Re: [one-users] "Wrong state to perform action" when onevm livemigrate'ing a RUNNING machine ?
Hi, Thanks Carlos, that did the trick. Yours, Steffen Just for the records: update history set body = '5node-b...3' where vid=1006 and seq=5; and of course with the correct values for vid and seq determined before, and the body XML with REASON edited. On Tue, 2012-09-25 at 18:47 +0200, Carlos Martín Sánchez wrote: > Hi Steffen, > > > Looks like you are right, a migration failed and somehow left the > history records corrupted. > > > If you only need to be able to migrate, change the reason of history 7 > to '3'. To do so, stop opennebula, do a DB backup, > and update the body column of the history table. As I said, do this > for history with SEQ 7. > > > The history seq. 8 is completely uninitialized, all the timestamps are > 0, and that will show strangely in the accounting. But if you just > need to migrate, changing the history 7 reason should be enough. > > > Regards > -- > Carlos Martín, MSc > Project Engineer > OpenNebula - The Open-source Solution for Data Center Virtualization > www.OpenNebula.org | cmar...@opennebula.org | @OpenNebula > > > > On Tue, Sep 25, 2012 at 10:05 AM, Steffen Neumann > wrote: > Hi, > > sorry for replying to self. I found a thread from august this > year, > http://comments.gmane.org/gmane.comp.distributed.opennebula.user/7554 > which I missed in my earlier searches, where Shantanu Pavgi > suggests > to clean the history and vm_pool tables. > > To avoid messing this up, which entries in history do I sql > delete ? > SEQHOSTNAME REASON STARTTIME > PTIME >0 node-blc28 user 08/15 13:07:13 40 02:47:35 00 > 00:00:15 >1 node-cc08d5 none 09/24 15:54:48 00 17:40:19 00 > 00:00:00 >2 node-blc28 user 09/24 15:54:48 00 00:01:22 00 > 00:00:00 >3 node-cc08d5 none 09/24 15:56:10 00 17:38:57 00 > 00:00:00 >4 node-blc28 user 09/24 15:56:10 00 00:20:27 00 > 00:00:00 >5 node-blc07 none 09/24 16:16:37 00 17:18:30 00 > 00:00:00 >6 node-blc28 user 09/24 16:16:37 00 00:28:09 00 > 00:00:00 >7 node-blc02 none 09/24 16:44:46 00 16:50:21 00 > 00:00:00 >8 node-blc28 none - 268 07:35:0 00 > 00:00:00 > > > So do I delete just SEQ 7 ? > > What do I replace in VM_POOL ? > Here the in column body seems to list only > 8 ? > > Yours, > Steffen > > > 100400oneadminoneadminone-100413485594223313450288060one-1004335544324312015720310559760171 > > > > > > 8node-blc28/cumulus/one/var/3700vmm_kvmtm_shared000 > > > > > On Tue, 2012-09-25 at 09:44 +0200, Steffen Neumann wrote: > > Hi, > > > > I have a working one-3.0 cluster, and for upgrading to 3.6 > > I need to move around VMs. The cluster has been up and > > running for about a year now. > > > > At least two VMs on one host are stubbornly refusing > > to live-migrate: > > > > oneadmin@cumulus:~$ onevm livemigrate 1004 55 > > [VirtualMachineDeploy] Wrong state to perform action > > > > As the machine is RUNNING, I can ssh into it, > > I am a bit at a loss here. Only strange thing is > > that current VM monitoring reports "Net_TX: -1" > > which happens since August 17th. > > > > Any idea ? > > > > sneumann: can you send onevm show of that machine > to the mailing list? > > sneumann: there could be something wrong with the > history and my pals may know what it is > > > > onehost list: > > ID NAME RVM TCPU FCPU ACPU TMEM > FMEM AMEM STAT > > 55 node-blc02 3 2400 2393 0 94.6G > 71.6G 46.6G on > > 37 node-blc28 2 2400766800 94.6G > 48.7G 30.6G on > > ... > > > > onevm show 1004: > > VIRTUAL MACHINE 1004 INFORMATION > > ID : 1004 > > NAME: one-1004 > > USER
Re: [one-users] "Wrong state to perform action" when onevm livemigrate'ing a RUNNING machine ?
Hi, sorry for replying to self. I found a thread from august this year, http://comments.gmane.org/gmane.comp.distributed.opennebula.user/7554 which I missed in my earlier searches, where Shantanu Pavgi suggests to clean the history and vm_pool tables. To avoid messing this up, which entries in history do I sql delete ? SEQHOSTNAME REASON STARTTIME PTIME 0 node-blc28 user 08/15 13:07:13 40 02:47:35 00 00:00:15 1 node-cc08d5 none 09/24 15:54:48 00 17:40:19 00 00:00:00 2 node-blc28 user 09/24 15:54:48 00 00:01:22 00 00:00:00 3 node-cc08d5 none 09/24 15:56:10 00 17:38:57 00 00:00:00 4 node-blc28 user 09/24 15:56:10 00 00:20:27 00 00:00:00 5 node-blc07 none 09/24 16:16:37 00 17:18:30 00 00:00:00 6 node-blc28 user 09/24 16:16:37 00 00:28:09 00 00:00:00 7 node-blc02 none 09/24 16:44:46 00 16:50:21 00 00:00:00 8 node-blc28 none - 268 07:35:0 00 00:00:00 So do I delete just SEQ 7 ? What do I replace in VM_POOL ? Here the in column body seems to list only 8 ? Yours, Steffen 100400oneadminoneadminone-100413485594223313450288060one-1004335544324312015720310559760171 8node-blc28/cumulus/one/var/3700vmm_kvmtm_shared000 On Tue, 2012-09-25 at 09:44 +0200, Steffen Neumann wrote: > Hi, > > I have a working one-3.0 cluster, and for upgrading to 3.6 > I need to move around VMs. The cluster has been up and > running for about a year now. > > At least two VMs on one host are stubbornly refusing > to live-migrate: > > oneadmin@cumulus:~$ onevm livemigrate 1004 55 > [VirtualMachineDeploy] Wrong state to perform action > > As the machine is RUNNING, I can ssh into it, > I am a bit at a loss here. Only strange thing is > that current VM monitoring reports "Net_TX: -1" > which happens since August 17th. > > Any idea ? > > sneumann: can you send onevm show of that machine to the mailing > list? > sneumann: there could be something wrong with the history and my > pals may know what it is > > onehost list: > ID NAME RVM TCPU FCPU ACPU TMEM FMEM AMEM STAT > 55 node-blc02 3 2400 2393 0 94.6G 71.6G 46.6G on > 37 node-blc28 2 2400766800 94.6G 48.7G 30.6G on > ... > > onevm show 1004: > VIRTUAL MACHINE 1004 INFORMATION > > ID : 1004 > NAME: one-1004 > USER: oneadmin > GROUP : oneadmin > STATE : ACTIVE > LCM_STATE : RUNNING > HOSTNAME: node-blc28 > START TIME : 08/15 13:06:46 > END TIME: - > DEPLOY ID : one-1004 > > VIRTUