud/network/Networks.java (c76c3d4) - api/test/com/cloud/network/NetworksTest.java (31114e8) View Diff https://reviews.apache.org/r/12849/diff/
Re: Review Request 13298: Put in upgrade path from 4.1.1 to 4.1.2
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13298/#review24813 --- Vijay, as the schema is empty, you don't have to have this file. Remove it pls, as well as the reference to it from the corresponding Java code. - Alena Prokharchyk On Aug. 5, 2013, 9:12 p.m., Venkata Siva Vijayendra Bhamidipati wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13298/ --- (Updated Aug. 5, 2013, 9:12 p.m.) Review request for cloudstack, Alena Prokharchyk, Chip Childers, and Min Chen. Bugs: CLOUDSTACK-4091 Repository: cloudstack-git Description --- Putting in upgrade path for db upgrade from 4.1.1 to 4.1.2. Prior to the patch, th mgmt server doesn't start up successfully on 4.1. Diffs - server/src/com/cloud/upgrade/DatabaseUpgradeChecker.java a7a4e61 server/src/com/cloud/upgrade/dao/Upgrade411to412.java PRE-CREATION setup/db/db/schema-411to412.sql PRE-CREATION Diff: https://reviews.apache.org/r/13298/diff/ Testing --- Mgmt server comes up successfully upon fresh install of 4.1, and the db version shows the right value: mysql select * from version; ++-+-+--+ | id | version | updated | step | ++-+-+--+ | 1 | 4.0.0 | 2013-08-05 07:40:38 | Complete | | 2 | 4.1.0 | 2013-08-05 14:42:20 | Complete | | 3 | 4.1.1 | 2013-08-05 14:42:20 | Complete | | 4 | 4.1.2 | 2013-08-05 14:42:20 | Complete | ++-+-+--+ 4 rows in set (0.00 sec) mysql Thanks, Venkata Siva Vijayendra Bhamidipati
[KVM] Helper for agent during HA operations
Hi, In our production setups we have seen some crashes of the KVM agent. This could happen for all kinds of reasons, but that's not what I wanted to discuss. Also see this issue: https://issues.apache.org/jira/browse/CLOUDSTACK-3954 What I've been writing for a PoC in our company is a small helper written in Python which runs on port 8251. The Investigator can query this webservice (attached) which will simply tell it which VMs are running on that host. It's online here: http://stack01.ceph.widodh.nl:8251/ You can also do a query like this: http://stack01.ceph.widodh.nl:8251/ping/i-2-6570-VM This way we can more reliably verify if a specific VM is still running if the Agent stops responding for some reason. A ICMP echo-request isn't safe since the Security Groups could prevent ICMP from coming through. I'd rather not have the management server query libvirt directly, since that would open a potential security whole. This webservice is read-only and on my production setups I have libvirt listening on the private bridge only. What do you think? Wido #!/usr/bin/python ''' This is a helper for the CloudStack Agent for High Availability checks It will run on port 8251 and it can tell the Management server whichs instances are running and if a particular instance is still running here. This is for cases where the main Agent crashes or becomes unrepsonsive and the HA Investigators start doing their work It provides an alternative way to see which Instances are still running on this host ''' from BaseHTTPServer import BaseHTTPRequestHandler,HTTPServer import json import libvirt tcp_port = 8251 class RequestHandler(BaseHTTPRequestHandler): def do_GET(self): content_type = 'application/json' conn = libvirt.openReadOnly('qemu:///system') if conn == None: self.send_response(503) self.end_headers() return virtdomains = map(conn.lookupByID, conn.listDomainsID()) domains = [] for domain in virtdomains: domains.append(domain.name()) if self.path == '/': self.send_response(200) self.send_header('Content-type', content_type) self.end_headers() self.wfile.write('{ instances: ' + json.dumps(domains) + '}') elif self.path.startswith('/ping/'): if self.path.count('/') == 2: alive, domain = self.path.lstrip('/').split('/') self.send_response(200) self.send_header('Content-type', content_type) self.end_headers() running = False for dom in domains: if dom == domain: running = True if running == True: result = true else: result = false self.wfile.write('{ alive: ' + result + ' }') else: self.send_response(405) else: self.send_response(405) conn.close() try: server = HTTPServer(('', tcp_port), RequestHandler) print 'Started httpserver on port ' , tcp_port server.serve_forever() except (KeyboardInterrupt, SystemExit): print '^C received, shutting down the web server' server.socket.close()
RE: IP Address Allocation
I think this is a consensus of where we are: We would follow the path of NetworkElement meaning the infrastructure to create a service offering, the provider concept and the use of agent/host infrastructure to support external providers. (The external NetworkGuru also uses many of the same infrastructure.) This would allow developing a plugin, if needed that could provide in addition to NetworkElement and NetworkGuru features the new External IP Address Allocation capability. The points Alex made was to complete the work in the NetworkManager and refactor out the IP Address Allocation out into a unique class that can be called by NetworkGurus directly rather than using the NetworkManager. This new Class would have logic to call to the external IP Address Providers or use internal database to provide allocation. I hope I captured the essence of where we are with this? -Soheil From: Murali Reddy [murali.re...@citrix.com] Sent: Wednesday, August 07, 2013 7:43 AM To: dev@cloudstack.apache.org Subject: Re: IP Address Allocation On 07/08/13 1:40 AM, Alex Huang alex.hu...@citrix.com wrote: I'll say it's a bad design. He he I designed it. :( We had to make some compromises due to time constraints. I always intended to move the ip allocation into NetworkGurus and actually attempted to several times but, at the end for various reasons, we just couldn't complete that task. I generally believe that ip address allocation can actually be done even without CloudStack's database, for example, if it came from a master dhcp server or some IT central database. So an interface like IpAddressReserver to abstract that functionality would be great. I have to say the number of implementations for NetworkGuru grew faster than I expected. In one sense, it's a good sign that everyone was able to use it but then it exposes the work that was incomplete to begin with. IIRC, there's basically two problems I left on NetworkGuru to be better designed. One is this ip address allocation. The other problem is how to identify which set of NetworkGurus should manage the network. I believe that selection process should be much more transparent to the administrator who setup the CS deployment but because at the time only VLAN isolation was supported, I did not put that into CS. As more and more SDN technologies come online, it's best we design that process. Can we follow the path of NetworkElement which was once a single base class implementing all service abstractions to set of derived interfaces for each service (DhcpServiceProvider, LoadBalancingServiceProvider etc). With this approach a network element can just implement the service it want to provide. Current guru implementation are tightly coupled with each other. For e.g Midonet, Nicira guru implement the network design for their isolation and for the rest of the functionality they fall back on Guest network guru. Perhaps we can split the abstraction of network guru into multiple derived interface specific to each function like isolation, network design, ip address management. So if Infoblox guru just want to provide IPAM, then it can just focus on it with out the knowledge of other functions of network guru. I think Hugo made some changes there but I haven't had a chance to study how that all works now in 4.x yet. There are some artifacts in the code to show I was thinking about this. For example, Network and Nic carries the Isolation URI instead of just VLAN tag because it was intended to be cast by the NetworkGuru and then transported to the hypervisor. We expected anyone who plugs into CloudStack have two places to plug into: orchestration and actual provisioning. So the NetworkGuru forms the URI for isolation and then it has code at the hypervisor side to resolve the URI to something that the hypervisor can understand. --Alex -Original Message- From: Soheil Eizadi [mailto:seiz...@infoblox.com] Sent: Tuesday, August 6, 2013 11:54 AM To: dev@cloudstack.apache.org Subject: RE: IP Address Allocation Agree that the IP Address allocation is different than the DHCP and DNS service offered by the NetworkElements. The third party NetworkGurus just reference the standard NetworkManager IP Address Allocation routines, either directly (e.g. _networkMgr.assignPublicIpAddress()) or by calling out to the GuestNetworkGuru. I would say in the current design the Network Manager manages the IP Addresses not the Network Gurus. I don't think this is a bad design, in that the NetworkGurus are sharing the same ip address allocation and get new features (e.g. IP reservation or VM migration) without implementing new code to support them. -Soheil From: Alex Huang [alex.hu...@citrix.com] Sent: Tuesday, August 06, 2013 11:27 AM To: Chiradeep Vittal; dev@cloudstack.apache.org Subject: RE: IP Address Allocation We can either break NetworkManager down
Re: weird test failure
I'd say the test needs to be changed to allow for a negative skew. On 8/7/13 2:05 AM, Daan Hoogland daan.hoogl...@gmail.com wrote: I under-kinda-stand, Does this mean I should live with the whole build failing for a while once in a while? Meaning should I change this test and submit a patch or is there some windows or java setting I should look at? On Tue, Aug 6, 2013 at 8:19 PM, Chiradeep Vittal chiradeep.vit...@citrix.com wrote: Some insight here http://www.javatuning.com/why-is-thread-sleep-inherently-inaccurate/ On 8/6/13 8:16 AM, Daan Hoogland daan.hoogl...@gmail.com wrote: H, I just had a time skew that seems impossible: if in the folowing test the sleep is 1000 it reports a failure saying the duration was 999. now I see at least some clock ticks around it so whats heppening? An overeager optimizer, maybe? I had to set it to 1001 to make it work. This is not new code, it has been there since jan 24! So i ran it a few times. public class TestProfiler extends Log4jEnabledTestCase { protected final static Logger s_logger = Logger.getLogger(TestProfiler.class); @Test public void testProfiler() { s_logger.info(testProfiler() started); Profiler pf = new Profiler(); pf.start(); try { Thread.sleep(1001); } catch (InterruptedException e) { } pf.stop(); s_logger.info(Duration : + pf.getDuration()); Assert.assertTrue(pf.getDuration() = 1000); s_logger.info(testProfiler() stopped); } } any clue welcome, Daan
[ACS42] Closing Resolved Defects
Hi All, Below are total and Blocker/Critical defects in resolved state. Pl take some time to close them. Even if you get to close Cortical/Blocker that would be great. Reporter Total Blocker/Critical Abhinandan Prateek 1 1 Abhinav Roy 8 7 Ahmad Emneina 5 5 Alena Prokharchyk 20 4 angeline shen3 3 Anthony Xu4 1 Bharat Kumar6 4 Brian Angus1 1 Brian Spindler1 1 Caleb Call 1 0 Chandan Purushothama 21 18 Chris Sears 1 0 Daan Hoogland 1 1 Dave Brosius 1 0 Dave Cahill 2 0 David Nalley 1 0 David Noland 2 0 david vz1 1 Devdeep Singh 3 3 Diogo Monteiro1 0 edison su 1 1 Fang Wang 2 2 Francois Gaudreault 5 3 frank zhang2 0 Girish Shilamkar2 1 Harikrishna Patnala 9 5 Hiroaki Kawai 2 1 Hongtu Zang 1 0 Hugo Trippaers 1 0 ilya musayev 3 2 Isaac Chiang 4 0 Jayapal Reddy 8 5 Jeronimo 1 0 Jessica Tomechak3 0 Jessica Wang 1 0 John Burwell 2 2 Kelven Yang 1 1 Kishan Kavala 8 2 Koushik Das4 2 Likitha Shetty 8 0 manasaveloori 3 2 Mice Xia 1 0 Michael Dickey 1 1 Mike Tutkowski6 0 Milamber2 0 Min Chen5 4 Minying Bao 4 4 Murali Reddy 9 3 Nick Wales 1 1 Nicolas Lamirault 1 1 Nik Martin 1 0 Nitin Mehta28 6 Parth Jagirdar8 2 Paul Angus 3 1 Pavan Kumar Bandarupally 4 4 Prachi Damle 10 1 Pranav Saxena 1 0 Prasanna Santhanam 46 31 prashant kumar mishra 7 4 Radhika Nair 1 0 Rajesh Battala 5 5 Rayees Namathponnan43 22 Rohit Yadav3 2 roxanne chang 1 0 sadhu suresh 3 3 Sailaja Mada 6 5 Saksham Srivastava 8 3 Sangeetha Hariharan 44 22 Sanjay Tripathi 6 2 Sanjeev N 12 9 Sateesh Chodapuneedi 19 13 satoru nakaya1 1 sebastien goasguen1 0 Shane Witbeck 1 0 Sharad Yadav 1 1 Sheng Yang 15 8 shweta agarwal4 3 Simon Weller 1 1 Sonny Chhen 1 0 Sowmya Krishnan8 2 Srikanteswararao Talluri23 22 Toshiaki Hatano2 0 Venkata Siva Vijayendra Bhamidipati 4 0 venkata swamybabu budumuru 1 1 Wido den Hollander4 0 Grand Total511 267 Thanks /sudha
Re: Values within commands.properties.in
I think 'private zones' are assigned to RESOURCE_DOMAIN_ADMIN On 8/7/13 1:29 AM, Ryan Lei ryan...@cht.com.tw wrote: Curious for long. When does the rule RESOURCE_DOMAIN_ADMIN come in place in CloudStack? A CloudStack account can only be root admin, domain admin, or user. Does RESOURCE_DOMAIN_ADMIN refer to something internal? -- - Yu-Heng (Ryan) Lei, Associate Reasearcher Chunghwa Telecom Laboratories / Cloud Computing Laboratory ryan...@cht.com.twhttps://email.cht.com.tw/owa/redir.aspx?C=-wE1FEC3G0SWY pVkiWo8SsDdf3ZqO9AIuAPTzRnFYCUi-z4YljtI_hyVKkNHfn9F1Bn-vUWJnQ4.URL=mailto %3aryanlei%40cht.com.tw or ryanlei750...@gmail.com On Fri, Aug 2, 2013 at 6:24 PM, Ian Duffy i...@ianduffy.ie wrote: Hi, This query is in relation to the permissions set for commands via client/tomcatconf/commands.properties.in Within the header of the file it states: bitmap of permissions at the end of each classname, 1 = ADMIN, 2 = RESOURCE_DOMAIN_ADMIN, 4 = DOMAIN_ADMIN, 8 = USER However, throughout the file the undocumented numbers of 3, 7 and 15 can be found. Can somebody explain what they mean? Looking within the RoleType enum only the following are defined: Admin = 1 ResourceAdmin = 2 Domain Admin = 4 User = 8 Unknown = 0 Thanks!
RE: [KVM] Helper for agent during HA operations
-Original Message- From: Wido den Hollander [mailto:w...@widodh.nl] Sent: Wednesday, August 07, 2013 10:53 AM To: dev@cloudstack.apache.org Cc: shadow...@gmail.com Subject: [KVM] Helper for agent during HA operations Hi, In our production setups we have seen some crashes of the KVM agent. If we can make sure KVM agent restarted immediately after crash, then you don't another separate service running on your KVM host. Not sure jsvc can automatically restart agent or not, I remember we have a small c daemon program in the 3.0.x source code, which can monitor agent. This could happen for all kinds of reasons, but that's not what I wanted to discuss. Also see this issue: https://issues.apache.org/jira/browse/CLOUDSTACK- 3954 What I've been writing for a PoC in our company is a small helper written in Python which runs on port 8251. The Investigator can query this webservice (attached) which will simply tell it which VMs are running on that host. It's online here: http://stack01.ceph.widodh.nl:8251/ You can also do a query like this: http://stack01.ceph.widodh.nl:8251/ping/i-2-6570-VM This way we can more reliably verify if a specific VM is still running if the Agent stops responding for some reason. A ICMP echo-request isn't safe since the Security Groups could prevent ICMP from coming through. I'd rather not have the management server query libvirt directly, since that would open a potential security whole. This webservice is read-only and on my production setups I have libvirt listening on the private bridge only. What do you think? Wido
Re: Review Request 13376: Fix QuickCloud secondary storage service ListTemplateCommand and ListVolumeCommand
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13376/#review24814 --- Ship it! In the future, please format patches so that they can be applied using git am (authorship info is maintained if you do this). This one has been applied using git apply-patch and git commit -s --author=Donal Lafferty donal.laffe...@citrix.com -m Get QuickCloud working again - Chiradeep Vittal On Aug. 7, 2013, 1:49 p.m., Donal Lafferty wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13376/ --- (Updated Aug. 7, 2013, 1:49 p.m.) Review request for cloudstack, Chiradeep Vittal and Devdeep Singh. Repository: cloudstack-git Description --- Fix non-systemVM execution. ListTemplateCommand and ListVolumeCommand need to return the corresponding Answer object. No bug was logged. Diffs - services/secondary-storage/src/org/apache/cloudstack/storage/resource/NfsSecondaryStorageResource.java 89ce8e009a36df53f10efda8ef7c3d8f7ed6e2dc Diff: https://reviews.apache.org/r/13376/diff/ Testing --- Integration test: launched mgmt srvr with Quick Cloud configuration, launched QuickCloud secondary storage server, verified that no exceptions no longer occurred when secondary storage server connected to mgmt server. Thanks, Donal Lafferty
Re: Review Request 13252: SHA256 timing attack and brute force attack fix
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13252/ --- (Updated Aug. 7, 2013, 7:01 p.m.) Review request for cloudstack and John Kinsella. Changes --- 1. Added test case to ensure (old) 20-byte salted passwords work 2. Test cases for general authentication 3. Test cases to ensure computation time for all different scenarios are within minutely small range of each other Bugs: https://issues.apache.org/jira/browse/CLOUDSTACK-2312 and https://issues.apache.org/jira/browse/CLOUDSTACK-2314 Repository: cloudstack-git Description --- 1. Fix timing attack by using a constant-time comparison function 2. Increase salt size 3. Make flow for invalid user go through full normal execution using a fake password and salt Diffs (updated) - plugins/user-authenticators/sha256salted/src/com/cloud/server/auth/SHA256SaltedUserAuthenticator.java da939273ea10bff3b2687c9684edf8a5d0ab4b2e plugins/user-authenticators/sha256salted/test/src/com/cloud/server/auth/test/AuthenticatorTest.java 4e23d14fe43b4e334203f48196aced038ca0a196 Diff: https://reviews.apache.org/r/13252/diff/ Testing --- Local environment Thanks, Amogh Vasekar
Re: Question about creating a template
Looks like a bug to me. On 8/6/13 11:28 AM, Mike Tutkowski mike.tutkow...@solidfire.com wrote: Hi, I have a small Linux ISO (http://www.damnsmalllinux.org/download.html) I've uploaded to my secondary storage NFS location. I can create VM instances just fine with it. I stopped one of my VMs that was created with this ISO and tried to create a template from its root disk. I received the following exception and was wondering what people's comments might be. Thanks! ERROR [cloud.async.AsyncJobManagerImpl] (Job-Executor-8:job-41 = [ 583f08f2-e32a-4d16-815a-2ac2953d2ae9 ]) Unexpected exception while executing org.apache.cloudstack.api.command.user.template.CreateTemplateCmd com.cloud.utils.exception.CloudRuntimeException: Failed to create templateUnsupported command issued:org.apache.cloudstack.storage.command.CopyCommand. Are you sure you got the right type of server? at com.cloud.template.TemplateManagerImpl.createPrivateTemplate(TemplateManag erImpl.java:1368) at com.cloud.utils.component.ComponentInstantiationPostProcessor$InterceptorD ispatcher.intercept(ComponentInstantiationPostProcessor.java:125) at org.apache.cloudstack.api.command.user.template.CreateTemplateCmd.execute( CreateTemplateCmd.java:263) at com.cloud.api.ApiDispatcher.dispatch(ApiDispatcher.java:158) at com.cloud.async.AsyncJobManagerImpl$1.run(AsyncJobManagerImpl.java:531) at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:439) at java.util.concurrent.FutureTask$Sync.innerRun(FutureTask.java:303) at java.util.concurrent.FutureTask.run(FutureTask.java:138) at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor. java:895) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java :918) at java.lang.Thread.run(Thread.java:680) -- *Mike Tutkowski* *Senior CloudStack Developer, SolidFire Inc.* e: mike.tutkow...@solidfire.com o: 303.746.7302 Advancing the way the world uses the cloudhttp://solidfire.com/solution/overview/?video=play **
Re: Review Request 13252: SHA256 timing attack and brute force attack fix
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13252/#review24818 --- Ship it! Ship It! - Chiradeep Vittal On Aug. 7, 2013, 7:01 p.m., Amogh Vasekar wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13252/ --- (Updated Aug. 7, 2013, 7:01 p.m.) Review request for cloudstack and John Kinsella. Bugs: https://issues.apache.org/jira/browse/CLOUDSTACK-2312 and https://issues.apache.org/jira/browse/CLOUDSTACK-2314 Repository: cloudstack-git Description --- 1. Fix timing attack by using a constant-time comparison function 2. Increase salt size 3. Make flow for invalid user go through full normal execution using a fake password and salt Diffs - plugins/user-authenticators/sha256salted/src/com/cloud/server/auth/SHA256SaltedUserAuthenticator.java da939273ea10bff3b2687c9684edf8a5d0ab4b2e plugins/user-authenticators/sha256salted/test/src/com/cloud/server/auth/test/AuthenticatorTest.java 4e23d14fe43b4e334203f48196aced038ca0a196 Diff: https://reviews.apache.org/r/13252/diff/ Testing --- Local environment Thanks, Amogh Vasekar
Re: weird test failure
ok, in the mean while please have another look at https://reviews.apache.org/r/12849/ please? On Wed, Aug 7, 2013 at 8:01 PM, Chiradeep Vittal chiradeep.vit...@citrix.com wrote: I'd say the test needs to be changed to allow for a negative skew. On 8/7/13 2:05 AM, Daan Hoogland daan.hoogl...@gmail.com wrote: I under-kinda-stand, Does this mean I should live with the whole build failing for a while once in a while? Meaning should I change this test and submit a patch or is there some windows or java setting I should look at? On Tue, Aug 6, 2013 at 8:19 PM, Chiradeep Vittal chiradeep.vit...@citrix.com wrote: Some insight here http://www.javatuning.com/why-is-thread-sleep-inherently-inaccurate/ On 8/6/13 8:16 AM, Daan Hoogland daan.hoogl...@gmail.com wrote: H, I just had a time skew that seems impossible: if in the folowing test the sleep is 1000 it reports a failure saying the duration was 999. now I see at least some clock ticks around it so whats heppening? An overeager optimizer, maybe? I had to set it to 1001 to make it work. This is not new code, it has been there since jan 24! So i ran it a few times. public class TestProfiler extends Log4jEnabledTestCase { protected final static Logger s_logger = Logger.getLogger(TestProfiler.class); @Test public void testProfiler() { s_logger.info(testProfiler() started); Profiler pf = new Profiler(); pf.start(); try { Thread.sleep(1001); } catch (InterruptedException e) { } pf.stop(); s_logger.info(Duration : + pf.getDuration()); Assert.assertTrue(pf.getDuration() = 1000); s_logger.info(testProfiler() stopped); } } any clue welcome, Daan
Re: [KVM] Helper for agent during HA operations
why not use systemd and launchd facilities ? On Wed, Aug 7, 2013 at 8:28 PM, Edison Su edison...@citrix.com wrote: -Original Message- From: Wido den Hollander [mailto:w...@widodh.nl] Sent: Wednesday, August 07, 2013 10:53 AM To: dev@cloudstack.apache.org Cc: shadow...@gmail.com Subject: [KVM] Helper for agent during HA operations Hi, In our production setups we have seen some crashes of the KVM agent. If we can make sure KVM agent restarted immediately after crash, then you don't another separate service running on your KVM host. Not sure jsvc can automatically restart agent or not, I remember we have a small c daemon program in the 3.0.x source code, which can monitor agent. This could happen for all kinds of reasons, but that's not what I wanted to discuss. Also see this issue: https://issues.apache.org/jira/browse/CLOUDSTACK- 3954 What I've been writing for a PoC in our company is a small helper written in Python which runs on port 8251. The Investigator can query this webservice (attached) which will simply tell it which VMs are running on that host. It's online here: http://stack01.ceph.widodh.nl:8251/ You can also do a query like this: http://stack01.ceph.widodh.nl:8251/ping/i-2-6570-VM This way we can more reliably verify if a specific VM is still running if the Agent stops responding for some reason. A ICMP echo-request isn't safe since the Security Groups could prevent ICMP from coming through. I'd rather not have the management server query libvirt directly, since that would open a potential security whole. This webservice is read-only and on my production setups I have libvirt listening on the private bridge only. What do you think? Wido -- Grtz, Jörgen Maas
Re: Help needed with Cloudstack Setup
Hi All, Thanks for your inputs. I have tried with the suggested changes but could not get it working. I started from scratch again to see if it helps. But I seem to be going wrong in some place. Can some one kinldy go though the below setup and tell me if its right. I am stuck since a week trying to get this running :( I have the following setup: Cloudstack Management Server Port 20 of Switch (Trunk all + with necessary Vlans) Hypervisor (Ubuntu + KVM) Port 22 of Switch (Trunk all + with necessary Vlans) I am thinking of having the following networks/sub-networks: mgmt/public/private/storage - 172.16.10.0/24 network Mangement Server - 172.16.10.1, Hypervisor – 172.16.10.10 Public 172.16.10.20 - 172.16.10.30 Private (System VMs e.t.c) 172.16.10.31 - 172.16.10.200 Guest 10.1.1.0/24 *Server Side Configuration* CloudStack Server: eth0 Corporate lab ip (lets say 1.1.1.1) eth1.3100 Subinterface of Eth1 with Vlan 3100, IP: 172.16.10.1 *Global settings* system.vm.use.local.storage = true *Zone Creation Wizard for Advanced Network * Name : Zone-1** DNS1 : 8.8.8.8 Internal DNS1 : 8.8.8.8 Hypervisor : KVM Guest CIDR = 10.1.1.0/24 Local Storage enabled = true *Network Setup* Physical Network 1 : Isolation Method Vlan mgmt -- gave cloudbr0 (on the hypervisor, I have eth1.3100 assigned with IP 172.16.10.10 for mgmt traffic, cant add a bridge to it) public -- cloudbr0 guest -- cloudbr1 *Public IP range * GW : 172.16.10.1/24 (ip of the CS server) Vlan : 3200(Public traffic Vlan) IP range : 172.16.10.20 - 172.16.10.30 *POD* GW : 172.16.10.1/24 (ip of the CS server) IP Range (Private n/w) : 172.16.10.31 - 172.16.10.200 *Guest* - Skipped Vlan range for guest traffic *Cluster* Name : Cluster-1 Hypervisor : KVM *Host* Name : Corporate Lab IP for Host (lets say 2.2.2.2) username password *Primary Storage* : Gets skipped as I have selected local storage to true. *Secondary Storage* Provider : NFS Name : SecondaryStorage-1 NFS Server : CS Server Corporate ip (1.1.1.1) path : /mnt/secondary *Hypervisor (Ubuntu 13.04 + KVM)* eth0 Corporate Lab ip (2.2.2.2) eth1.3100 172.16.10.10 GW: 172.16.10.1 eth1.3200 Added to bridge Cloudbr0 eth1.3300 Added to bridge Cloudbr1 Agent is setup properly and is able to communicate with the server. *Is this configuration correct or am I missing something??* I have seeded the SystemVM with the proper 4.2 templates as per Prasanna's suggestion today. My Setup due to some reason is not able to start the console proxy SSVMs. Thanks in advance for your suggestions. regards, Sarath
4.2 build failing with latest head
4.2 build failing with latest head, someone please look this ASAP ? [INFO] --- maven-resources-plugin:2.5:testResources (default-testResources) @ cloud-api --- [debug] execute contextualize [INFO] Using 'UTF-8' encoding to copy filtered resources. [INFO] skip non existing resourceDirectory /root/jenkins/build/workspace/CloudPlatform-4.2-rhel63/cloudplatform/cloudstack/dist/rpmbuild/BUILD/cloudstack-4.2.0-SNAPSHOT/api/test/resources [INFO] Copying 3 resources [INFO] Copying 3 resources [INFO] [INFO] --- maven-compiler-plugin:2.5.1:testCompile (default-testCompile) @ cloud-api --- [INFO] Compiling 13 source files to /root/jenkins/build/workspace/CloudPlatform-4.2-rhel63/cloudplatform/cloudstack/dist/rpmbuild/BUILD/cloudstack-4.2.0-SNAPSHOT/api/target/test-classes [INFO] - [ERROR] COMPILATION ERROR : [INFO] - [ERROR] AddSecondaryStorageCmdTest.java:[22,32] cannot find symbol symbol : class ResponseGenerator location: package org.apache.cloudstack.api [ERROR] AddSecondaryStorageCmdTest.java:[23,32] cannot find symbol symbol : class ServerApiException location: package org.apache.cloudstack.api [ERROR] AddSecondaryStorageCmdTest.java:[24,54] package org.apache.cloudstack.api.command.admin.storage does not exist [ERROR] AddSecondaryStorageCmdTest.java:[25,41] package org.apache.cloudstack.api.response does not exist [ERROR] AddSecondaryStorageCmdTest.java:[26,41] package org.apache.cloudstack.api.response does not exist [ERROR] AddSecondaryStorageCmdTest.java:[33,24] package com.cloud.storage does not exist [ERROR] AddSecondaryStorageCmdTest.java:[34,24] package com.cloud.storage does not exist [ERROR] AddSecondaryStorageCmdTest.java:[38,12] cannot find symbol symbol : class AddImageStoreCmd location: class org.apache.cloudstack.api.command.test.AddSecondaryStorageCmdTest [ERROR] AddHostCmdTest.java:[22,32] cannot find symbol symbol : class ResponseGenerator location: package org.apache.cloudstack.api [ERROR] AddHostCmdTest.java:[23,32] cannot find symbol symbol : class ServerApiException location: package org.apache.cloudstack.api [ERROR] AddHostCmdTest.java:[24,51] package org.apache.cloudstack.api.command.admin.host does not exist [ERROR] AddHostCmdTest.java:[25,41] package org.apache.cloudstack.api.response does not exist [ERROR] AddHostCmdTest.java:[26,41] package org.apache.cloudstack.api.response does not exist [ERROR] AddHostCmdTest.java:[33,26] package com.cloud.exception does not exist [ERROR] AddHostCmdTest.java:[34,26] package com.cloud.exception does not exist [ERROR] AddHostCmdTest.java:[35,21] package com.cloud.host does not exist [ERROR] AddHostCmdTest.java:[36,25] package com.cloud.resource does not exist [ERROR] AddHostCmdTest.java:[42,12] cannot find symbol symbol : class AddHostCmd location: class org.apache.cloudstack.api.command.test.AddHostCmdTest [ERROR] AddHostCmdTest.java:[43,12] cannot find symbol symbol : class ResourceService location: class org.apache.cloudstack.api.command.test.AddHostCmdTest Regards, Rayees
Re: Help needed with Cloudstack Setup
Local storage wasn't working at all yesterday in 4.2. Fixed last night. Not sure if that is related. On Aug 7, 2013 1:32 PM, Sarath Chandra sarath.cloudst...@gmail.com wrote: Hi All, Thanks for your inputs. I have tried with the suggested changes but could not get it working. I started from scratch again to see if it helps. But I seem to be going wrong in some place. Can some one kinldy go though the below setup and tell me if its right. I am stuck since a week trying to get this running :( I have the following setup: Cloudstack Management Server Port 20 of Switch (Trunk all + with necessary Vlans) Hypervisor (Ubuntu + KVM) Port 22 of Switch (Trunk all + with necessary Vlans) I am thinking of having the following networks/sub-networks: mgmt/public/private/storage - 172.16.10.0/24 network Mangement Server - 172.16.10.1, Hypervisor – 172.16.10.10 Public 172.16.10.20 - 172.16.10.30 Private (System VMs e.t.c) 172.16.10.31 - 172.16.10.200 Guest 10.1.1.0/24 *Server Side Configuration* CloudStack Server: eth0 Corporate lab ip (lets say 1.1.1.1) eth1.3100 Subinterface of Eth1 with Vlan 3100, IP: 172.16.10.1 *Global settings* system.vm.use.local.storage = true *Zone Creation Wizard for Advanced Network * Name : Zone-1** DNS1 : 8.8.8.8 Internal DNS1 : 8.8.8.8 Hypervisor : KVM Guest CIDR = 10.1.1.0/24 Local Storage enabled = true *Network Setup* Physical Network 1 : Isolation Method Vlan mgmt -- gave cloudbr0 (on the hypervisor, I have eth1.3100 assigned with IP 172.16.10.10 for mgmt traffic, cant add a bridge to it) public -- cloudbr0 guest -- cloudbr1 *Public IP range * GW : 172.16.10.1/24 (ip of the CS server) Vlan : 3200(Public traffic Vlan) IP range : 172.16.10.20 - 172.16.10.30 *POD* GW : 172.16.10.1/24 (ip of the CS server) IP Range (Private n/w) : 172.16.10.31 - 172.16.10.200 *Guest* - Skipped Vlan range for guest traffic *Cluster* Name : Cluster-1 Hypervisor : KVM *Host* Name : Corporate Lab IP for Host (lets say 2.2.2.2) username password *Primary Storage* : Gets skipped as I have selected local storage to true. *Secondary Storage* Provider : NFS Name : SecondaryStorage-1 NFS Server : CS Server Corporate ip (1.1.1.1) path : /mnt/secondary *Hypervisor (Ubuntu 13.04 + KVM)* eth0 Corporate Lab ip (2.2.2.2) eth1.3100 172.16.10.10 GW: 172.16.10.1 eth1.3200 Added to bridge Cloudbr0 eth1.3300 Added to bridge Cloudbr1 Agent is setup properly and is able to communicate with the server. *Is this configuration correct or am I missing something??* I have seeded the SystemVM with the proper 4.2 templates as per Prasanna's suggestion today. My Setup due to some reason is not able to start the console proxy SSVMs. Thanks in advance for your suggestions. regards, Sarath
Re: Help needed with Cloudstack Setup
On Wed, Aug 07, 2013 at 01:50:28PM -0600, Marcus Sorensen wrote: Local storage wasn't working at all yesterday in 4.2. Fixed last night. Not sure if that is related. The master / 4.2 quality seems to be churning a ton... this isn't a good sign IMO. Did we cram in too much at the end again?
Re: Help needed with Cloudstack Setup
No idea, it's been mostly off of my radar. I just happened to try to test a few fixes and 4.2 was having problems with local storage. At the least I'd say it's not a good idea for someone to use it (or any non-production version) right now if they're setting up cloudstack for the first time. On Wed, Aug 7, 2013 at 2:04 PM, Chip Childers chip.child...@sungard.com wrote: On Wed, Aug 07, 2013 at 01:50:28PM -0600, Marcus Sorensen wrote: Local storage wasn't working at all yesterday in 4.2. Fixed last night. Not sure if that is related. The master / 4.2 quality seems to be churning a ton... this isn't a good sign IMO. Did we cram in too much at the end again?
RE: 4.2 build failing with latest head
Latest build works; please ignore -Original Message- From: Rayees Namathponnan [mailto:rayees.namathpon...@citrix.com] Sent: Wednesday, August 07, 2013 12:45 PM To: dev@cloudstack.apache.org Subject: 4.2 build failing with latest head 4.2 build failing with latest head, someone please look this ASAP ? [INFO] --- maven-resources-plugin:2.5:testResources (default-testResources) @ cloud-api --- [debug] execute contextualize [INFO] Using 'UTF-8' encoding to copy filtered resources. [INFO] skip non existing resourceDirectory /root/jenkins/build/workspace/CloudPlatform-4.2-rhel63/cloudplatform/cloudstack/dist/rpmbuild/BUILD/cloudstack-4.2.0-SNAPSHOT/api/test/resources [INFO] Copying 3 resources [INFO] Copying 3 resources [INFO] [INFO] --- maven-compiler-plugin:2.5.1:testCompile (default-testCompile) @ cloud-api --- [INFO] Compiling 13 source files to /root/jenkins/build/workspace/CloudPlatform-4.2-rhel63/cloudplatform/cloudstack/dist/rpmbuild/BUILD/cloudstack-4.2.0-SNAPSHOT/api/target/test-classes [INFO] - [ERROR] COMPILATION ERROR : [INFO] - [ERROR] AddSecondaryStorageCmdTest.java:[22,32] cannot find symbol symbol : class ResponseGenerator location: package org.apache.cloudstack.api [ERROR] AddSecondaryStorageCmdTest.java:[23,32] cannot find symbol symbol : class ServerApiException location: package org.apache.cloudstack.api [ERROR] AddSecondaryStorageCmdTest.java:[24,54] package org.apache.cloudstack.api.command.admin.storage does not exist [ERROR] AddSecondaryStorageCmdTest.java:[25,41] package org.apache.cloudstack.api.response does not exist [ERROR] AddSecondaryStorageCmdTest.java:[26,41] package org.apache.cloudstack.api.response does not exist [ERROR] AddSecondaryStorageCmdTest.java:[33,24] package com.cloud.storage does not exist [ERROR] AddSecondaryStorageCmdTest.java:[34,24] package com.cloud.storage does not exist [ERROR] AddSecondaryStorageCmdTest.java:[38,12] cannot find symbol symbol : class AddImageStoreCmd location: class org.apache.cloudstack.api.command.test.AddSecondaryStorageCmdTest [ERROR] AddHostCmdTest.java:[22,32] cannot find symbol symbol : class ResponseGenerator location: package org.apache.cloudstack.api [ERROR] AddHostCmdTest.java:[23,32] cannot find symbol symbol : class ServerApiException location: package org.apache.cloudstack.api [ERROR] AddHostCmdTest.java:[24,51] package org.apache.cloudstack.api.command.admin.host does not exist [ERROR] AddHostCmdTest.java:[25,41] package org.apache.cloudstack.api.response does not exist [ERROR] AddHostCmdTest.java:[26,41] package org.apache.cloudstack.api.response does not exist [ERROR] AddHostCmdTest.java:[33,26] package com.cloud.exception does not exist [ERROR] AddHostCmdTest.java:[34,26] package com.cloud.exception does not exist [ERROR] AddHostCmdTest.java:[35,21] package com.cloud.host does not exist [ERROR] AddHostCmdTest.java:[36,25] package com.cloud.resource does not exist [ERROR] AddHostCmdTest.java:[42,12] cannot find symbol symbol : class AddHostCmd location: class org.apache.cloudstack.api.command.test.AddHostCmdTest [ERROR] AddHostCmdTest.java:[43,12] cannot find symbol symbol : class ResourceService location: class org.apache.cloudstack.api.command.test.AddHostCmdTest Regards, Rayees
[DISCUSS] KVM HA
There is long time issue related to KVM HA, see bug: CLOUDSTACK-3535. Basically, HA won't be triggered, if KVM agent is stopped either normally nor abnormally, HA only be triggered if the network between mgt server and kvm host is disconnected and the network between KVM hosts in the same cluster is disconnected. Here is how the KVM HA works after the fix for CLOUDSTACK-3535: 1. If agent is stopped, agent will send a shutdown request to mgt server, mgt server will mark the host as disconnected, while still maintain the host in pingmap. Code is in AgentManagerImpl-AgentHandler- ProcessRequest- disconnectWithoutInvestigation 2. After ping.interval, mgt server will find the host is ping timeout, then start HA investigation for the host. Code is in AgentMonitor-run- disconnectWithInvestigation 3. Mgt server will call all the available Investigators to investigate the status of host. The current investigators will be called for KVM host: UserVmDomRInvestigator-isAgentAlive, will send PingTestCommand to the host's neighbor. PingTestCommand will ping host's private ip address, if ping is reachable, means host is up, otherwise, host's state is unknown. So this investigator can only detect host is in up state. KVMInvestigator, which is newly added, will send a CheckOnHostCommand to host's neighbor. CheckOnHostCommand will check the heartbeat of host(heartbeat is stored on shared primary storage). Ideally, it will detect host is down or up. Combined with UserVmDomRInvestigator and KVMInvestigator, mgt server should find out the status of host. But there is case, these two investigators can report wrong status of host: Host is in a network partition, while the KVM agent is down(thus heartbeat is stopped) 4. After investigator reports status of host, if host is down, then start HA for VMs created on the host. Improvement: Per suggestion from Lennert den Teuling, we'd better use IPMI to detect host status, which is more reliable than ping and heartbeat, as IPMI has its own network, less likely has network partition.
Re: [KVM] Helper for agent during HA operations
For me the most useful part is that I can take down the agent without anything freaking out. It's fairly common to upgrade the agent (or stop for a minute or two) without putting the host into maintenance and dealing with migrations. That's by far the normal case for when the agent isn't running, I don't think I've seen it crash yet or become unresponsive (knock on wood). This works fine now, but it probably won't once the HA stuff is patched up, so something like this is nice. However, I assume that it would be a part of the agent packaging, thus anything running this would also be killed/upgraded during such a procedure. I suppose it doesn't have to be like that, but it seems like a bit of a juggling act to keep two services going and make sure at least one is running at all times during upgrades. On Wed, Aug 7, 2013 at 1:30 PM, Jörgen Maas jorgen.m...@gmail.com wrote: why not use systemd and launchd facilities ? On Wed, Aug 7, 2013 at 8:28 PM, Edison Su edison...@citrix.com wrote: -Original Message- From: Wido den Hollander [mailto:w...@widodh.nl] Sent: Wednesday, August 07, 2013 10:53 AM To: dev@cloudstack.apache.org Cc: shadow...@gmail.com Subject: [KVM] Helper for agent during HA operations Hi, In our production setups we have seen some crashes of the KVM agent. If we can make sure KVM agent restarted immediately after crash, then you don't another separate service running on your KVM host. Not sure jsvc can automatically restart agent or not, I remember we have a small c daemon program in the 3.0.x source code, which can monitor agent. This could happen for all kinds of reasons, but that's not what I wanted to discuss. Also see this issue: https://issues.apache.org/jira/browse/CLOUDSTACK- 3954 What I've been writing for a PoC in our company is a small helper written in Python which runs on port 8251. The Investigator can query this webservice (attached) which will simply tell it which VMs are running on that host. It's online here: http://stack01.ceph.widodh.nl:8251/ You can also do a query like this: http://stack01.ceph.widodh.nl:8251/ping/i-2-6570-VM This way we can more reliably verify if a specific VM is still running if the Agent stops responding for some reason. A ICMP echo-request isn't safe since the Security Groups could prevent ICMP from coming through. I'd rather not have the management server query libvirt directly, since that would open a potential security whole. This webservice is read-only and on my production setups I have libvirt listening on the private bridge only. What do you think? Wido -- Grtz, Jörgen Maas
Re: [DISCUSS] KVM HA
Does KVMInvestigator work on all shared primary storage, or just NFS? I'm only familiar with the NFS KVMHA directories. From this it seems like a clean stop of the KVM agent still shouldn't trigger any issues/HA, correct? On Wed, Aug 7, 2013 at 2:28 PM, Edison Su edison...@citrix.com wrote: There is long time issue related to KVM HA, see bug: CLOUDSTACK-3535. Basically, HA won't be triggered, if KVM agent is stopped either normally nor abnormally, HA only be triggered if the network between mgt server and kvm host is disconnected and the network between KVM hosts in the same cluster is disconnected. Here is how the KVM HA works after the fix for CLOUDSTACK-3535: 1. If agent is stopped, agent will send a shutdown request to mgt server, mgt server will mark the host as disconnected, while still maintain the host in pingmap. Code is in AgentManagerImpl-AgentHandler- ProcessRequest- disconnectWithoutInvestigation 2. After ping.interval, mgt server will find the host is ping timeout, then start HA investigation for the host. Code is in AgentMonitor-run- disconnectWithInvestigation 3. Mgt server will call all the available Investigators to investigate the status of host. The current investigators will be called for KVM host: UserVmDomRInvestigator-isAgentAlive, will send PingTestCommand to the host's neighbor. PingTestCommand will ping host's private ip address, if ping is reachable, means host is up, otherwise, host's state is unknown. So this investigator can only detect host is in up state. KVMInvestigator, which is newly added, will send a CheckOnHostCommand to host's neighbor. CheckOnHostCommand will check the heartbeat of host(heartbeat is stored on shared primary storage). Ideally, it will detect host is down or up. Combined with UserVmDomRInvestigator and KVMInvestigator, mgt server should find out the status of host. But there is case, these two investigators can report wrong status of host: Host is in a network partition, while the KVM agent is down(thus heartbeat is stopped) 4. After investigator reports status of host, if host is down, then start HA for VMs created on the host. Improvement: Per suggestion from Lennert den Teuling, we'd better use IPMI to detect host status, which is more reliable than ping and heartbeat, as IPMI has its own network, less likely has network partition.
Re: Review Request 13323: if a failure occurs while adding VM to another network (this should be the first vm in the subnet). The ip alias created as a part of this process is not removed.
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13323/#review24826 --- Ship it! Ship It! - Sheng Yang On Aug. 7, 2013, 8:53 a.m., bharat kumar wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13323/ --- (Updated Aug. 7, 2013, 8:53 a.m.) Review request for cloudstack, Alena Prokharchyk and Sheng Yang. Bugs: Cloudstack-4083 Repository: cloudstack-git Description --- https://issues.apache.org/jira/browse/CLOUDSTACK-4083 if a failure occurs while adding VM to another network (this should be the first vm in the subnet). The ip alias created as a part of this process is not removed. This occurred because we were not cleaning the alias ips in the event of a failure. As a part of the fix. 1.) moved the function removeDhcpServiceInsubnet and listLastNicsInSubnet to NetworkManager (These were in VirtualMachineManagerImpl earlier.) 2.) add the call to clean ipAlias in the remove nic function of the networkManager. 3.) Modified the removeDhcpServiceInsubnet to take network ad an argument. This will help in removing only the ipAlias which belong to a particular network. Diffs - server/src/com/cloud/network/NetworkManager.java dab7a13 server/src/com/cloud/network/NetworkManagerImpl.java ad79a2f server/src/com/cloud/vm/VirtualMachineManagerImpl.java b33ee49 server/test/com/cloud/network/MockNetworkManagerImpl.java 7f34e55 server/test/com/cloud/vpc/MockNetworkManagerImpl.java 178ebba Diff: https://reviews.apache.org/r/13323/diff/ Testing --- Tested on 4.2. created two guest networks guest1 and guest2. created VMS in both the networks. Added a VM(with no PV drivers) from guest1 to guest2. this failed and on failure ipAlias configured as part of nic creation was removed. Deleting the vm causes all the removal of all ipAliases from all the subnets in which this is the lastvm. Thanks, bharat kumar
Re: [DISCUSS] KVM HA
I'm not sure we can rely on IPMI to tell us much about the host status itself. It's easy to use it for checking on basic poweron/poweroff, temperature, etc, but not so easy to tell if something is wrong with the OS, config, or at the software level. However, I did mention support in that thread early on for sending an IPMI poweroff for hosts that cloudstack has determined are down and starts migrating vms for, as a safety precaution. On Wed, Aug 7, 2013 at 2:41 PM, Marcus Sorensen shadow...@gmail.com wrote: Does KVMInvestigator work on all shared primary storage, or just NFS? I'm only familiar with the NFS KVMHA directories. From this it seems like a clean stop of the KVM agent still shouldn't trigger any issues/HA, correct? On Wed, Aug 7, 2013 at 2:28 PM, Edison Su edison...@citrix.com wrote: There is long time issue related to KVM HA, see bug: CLOUDSTACK-3535. Basically, HA won't be triggered, if KVM agent is stopped either normally nor abnormally, HA only be triggered if the network between mgt server and kvm host is disconnected and the network between KVM hosts in the same cluster is disconnected. Here is how the KVM HA works after the fix for CLOUDSTACK-3535: 1. If agent is stopped, agent will send a shutdown request to mgt server, mgt server will mark the host as disconnected, while still maintain the host in pingmap. Code is in AgentManagerImpl-AgentHandler- ProcessRequest- disconnectWithoutInvestigation 2. After ping.interval, mgt server will find the host is ping timeout, then start HA investigation for the host. Code is in AgentMonitor-run- disconnectWithInvestigation 3. Mgt server will call all the available Investigators to investigate the status of host. The current investigators will be called for KVM host: UserVmDomRInvestigator-isAgentAlive, will send PingTestCommand to the host's neighbor. PingTestCommand will ping host's private ip address, if ping is reachable, means host is up, otherwise, host's state is unknown. So this investigator can only detect host is in up state. KVMInvestigator, which is newly added, will send a CheckOnHostCommand to host's neighbor. CheckOnHostCommand will check the heartbeat of host(heartbeat is stored on shared primary storage). Ideally, it will detect host is down or up. Combined with UserVmDomRInvestigator and KVMInvestigator, mgt server should find out the status of host. But there is case, these two investigators can report wrong status of host: Host is in a network partition, while the KVM agent is down(thus heartbeat is stopped) 4. After investigator reports status of host, if host is down, then start HA for VMs created on the host. Improvement: Per suggestion from Lennert den Teuling, we'd better use IPMI to detect host status, which is more reliable than ping and heartbeat, as IPMI has its own network, less likely has network partition.
Re: IP Address Allocation
Yes, I think so On 8/7/13 10:59 AM, Soheil Eizadi seiz...@infoblox.com wrote: I think this is a consensus of where we are: We would follow the path of NetworkElement meaning the infrastructure to create a service offering, the provider concept and the use of agent/host infrastructure to support external providers. (The external NetworkGuru also uses many of the same infrastructure.) This would allow developing a plugin, if needed that could provide in addition to NetworkElement and NetworkGuru features the new External IP Address Allocation capability. The points Alex made was to complete the work in the NetworkManager and refactor out the IP Address Allocation out into a unique class that can be called by NetworkGurus directly rather than using the NetworkManager. This new Class would have logic to call to the external IP Address Providers or use internal database to provide allocation. I hope I captured the essence of where we are with this? -Soheil From: Murali Reddy [murali.re...@citrix.com] Sent: Wednesday, August 07, 2013 7:43 AM To: dev@cloudstack.apache.org Subject: Re: IP Address Allocation On 07/08/13 1:40 AM, Alex Huang alex.hu...@citrix.com wrote: I'll say it's a bad design. He he I designed it. :( We had to make some compromises due to time constraints. I always intended to move the ip allocation into NetworkGurus and actually attempted to several times but, at the end for various reasons, we just couldn't complete that task. I generally believe that ip address allocation can actually be done even without CloudStack's database, for example, if it came from a master dhcp server or some IT central database. So an interface like IpAddressReserver to abstract that functionality would be great. I have to say the number of implementations for NetworkGuru grew faster than I expected. In one sense, it's a good sign that everyone was able to use it but then it exposes the work that was incomplete to begin with. IIRC, there's basically two problems I left on NetworkGuru to be better designed. One is this ip address allocation. The other problem is how to identify which set of NetworkGurus should manage the network. I believe that selection process should be much more transparent to the administrator who setup the CS deployment but because at the time only VLAN isolation was supported, I did not put that into CS. As more and more SDN technologies come online, it's best we design that process. Can we follow the path of NetworkElement which was once a single base class implementing all service abstractions to set of derived interfaces for each service (DhcpServiceProvider, LoadBalancingServiceProvider etc). With this approach a network element can just implement the service it want to provide. Current guru implementation are tightly coupled with each other. For e.g Midonet, Nicira guru implement the network design for their isolation and for the rest of the functionality they fall back on Guest network guru. Perhaps we can split the abstraction of network guru into multiple derived interface specific to each function like isolation, network design, ip address management. So if Infoblox guru just want to provide IPAM, then it can just focus on it with out the knowledge of other functions of network guru. I think Hugo made some changes there but I haven't had a chance to study how that all works now in 4.x yet. There are some artifacts in the code to show I was thinking about this. For example, Network and Nic carries the Isolation URI instead of just VLAN tag because it was intended to be cast by the NetworkGuru and then transported to the hypervisor. We expected anyone who plugs into CloudStack have two places to plug into: orchestration and actual provisioning. So the NetworkGuru forms the URI for isolation and then it has code at the hypervisor side to resolve the URI to something that the hypervisor can understand. --Alex -Original Message- From: Soheil Eizadi [mailto:seiz...@infoblox.com] Sent: Tuesday, August 6, 2013 11:54 AM To: dev@cloudstack.apache.org Subject: RE: IP Address Allocation Agree that the IP Address allocation is different than the DHCP and DNS service offered by the NetworkElements. The third party NetworkGurus just reference the standard NetworkManager IP Address Allocation routines, either directly (e.g. _networkMgr.assignPublicIpAddress()) or by calling out to the GuestNetworkGuru. I would say in the current design the Network Manager manages the IP Addresses not the Network Gurus. I don't think this is a bad design, in that the NetworkGurus are sharing the same ip address allocation and get new features (e.g. IP reservation or VM migration) without implementing new code to support them. -Soheil From: Alex Huang [alex.hu...@citrix.com] Sent: Tuesday, August 06, 2013 11:27 AM To: Chiradeep Vittal; dev@cloudstack.apache.org Subject:
Re: Review Request 13323: if a failure occurs while adding VM to another network (this should be the first vm in the subnet). The ip alias created as a part of this process is not removed.
On Aug. 7, 2013, 8:46 p.m., Sheng Yang wrote: Ship It! Please uppercase the bug id(CLOUDSTACK- rather than Cloudstack-xxx) next time, otherwise jira won't recognize it. I also removed the unnecessary import and blank line. - Sheng --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13323/#review24826 --- On Aug. 7, 2013, 8:53 a.m., bharat kumar wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13323/ --- (Updated Aug. 7, 2013, 8:53 a.m.) Review request for cloudstack, Alena Prokharchyk and Sheng Yang. Bugs: Cloudstack-4083 Repository: cloudstack-git Description --- https://issues.apache.org/jira/browse/CLOUDSTACK-4083 if a failure occurs while adding VM to another network (this should be the first vm in the subnet). The ip alias created as a part of this process is not removed. This occurred because we were not cleaning the alias ips in the event of a failure. As a part of the fix. 1.) moved the function removeDhcpServiceInsubnet and listLastNicsInSubnet to NetworkManager (These were in VirtualMachineManagerImpl earlier.) 2.) add the call to clean ipAlias in the remove nic function of the networkManager. 3.) Modified the removeDhcpServiceInsubnet to take network ad an argument. This will help in removing only the ipAlias which belong to a particular network. Diffs - server/src/com/cloud/network/NetworkManager.java dab7a13 server/src/com/cloud/network/NetworkManagerImpl.java ad79a2f server/src/com/cloud/vm/VirtualMachineManagerImpl.java b33ee49 server/test/com/cloud/network/MockNetworkManagerImpl.java 7f34e55 server/test/com/cloud/vpc/MockNetworkManagerImpl.java 178ebba Diff: https://reviews.apache.org/r/13323/diff/ Testing --- Tested on 4.2. created two guest networks guest1 and guest2. created VMS in both the networks. Added a VM(with no PV drivers) from guest1 to guest2. this failed and on failure ipAlias configured as part of nic creation was removed. Deleting the vm causes all the removal of all ipAliases from all the subnets in which this is the lastvm. Thanks, bharat kumar
Re: Review Request 13323: if a failure occurs while adding VM to another network (this should be the first vm in the subnet). The ip alias created as a part of this process is not removed.
On Aug. 7, 2013, 8:46 p.m., Sheng Yang wrote: Ship It! Sheng Yang wrote: Please uppercase the bug id(CLOUDSTACK- rather than Cloudstack-xxx) next time, otherwise jira won't recognize it. I also removed the unnecessary import and blank line. Pushed to 4.2 Bharat, please send a patch for MASTER as well. It's not applied now. - Sheng --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13323/#review24826 --- On Aug. 7, 2013, 8:53 a.m., bharat kumar wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13323/ --- (Updated Aug. 7, 2013, 8:53 a.m.) Review request for cloudstack, Alena Prokharchyk and Sheng Yang. Bugs: Cloudstack-4083 Repository: cloudstack-git Description --- https://issues.apache.org/jira/browse/CLOUDSTACK-4083 if a failure occurs while adding VM to another network (this should be the first vm in the subnet). The ip alias created as a part of this process is not removed. This occurred because we were not cleaning the alias ips in the event of a failure. As a part of the fix. 1.) moved the function removeDhcpServiceInsubnet and listLastNicsInSubnet to NetworkManager (These were in VirtualMachineManagerImpl earlier.) 2.) add the call to clean ipAlias in the remove nic function of the networkManager. 3.) Modified the removeDhcpServiceInsubnet to take network ad an argument. This will help in removing only the ipAlias which belong to a particular network. Diffs - server/src/com/cloud/network/NetworkManager.java dab7a13 server/src/com/cloud/network/NetworkManagerImpl.java ad79a2f server/src/com/cloud/vm/VirtualMachineManagerImpl.java b33ee49 server/test/com/cloud/network/MockNetworkManagerImpl.java 7f34e55 server/test/com/cloud/vpc/MockNetworkManagerImpl.java 178ebba Diff: https://reviews.apache.org/r/13323/diff/ Testing --- Tested on 4.2. created two guest networks guest1 and guest2. created VMS in both the networks. Added a VM(with no PV drivers) from guest1 to guest2. this failed and on failure ipAlias configured as part of nic creation was removed. Deleting the vm causes all the removal of all ipAliases from all the subnets in which this is the lastvm. Thanks, bharat kumar
Re: Comment about Data Disk Offering when creating a VM from an ISO
I went ahead and logged a JIRA ticket for this issue. On Wed, Aug 7, 2013 at 10:14 AM, Mike Tutkowski mike.tutkow...@solidfire.com wrote: That sounds reasonable to me. Do you know if a bug exists for this or perhaps I should create one? On Wed, Aug 7, 2013 at 2:03 AM, Ryan Lei ryan...@cht.com.tw wrote: I noticed this issue long ago, and I agree with you. In my understanding, while the tab in the VM wizard always says Data Disk Offering, if you create a VM using ISO, this step actually means Root Disk Offering. If this is the case, the tab name should depend on whether the user chooses Select Template (Data Disk Offering) or Select ISO (Root Disk Offering). --- Yu-Heng (Ryan) Lei, Associate Reasearcher Chunghwa Telecom Laboratories / Cloud Computing Laboratory ryan...@cht.com.tw https://email.cht.com.tw/owa/redir.aspx?C=-wE1FEC3G0SWYpVkiWo8SsDdf3ZqO9AIuAPTzRnFYCUi-z4YljtI_hyVKkNHfn9F1Bn-vUWJnQ4.URL=mailto%3aryanlei%40cht.com.tw or ryanlei750...@gmail.com On Wed, Aug 7, 2013 at 9:12 AM, Mike Tutkowski mike.tutkow...@solidfire.com wrote: Hi, I noticed while creating a VM from an ISO today that I am taken to a page in the wizard that talks about a Data Disk Offering. It seems a bit confusing to me because I don't actually get a data disk once the wizard has spun up the VM. This Data Disk Offering page is displayed to collect data regarding the size of the root disk the VM will have. Any thoughts on this? Is this a known issue or should I log a bug? Thanks! -- *Mike Tutkowski* *Senior CloudStack Developer, SolidFire Inc.* e: mike.tutkow...@solidfire.com o: 303.746.7302 Advancing the way the world uses the cloudhttp://solidfire.com/solution/overview/?video=play *™* -- *Mike Tutkowski* *Senior CloudStack Developer, SolidFire Inc.* e: mike.tutkow...@solidfire.com o: 303.746.7302 Advancing the way the world uses the cloudhttp://solidfire.com/solution/overview/?video=play *™* -- *Mike Tutkowski* *Senior CloudStack Developer, SolidFire Inc.* e: mike.tutkow...@solidfire.com o: 303.746.7302 Advancing the way the world uses the cloudhttp://solidfire.com/solution/overview/?video=play *™*
RE: [ACS42] Closing Resolved Defects
I bulk updated the comment for these tickets asking for verifying the issues Animesh -Original Message- From: Sudha Ponnaganti [mailto:sudha.ponnaga...@citrix.com] Sent: Wednesday, August 07, 2013 11:02 AM To: dev@cloudstack.apache.org Subject: [ACS42] Closing Resolved Defects Hi All, Below are total and Blocker/Critical defects in resolved state. Pl take some time to close them. Even if you get to close Cortical/Blocker that would be great. Reporter Total Blocker/Critical Abhinandan Prateek 1 1 Abhinav Roy 8 7 Ahmad Emneina 5 5 Alena Prokharchyk 20 4 angeline shen3 3 Anthony Xu4 1 Bharat Kumar6 4 Brian Angus1 1 Brian Spindler1 1 Caleb Call 1 0 Chandan Purushothama 21 18 Chris Sears 1 0 Daan Hoogland 1 1 Dave Brosius 1 0 Dave Cahill 2 0 David Nalley 1 0 David Noland 2 0 david vz1 1 Devdeep Singh 3 3 Diogo Monteiro1 0 edison su 1 1 Fang Wang 2 2 Francois Gaudreault 5 3 frank zhang2 0 Girish Shilamkar2 1 Harikrishna Patnala 9 5 Hiroaki Kawai 2 1 Hongtu Zang 1 0 Hugo Trippaers 1 0 ilya musayev 3 2 Isaac Chiang 4 0 Jayapal Reddy 8 5 Jeronimo 1 0 Jessica Tomechak3 0 Jessica Wang 1 0 John Burwell 2 2 Kelven Yang 1 1 Kishan Kavala 8 2 Koushik Das4 2 Likitha Shetty 8 0 manasaveloori 3 2 Mice Xia 1 0 Michael Dickey 1 1 Mike Tutkowski6 0 Milamber2 0 Min Chen5 4 Minying Bao 4 4 Murali Reddy 9 3 Nick Wales 1 1 Nicolas Lamirault 1 1 Nik Martin 1 0 Nitin Mehta28 6 Parth Jagirdar8 2 Paul Angus 3 1 Pavan Kumar Bandarupally 4 4 Prachi Damle 10 1 Pranav Saxena 1 0 Prasanna Santhanam 46 31 prashant kumar mishra 7 4 Radhika Nair 1 0 Rajesh Battala 5 5 Rayees Namathponnan43 22 Rohit Yadav3 2 roxanne chang 1 0 sadhu suresh 3 3 Sailaja Mada 6 5 Saksham Srivastava 8 3 Sangeetha Hariharan 44 22 Sanjay Tripathi 6 2 Sanjeev N 12 9 Sateesh Chodapuneedi 19 13 satoru nakaya1 1 sebastien goasguen1 0 Shane Witbeck 1 0 Sharad Yadav 1 1 Sheng Yang 15 8 shweta agarwal4 3 Simon Weller 1 1 Sonny Chhen 1 0 Sowmya Krishnan8 2 Srikanteswararao Talluri23 22 Toshiaki Hatano2 0 Venkata Siva Vijayendra Bhamidipati 4 0 venkata swamybabu budumuru 1 1 Wido den Hollander4 0 Grand Total511 267 Thanks /sudha
RE: [DISCUSS] KVM HA
-Original Message- From: Marcus Sorensen [mailto:shadow...@gmail.com] Sent: Wednesday, August 07, 2013 1:42 PM To: dev@cloudstack.apache.org Subject: Re: [DISCUSS] KVM HA Does KVMInvestigator work on all shared primary storage, or just NFS? Right now, the agent only writes heartbeat into NFS primary storage, so the current implementation only works with NFS primary storage. But the heartbeat file is just only one implementation, if using IPMI, it should be able to detect host status also. I'm only familiar with the NFS KVMHA directories. From this it seems like a clean stop of the KVM agent still shouldn't trigger any issues/HA, correct? Yes, a clean stop wont' trigger HA. reboot -f or plug off power supply will trigger HA. On Wed, Aug 7, 2013 at 2:28 PM, Edison Su edison...@citrix.com wrote: There is long time issue related to KVM HA, see bug: CLOUDSTACK-3535. Basically, HA won't be triggered, if KVM agent is stopped either normally nor abnormally, HA only be triggered if the network between mgt server and kvm host is disconnected and the network between KVM hosts in the same cluster is disconnected. Here is how the KVM HA works after the fix for CLOUDSTACK-3535: 1. If agent is stopped, agent will send a shutdown request to mgt server, mgt server will mark the host as disconnected, while still maintain the host in pingmap. Code is in AgentManagerImpl-AgentHandler- ProcessRequest- disconnectWithoutInvestigation 2. After ping.interval, mgt server will find the host is ping timeout, then start HA investigation for the host. Code is in AgentMonitor-run- disconnectWithInvestigation 3. Mgt server will call all the available Investigators to investigate the status of host. The current investigators will be called for KVM host: UserVmDomRInvestigator-isAgentAlive, will send PingTestCommand to the host's neighbor. PingTestCommand will ping host's private ip address, if ping is reachable, means host is up, otherwise, host's state is unknown. So this investigator can only detect host is in up state. KVMInvestigator, which is newly added, will send a CheckOnHostCommand to host's neighbor. CheckOnHostCommand will check the heartbeat of host(heartbeat is stored on shared primary storage). Ideally, it will detect host is down or up. Combined with UserVmDomRInvestigator and KVMInvestigator, mgt server should find out the status of host. But there is case, these two investigators can report wrong status of host: Host is in a network partition, while the KVM agent is down(thus heartbeat is stopped) 4. After investigator reports status of host, if host is down, then start HA for VMs created on the host. Improvement: Per suggestion from Lennert den Teuling, we'd better use IPMI to detect host status, which is more reliable than ping and heartbeat, as IPMI has its own network, less likely has network partition.
[DISCUSS] Binary diff and hot patch
I was wondering if we can use binary diffs to hotpatch ACS. Currently, the way for me to identify what patch could have caused issue X, would be - try to find the suspects - it could be several. Revert 1 commit, rebuild and retry, if failed, restore previously reverted commit, revert 2nd commit, rebuild and retry - etc.. As you can see, this is very time consuming and painful, I'm wondering if there is a better way to do this. I came across a project called javaxdelta on sourceforge - curious what your thought on this may be, and if there is a better way to try out patches instead of what I'm doing now. Thanks ilya
RE: [DISCUSS] KVM HA
-Original Message- From: Marcus Sorensen [mailto:shadow...@gmail.com] Sent: Wednesday, August 07, 2013 1:46 PM To: dev@cloudstack.apache.org Subject: Re: [DISCUSS] KVM HA I'm not sure we can rely on IPMI to tell us much about the host status itself. It's easy to use it for checking on basic poweron/poweroff, temperature, etc, but not so easy to tell if something is wrong with the OS, config, or at the software level. Check the host status is enough, the investigator only needs to know, the host is power on or off. If host is power off, then cloudstack can start VM on another host, without worried about the split brain issue. However, I did mention support in that thread early on for sending an IPMI poweroff for hosts that cloudstack has determined are down and starts migrating vms for, as a safety precaution. On Wed, Aug 7, 2013 at 2:41 PM, Marcus Sorensen shadow...@gmail.com wrote: Does KVMInvestigator work on all shared primary storage, or just NFS? I'm only familiar with the NFS KVMHA directories. From this it seems like a clean stop of the KVM agent still shouldn't trigger any issues/HA, correct? On Wed, Aug 7, 2013 at 2:28 PM, Edison Su edison...@citrix.com wrote: There is long time issue related to KVM HA, see bug: CLOUDSTACK-3535. Basically, HA won't be triggered, if KVM agent is stopped either normally nor abnormally, HA only be triggered if the network between mgt server and kvm host is disconnected and the network between KVM hosts in the same cluster is disconnected. Here is how the KVM HA works after the fix for CLOUDSTACK-3535: 1. If agent is stopped, agent will send a shutdown request to mgt server, mgt server will mark the host as disconnected, while still maintain the host in pingmap. Code is in AgentManagerImpl-AgentHandler- ProcessRequest- disconnectWithoutInvestigation 2. After ping.interval, mgt server will find the host is ping timeout, then start HA investigation for the host. Code is in AgentMonitor-run- disconnectWithInvestigation 3. Mgt server will call all the available Investigators to investigate the status of host. The current investigators will be called for KVM host: UserVmDomRInvestigator-isAgentAlive, will send PingTestCommand to the host's neighbor. PingTestCommand will ping host's private ip address, if ping is reachable, means host is up, otherwise, host's state is unknown. So this investigator can only detect host is in up state. KVMInvestigator, which is newly added, will send a CheckOnHostCommand to host's neighbor. CheckOnHostCommand will check the heartbeat of host(heartbeat is stored on shared primary storage). Ideally, it will detect host is down or up. Combined with UserVmDomRInvestigator and KVMInvestigator, mgt server should find out the status of host. But there is case, these two investigators can report wrong status of host: Host is in a network partition, while the KVM agent is down(thus heartbeat is stopped) 4. After investigator reports status of host, if host is down, then start HA for VMs created on the host. Improvement: Per suggestion from Lennert den Teuling, we'd better use IPMI to detect host status, which is more reliable than ping and heartbeat, as IPMI has its own network, less likely has network partition.
Re: Review Request 13298: Put in upgrade path from 4.1.1 to 4.1.2
Let's get this applied! 4.1 branch has been broken for over a week! On Wed, Aug 7, 2013 at 11:48 AM, Alena Prokharchyk alena.prokharc...@citrix.com wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13298/#review24813 --- Vijay, as the schema is empty, you don't have to have this file. Remove it pls, as well as the reference to it from the corresponding Java code. - Alena Prokharchyk On Aug. 5, 2013, 9:12 p.m., Venkata Siva Vijayendra Bhamidipati wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13298/ --- (Updated Aug. 5, 2013, 9:12 p.m.) Review request for cloudstack, Alena Prokharchyk, Chip Childers, and Min Chen. Bugs: CLOUDSTACK-4091 Repository: cloudstack-git Description --- Putting in upgrade path for db upgrade from 4.1.1 to 4.1.2. Prior to the patch, th mgmt server doesn't start up successfully on 4.1. Diffs - server/src/com/cloud/upgrade/DatabaseUpgradeChecker.java a7a4e61 server/src/com/cloud/upgrade/dao/Upgrade411to412.java PRE-CREATION setup/db/db/schema-411to412.sql PRE-CREATION Diff: https://reviews.apache.org/r/13298/diff/ Testing --- Mgmt server comes up successfully upon fresh install of 4.1, and the db version shows the right value: mysql select * from version; ++-+-+--+ | id | version | updated | step | ++-+-+--+ | 1 | 4.0.0 | 2013-08-05 07:40:38 | Complete | | 2 | 4.1.0 | 2013-08-05 14:42:20 | Complete | | 3 | 4.1.1 | 2013-08-05 14:42:20 | Complete | | 4 | 4.1.2 | 2013-08-05 14:42:20 | Complete | ++-+-+--+ 4 rows in set (0.00 sec) mysql Thanks, Venkata Siva Vijayendra Bhamidipati
RE: Review Request 13298: Put in upgrade path from 4.1.1 to 4.1.2
Will send out a patch shortly. Regards, Vijay -Original Message- From: Marcus Sorensen [mailto:shadow...@gmail.com] Sent: Wednesday, August 07, 2013 3:09 PM To: dev@cloudstack.apache.org; Alena Prokharchyk Cc: Min Chen; Chip Childers; cloudstack; Vijayendra Bhamidipati Subject: Re: Review Request 13298: Put in upgrade path from 4.1.1 to 4.1.2 Let's get this applied! 4.1 branch has been broken for over a week! On Wed, Aug 7, 2013 at 11:48 AM, Alena Prokharchyk alena.prokharc...@citrix.com wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13298/#review24813 --- Vijay, as the schema is empty, you don't have to have this file. Remove it pls, as well as the reference to it from the corresponding Java code. - Alena Prokharchyk On Aug. 5, 2013, 9:12 p.m., Venkata Siva Vijayendra Bhamidipati wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13298/ --- (Updated Aug. 5, 2013, 9:12 p.m.) Review request for cloudstack, Alena Prokharchyk, Chip Childers, and Min Chen. Bugs: CLOUDSTACK-4091 Repository: cloudstack-git Description --- Putting in upgrade path for db upgrade from 4.1.1 to 4.1.2. Prior to the patch, th mgmt server doesn't start up successfully on 4.1. Diffs - server/src/com/cloud/upgrade/DatabaseUpgradeChecker.java a7a4e61 server/src/com/cloud/upgrade/dao/Upgrade411to412.java PRE-CREATION setup/db/db/schema-411to412.sql PRE-CREATION Diff: https://reviews.apache.org/r/13298/diff/ Testing --- Mgmt server comes up successfully upon fresh install of 4.1, and the db version shows the right value: mysql select * from version; ++-+-+--+ | id | version | updated | step | ++-+-+--+ | 1 | 4.0.0 | 2013-08-05 07:40:38 | Complete | | 2 | 4.1.0 | 2013-08-05 14:42:20 | Complete | | 3 | 4.1.1 | 2013-08-05 14:42:20 | Complete | | 4 | 4.1.2 | 2013-08-05 14:42:20 | Complete | ++-+-+--+ 4 rows in set (0.00 sec) mysql Thanks, Venkata Siva Vijayendra Bhamidipati
RE: Question about creating a template
Hi Mike, The problem is with handling the command org.apache.cloudstack.storage.command.CopyCommand, look at the log before this error and see if there is any problem with the registered agent that is suppose to handle this command. -Soheil From: Mike Tutkowski [mike.tutkow...@solidfire.com] Sent: Tuesday, August 06, 2013 11:28 AM To: dev@cloudstack.apache.org Subject: Question about creating a template Hi, I have a small Linux ISO (http://www.damnsmalllinux.org/download.html) I've uploaded to my secondary storage NFS location. I can create VM instances just fine with it. I stopped one of my VMs that was created with this ISO and tried to create a template from its root disk. I received the following exception and was wondering what people's comments might be. Thanks! ERROR [cloud.async.AsyncJobManagerImpl] (Job-Executor-8:job-41 = [ 583f08f2-e32a-4d16-815a-2ac2953d2ae9 ]) Unexpected exception while executing org.apache.cloudstack.api.command.user.template.CreateTemplateCmd com.cloud.utils.exception.CloudRuntimeException: Failed to create templateUnsupported command issued:org.apache.cloudstack.storage.command.CopyCommand. Are you sure you got the right type of server? at com.cloud.template.TemplateManagerImpl.createPrivateTemplate(TemplateManagerImpl.java:1368) at com.cloud.utils.component.ComponentInstantiationPostProcessor$InterceptorDispatcher.intercept(ComponentInstantiationPostProcessor.java:125) at org.apache.cloudstack.api.command.user.template.CreateTemplateCmd.execute(CreateTemplateCmd.java:263) at com.cloud.api.ApiDispatcher.dispatch(ApiDispatcher.java:158) at com.cloud.async.AsyncJobManagerImpl$1.run(AsyncJobManagerImpl.java:531) at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:439) at java.util.concurrent.FutureTask$Sync.innerRun(FutureTask.java:303) at java.util.concurrent.FutureTask.run(FutureTask.java:138) at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:895) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:918) at java.lang.Thread.run(Thread.java:680) -- *Mike Tutkowski* *Senior CloudStack Developer, SolidFire Inc.* e: mike.tutkow...@solidfire.com o: 303.746.7302 Advancing the way the world uses the cloudhttp://solidfire.com/solution/overview/?video=play *™*
Re: Review Request 13373: keep the existing changes in cloud-set-guest-sshkey.in to make sure we do not break the backward compatibility.
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13373/#review24838 --- Look fine to me. Make sure when you test it, you tested both way(with data-center or not). - Sheng Yang On Aug. 7, 2013, 1:25 p.m., bharat kumar wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13373/ --- (Updated Aug. 7, 2013, 1:25 p.m.) Review request for cloudstack, Jayapal Reddy and Sheng Yang. Bugs: Cloudstack-4138 Repository: cloudstack-git Description --- keep the existing changes in cloud-set-guest-sshkey.in to make sure we do not break the backward compatibility. Diffs - setup/bindir/cloud-set-guest-sshkey.in 9436b03 Diff: https://reviews.apache.org/r/13373/diff/ Testing --- Tested on 4.2 Thanks, bharat kumar
Re: Review Request 13298: Put in upgrade path from 4.1.1 to 4.1.2
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13298/#review24840 --- Pushed the patch to 4.1 branch with commit id ae8159b8560181f5935443876862f8b826bc0a31. Not closing the ticket as another patch has to be submitted to master branch. - Alena Prokharchyk On Aug. 7, 2013, 10:44 p.m., Venkata Siva Vijayendra Bhamidipati wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13298/ --- (Updated Aug. 7, 2013, 10:44 p.m.) Review request for cloudstack, Alena Prokharchyk, Chip Childers, and Min Chen. Bugs: CLOUDSTACK-4091 Repository: cloudstack-git Description --- Putting in upgrade path for db upgrade from 4.1.1 to 4.1.2. Prior to the patch, th mgmt server doesn't start up successfully on 4.1. Diffs - server/src/com/cloud/upgrade/DatabaseUpgradeChecker.java a7a4e61 server/src/com/cloud/upgrade/dao/Upgrade411to412.java PRE-CREATION Diff: https://reviews.apache.org/r/13298/diff/ Testing --- Mgmt server comes up successfully upon fresh install of 4.1, and the db version shows the right value: mysql select * from version; ++-+-+--+ | id | version | updated | step | ++-+-+--+ | 1 | 4.0.0 | 2013-08-05 07:40:38 | Complete | | 2 | 4.1.0 | 2013-08-05 14:42:20 | Complete | | 3 | 4.1.1 | 2013-08-05 14:42:20 | Complete | | 4 | 4.1.2 | 2013-08-05 14:42:20 | Complete | ++-+-+--+ 4 rows in set (0.00 sec) mysql Thanks, Venkata Siva Vijayendra Bhamidipati
Re: Review Request 13298: Put in upgrade path from 4.1.1 to 4.1.2
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13298/#review24841 --- Ship it! Ship It! - Alena Prokharchyk On Aug. 7, 2013, 10:44 p.m., Venkata Siva Vijayendra Bhamidipati wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13298/ --- (Updated Aug. 7, 2013, 10:44 p.m.) Review request for cloudstack, Alena Prokharchyk, Chip Childers, and Min Chen. Bugs: CLOUDSTACK-4091 Repository: cloudstack-git Description --- Putting in upgrade path for db upgrade from 4.1.1 to 4.1.2. Prior to the patch, th mgmt server doesn't start up successfully on 4.1. Diffs - server/src/com/cloud/upgrade/DatabaseUpgradeChecker.java a7a4e61 server/src/com/cloud/upgrade/dao/Upgrade411to412.java PRE-CREATION Diff: https://reviews.apache.org/r/13298/diff/ Testing --- Mgmt server comes up successfully upon fresh install of 4.1, and the db version shows the right value: mysql select * from version; ++-+-+--+ | id | version | updated | step | ++-+-+--+ | 1 | 4.0.0 | 2013-08-05 07:40:38 | Complete | | 2 | 4.1.0 | 2013-08-05 14:42:20 | Complete | | 3 | 4.1.1 | 2013-08-05 14:42:20 | Complete | | 4 | 4.1.2 | 2013-08-05 14:42:20 | Complete | ++-+-+--+ 4 rows in set (0.00 sec) mysql Thanks, Venkata Siva Vijayendra Bhamidipati
Re: Review Request 13298: Put in upgrade path from 4.1.1 to 4.1.2
On Aug. 7, 2013, 10:58 p.m., Alena Prokharchyk wrote: Ship It! Vijay will open a separate reviewboard request for 411-412 upgrade path for master branch. The patch is going to be different given the presence of 4.2 release. - Alena --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13298/#review24841 --- On Aug. 7, 2013, 10:44 p.m., Venkata Siva Vijayendra Bhamidipati wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13298/ --- (Updated Aug. 7, 2013, 10:44 p.m.) Review request for cloudstack, Alena Prokharchyk, Chip Childers, and Min Chen. Bugs: CLOUDSTACK-4091 Repository: cloudstack-git Description --- Putting in upgrade path for db upgrade from 4.1.1 to 4.1.2. Prior to the patch, th mgmt server doesn't start up successfully on 4.1. Diffs - server/src/com/cloud/upgrade/DatabaseUpgradeChecker.java a7a4e61 server/src/com/cloud/upgrade/dao/Upgrade411to412.java PRE-CREATION Diff: https://reviews.apache.org/r/13298/diff/ Testing --- Mgmt server comes up successfully upon fresh install of 4.1, and the db version shows the right value: mysql select * from version; ++-+-+--+ | id | version | updated | step | ++-+-+--+ | 1 | 4.0.0 | 2013-08-05 07:40:38 | Complete | | 2 | 4.1.0 | 2013-08-05 14:42:20 | Complete | | 3 | 4.1.1 | 2013-08-05 14:42:20 | Complete | | 4 | 4.1.2 | 2013-08-05 14:42:20 | Complete | ++-+-+--+ 4 rows in set (0.00 sec) mysql Thanks, Venkata Siva Vijayendra Bhamidipati
Re: [jira] [Commented] (CLOUDSTACK-2729) [Automation] Libvirt failed find primary storage and VM deployment failed
Here's the issue... local storage isn't mounted like remote storage is. This is invalid for local storage: String lockFile = _mountPoint + File.separator + pool.getUUIDString() + File.separator + _lockfile; Need to look at storage pool type and change lockfile to path for local storage On Wed, Aug 7, 2013 at 5:18 PM, Marcus Sorensen (JIRA) j...@apache.org wrote: [ https://issues.apache.org/jira/browse/CLOUDSTACK-2729?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13732906#comment-13732906 ] Marcus Sorensen commented on CLOUDSTACK-2729: - Sorry, I think the KVMFILELOCK is left over on secondary storage because it's unable to lock the primary storage, and doesn't clean up after itself. I'm not sure why it can't take the lock out on primary yet... troubleshooting that. [Automation] Libvirt failed find primary storage and VM deployment failed Key: CLOUDSTACK-2729 URL: https://issues.apache.org/jira/browse/CLOUDSTACK-2729 Project: CloudStack Issue Type: Bug Security Level: Public(Anyone can view this level - this is the default.) Components: Storage Controller Affects Versions: 4.2.0 Environment: Master branch build Automation environment - KVM Reporter: Rayees Namathponnan Assignee: edison su Priority: Blocker Fix For: 4.2.0 Attachments: CLOUDSTACK-2729.rar Failed to deploy VM in automation, environment created with , 2 hosts and 2 primary storages in a cluster 1) /export/home/rayees/SC_QA_AUTO4/primary2 2) /export/home/rayees/SC_QA_AUTO4/primary Libvirt failed to find the primary storage /export/home/rayees/SC_QA_AUTO4/primary and VM deployment failed MS log 2013-05-28 21:11:44,540 DEBUG [agent.transport.Request] (consoleproxy-1:null) Seq 4-936706756: Sending { Cmd , MgmtId: 29066118877352, via: 4, Ver: v1, Flags: 100111, [{StartCommand:{vm:{id:29,name:v-29-QA,type:ConsoleProxy,cpus:1,minSpeed:500,maxSpeed:500,minRam:1073741824,maxRam:1073741824,arch:x86_64,os:Debian GNU/Linux 5.0 (32-bit),bootArgs: template=domP type=consoleproxy host=10.223.49.195 port=8250 name=v-29-QA premium=true zone=1 pod=1 guid=Proxy.29 proxy_vm=29 disable_rp_filter=true eth2ip=10.223.122.73 eth2mask=255.255.255.192 gateway=10.223.122.65 eth0ip=169.254.0.154 eth0mask=255.255.0.0 eth1ip=10.223.50.96 eth1mask=255.255.255.192 mgmtcidr=10.223.49.192/26 localgw=10.223.50.65 internaldns1=10.223.110.254 dns1=72.52.126.11,rebootOnCrash:false,enableHA:false,limitCpuUse:false,vncPassword:d099568827911cef,params:{},uuid:5a146833-6a8c-44e5-83c0-50f34accf513,disks:[{id:32,name:ROOT-29,mountPoint:/export/home/rayees/SC_QA_AUTO4/primary,path:f6f8d865-e9c0-4188-8a33-6c6383ca5075,size:276406784,type:ROOT,storagePoolType:NetworkFilesystem,storagePoolUuid:fff90cb5-06dd-33b3-8815-d78c08ca01d9,deviceId:0}],nics:[{deviceId:2,networkRateMbps:-1,defaultNic:true,uuid:21175978-96bd-4160-8228-8ad15aa40c66,ip:10.223.122.73,netmask:255.255.255.192,gateway:10.223.122.65,mac:06:2e:5a:00:00:42,dns1:72.52.126.11,broadcastType:Vlan,type:Public,broadcastUri:vlan://1221,isolationUri:vlan://1221,isSecurityGroupEnabled:false},{deviceId:0,networkRateMbps:-1,defaultNic:false,uuid:e0651452-a76e-4564-96b2-3d5d51e9bcd6,ip:169.254.0.154,netmask:255.255.0.0,gateway:169.254.0.1,mac:0e:00:a9:fe:00:9a,broadcastType:LinkLocal,type:Control,isSecurityGroupEnabled:false},{deviceId:1,networkRateMbps:-1,defaultNic:false,uuid:0765c229-468e-4dfd-8382-24ac49791a8d,ip:10.223.50.96,netmask:255.255.255.192,gateway:10.223.50.65,mac:06:a8:e8:00:00:1d,broadcastType:Native,type:Management,isSecurityGroupEnabled:false}]},hostIp:10.223.50.66,wait:0}},{check.CheckSshCommand:{ip:169.254.0.154,port:3922,interval:6,retries:100,name:v-29-QA,wait:0}}] } 2013-05-28 21:11:44,552 DEBUG [agent.transport.Request] (AgentManager-Handler-1:null) Seq 4-936706756: Processing: { Ans: , MgmtId: 29066118877352, via: 4, Ver: v1, Flags: 110, [{Answer:{result:false,details:com.cloud.utils.exception.CloudRuntimeException: org.libvirt.LibvirtException: Storage pool not found: no pool with matching uuid\n\tat com.cloud.hypervisor.kvm.storage.LibvirtStorageAdaptor.getStoragePool(LibvirtStorageAdaptor.java:380)\n\tat com.cloud.hypervisor.kvm.storage.KVMStoragePoolManager.getStoragePool(KVMStoragePoolManager.java:72)\n\tat com.cloud.hypervisor.kvm.resource.LibvirtComputingResource.createVbd(LibvirtComputingResource.java:3399)\n\tat com.cloud.hypervisor.kvm.resource.LibvirtComputingResource.execute(LibvirtComputingResource.java:3293)\n\tat com.cloud.hypervisor.kvm.resource.LibvirtComputingResource.executeRequest(LibvirtComputingResource.java:1187)\n\tat com.cloud.agent.Agent.processRequest(Agent.java:525)\n\tat
Re: [jira] [Commented] (CLOUDSTACK-2729) [Automation] Libvirt failed find primary storage and VM deployment failed
Here's the debugging I put in: 2013-08-07 17:39:12,286 DEBUG [kvm.storage.LibvirtStorageAdaptor] (agentRequest-Handler-3:null) attempting to lock pool 2a9c418e-dd56-3f2c-9081-032b877d5644 with file '/mnt/2a9c418e-dd56-3f2c-9081-032b877d5644/KVMFILELOCK/.lock' 2013-08-07 17:39:12,425 DEBUG [kvm.storage.LibvirtStorageAdaptor] (agentRequest-Handler-3:null) attempting to lock pool 2fa10069-1fb3-4eea-8dc4-f773e7be0d19 with file '/mnt/2fa10069-1fb3-4eea-8dc4-f773e7be0d19/KVMFILELOCK/.lock' 2013-08-07 17:39:12,479 DEBUG [cloud.agent.Agent] (agentRequest-Handler-3:null) Seq 1-1276706825: { Ans: , MgmtId: 90520735907663, via: 1, Ver: v1, Flags: 110, [{storage.PrimaryStorageDownloadAnswer:{templateSize:0,result:false,details:com.cloud.utils.exception.CloudRuntimeException: Can not get file lock to refresh the pool 2fa10069-1fb3-4eea-8dc4-f773e7be0d19,wait:0}}] } 2013-08-07 17:39:12,536 DEBUG [cloud.agent.Agent] (agentRequest-Handler-4:null) Request:Seq 1-1276706826: { Cmd , MgmtId: 90520735907663, via: 1, Ver: v1, Flags: 100111, [{StopCommand:{isProxy:false,vmName:s-57-VM,wait:0}}] } Note that 2a9c418e-dd56-3f2c-9081-032b877d5644 is my secondary storage, 2fa10069-1fb3-4eea-8dc4-f773e7be0d19 is local storage in /var/lib/libvirt/images On Wed, Aug 7, 2013 at 5:41 PM, Marcus Sorensen shadow...@gmail.com wrote: Here's the issue... local storage isn't mounted like remote storage is. This is invalid for local storage: String lockFile = _mountPoint + File.separator + pool.getUUIDString() + File.separator + _lockfile; Need to look at storage pool type and change lockfile to path for local storage On Wed, Aug 7, 2013 at 5:18 PM, Marcus Sorensen (JIRA) j...@apache.org wrote: [ https://issues.apache.org/jira/browse/CLOUDSTACK-2729?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13732906#comment-13732906 ] Marcus Sorensen commented on CLOUDSTACK-2729: - Sorry, I think the KVMFILELOCK is left over on secondary storage because it's unable to lock the primary storage, and doesn't clean up after itself. I'm not sure why it can't take the lock out on primary yet... troubleshooting that. [Automation] Libvirt failed find primary storage and VM deployment failed Key: CLOUDSTACK-2729 URL: https://issues.apache.org/jira/browse/CLOUDSTACK-2729 Project: CloudStack Issue Type: Bug Security Level: Public(Anyone can view this level - this is the default.) Components: Storage Controller Affects Versions: 4.2.0 Environment: Master branch build Automation environment - KVM Reporter: Rayees Namathponnan Assignee: edison su Priority: Blocker Fix For: 4.2.0 Attachments: CLOUDSTACK-2729.rar Failed to deploy VM in automation, environment created with , 2 hosts and 2 primary storages in a cluster 1) /export/home/rayees/SC_QA_AUTO4/primary2 2) /export/home/rayees/SC_QA_AUTO4/primary Libvirt failed to find the primary storage /export/home/rayees/SC_QA_AUTO4/primary and VM deployment failed MS log 2013-05-28 21:11:44,540 DEBUG [agent.transport.Request] (consoleproxy-1:null) Seq 4-936706756: Sending { Cmd , MgmtId: 29066118877352, via: 4, Ver: v1, Flags: 100111, [{StartCommand:{vm:{id:29,name:v-29-QA,type:ConsoleProxy,cpus:1,minSpeed:500,maxSpeed:500,minRam:1073741824,maxRam:1073741824,arch:x86_64,os:Debian GNU/Linux 5.0 (32-bit),bootArgs: template=domP type=consoleproxy host=10.223.49.195 port=8250 name=v-29-QA premium=true zone=1 pod=1 guid=Proxy.29 proxy_vm=29 disable_rp_filter=true eth2ip=10.223.122.73 eth2mask=255.255.255.192 gateway=10.223.122.65 eth0ip=169.254.0.154 eth0mask=255.255.0.0 eth1ip=10.223.50.96 eth1mask=255.255.255.192 mgmtcidr=10.223.49.192/26 localgw=10.223.50.65 internaldns1=10.223.110.254
RE: [jira] [Commented] (CLOUDSTACK-2729) [Automation] Libvirt failed find primary storage and VM deployment failed
The local path of storage pool should be able to get from LibvirtStoragePoolDef. -Original Message- From: Marcus Sorensen [mailto:shadow...@gmail.com] Sent: Wednesday, August 07, 2013 4:43 PM To: dev@cloudstack.apache.org Cc: cloudstack-iss...@incubator.apache.org Subject: Re: [jira] [Commented] (CLOUDSTACK-2729) [Automation] Libvirt failed find primary storage and VM deployment failed Here's the debugging I put in: 2013-08-07 17:39:12,286 DEBUG [kvm.storage.LibvirtStorageAdaptor] (agentRequest-Handler-3:null) attempting to lock pool 2a9c418e-dd56-3f2c-9081-032b877d5644 with file '/mnt/2a9c418e-dd56-3f2c- 9081-032b877d5644/KVMFILELOCK/.lock' 2013-08-07 17:39:12,425 DEBUG [kvm.storage.LibvirtStorageAdaptor] (agentRequest-Handler-3:null) attempting to lock pool 2fa10069-1fb3-4eea-8dc4-f773e7be0d19 with file '/mnt/2fa10069-1fb3-4eea- 8dc4-f773e7be0d19/KVMFILELOCK/.lock' 2013-08-07 17:39:12,479 DEBUG [cloud.agent.Agent] (agentRequest-Handler-3:null) Seq 1-1276706825: { Ans: , MgmtId: 90520735907663, via: 1, Ver: v1, Flags: 110, [{storage.PrimaryStorageDownloadAnswer:{templateSize:0,result:fals e,details:com.cloud.utils.exception.CloudRuntimeException: Can not get file lock to refresh the pool 2fa10069-1fb3-4eea-8dc4- f773e7be0d19,wait:0}}] } 2013-08-07 17:39:12,536 DEBUG [cloud.agent.Agent] (agentRequest-Handler-4:null) Request:Seq 1-1276706826: { Cmd , MgmtId: 90520735907663, via: 1, Ver: v1, Flags: 100111, [{StopCommand:{isProxy:false,vmName:s-57-VM,wait:0}}] } Note that 2a9c418e-dd56-3f2c-9081-032b877d5644 is my secondary storage, 2fa10069-1fb3-4eea-8dc4-f773e7be0d19 is local storage in /var/lib/libvirt/images On Wed, Aug 7, 2013 at 5:41 PM, Marcus Sorensen shadow...@gmail.com wrote: Here's the issue... local storage isn't mounted like remote storage is. This is invalid for local storage: String lockFile = _mountPoint + File.separator + pool.getUUIDString() + File.separator + _lockfile; Need to look at storage pool type and change lockfile to path for local storage On Wed, Aug 7, 2013 at 5:18 PM, Marcus Sorensen (JIRA) j...@apache.org wrote: [ https://issues.apache.org/jira/browse/CLOUDSTACK- 2729?page=com.atlass ian.jira.plugin.system.issuetabpanels:comment- tabpanelfocusedComment Id=13732906#comment-13732906 ] Marcus Sorensen commented on CLOUDSTACK-2729: - Sorry, I think the KVMFILELOCK is left over on secondary storage because it's unable to lock the primary storage, and doesn't clean up after itself. I'm not sure why it can't take the lock out on primary yet... troubleshooting that. [Automation] Libvirt failed find primary storage and VM deployment failed Key: CLOUDSTACK-2729 URL: https://issues.apache.org/jira/browse/CLOUDSTACK-2729 Project: CloudStack Issue Type: Bug Security Level: Public(Anyone can view this level - this is the default.) Components: Storage Controller Affects Versions: 4.2.0 Environment: Master branch build Automation environment - KVM Reporter: Rayees Namathponnan Assignee: edison su Priority: Blocker Fix For: 4.2.0 Attachments: CLOUDSTACK-2729.rar Failed to deploy VM in automation, environment created with , 2 hosts and 2 primary storages in a cluster 1) /export/home/rayees/SC_QA_AUTO4/primary2 2) /export/home/rayees/SC_QA_AUTO4/primary Libvirt failed to find the primary storage /export/home/rayees/SC_QA_AUTO4/primary and VM deployment failed MS log 2013-05-28 21:11:44,540 DEBUG [agent.transport.Request] (consoleproxy-1:null) Seq 4-936706756: Sending { Cmd , MgmtId: 29066118877352, via: 4, Ver: v1, Flags: 100111, [{StartCommand:{vm:{id:29,name:v-29- QA,type:ConsoleProx y,cpus:1,minSpeed:500,maxSpeed:500,minRam:1073741824,maxRa m:1073741824,arch:x86_64,os:Debian GNU/Linux 5.0 (32-bit),bootArgs: template=domP type=consoleproxy host=10.223.49.195 port=8250 name=v-29-QA premium=true zone=1 pod=1 guid=Proxy.29 proxy_vm=29 disable_rp_filter=true eth2ip=10.223.122.73 eth2mask=255.255.255.192 gateway=10.223.122.65 eth0ip=169.254.0.154 eth0mask=255.255.0.0 eth1ip=10.223.50.96 eth1mask=255.255.255.192 mgmtcidr=10.223.49.192/26 localgw=10.223.50.65 internaldns1=10.223.110.254 dns1=72.52.126.11,rebootOnCrash:false,enableHA:false,limitCpuU se:false,vncPassword:d099568827911cef,params:{},uuid:5a146 833-6a8c-44e5-83c0-50f34accf513,disks:[{id:32,name:ROOT-29, mountPoint:/export/home/rayees/SC_QA_AUTO4/primary,path:f6f8 d 865-e9c0-4188-8a33-6c6383ca5075,size:276406784,type:ROOT,sto ragePoolType:NetworkFilesystem,storagePoolUuid:fff90cb5-06dd- 3 3b3-8815-d78c08ca01d9,deviceId:0}],nics:[{deviceId:2,network
Re: [ACS42]Attach volumes to Instance which has VM snapshots
It is currently a limitation. After VM is running in VM-snapshot mode, it is not allowed to change VM's disk membership status Kelven On 8/7/13 10:09 AM, Sailaja Mada sailaja.m...@citrix.com wrote: Hi, I have an instance which has VM snapshots. (VMWARE hypervisor) Now I am trying to attach DATA volume to this instance. It failed saying Unable to attach volume, please specify a VM that does not have VM snapshots Is there any limitation not to add additional volumes if it has VM snapshots? Thanks, Sailaja.M
Review Board messages
Hi, I noticed that Review Board messages are still going to cloudstack-...@incubator.apache.org. Should we switch that to dev@cloudstack.apache.org ? If so, who has the access to do so? Thanks, Dave.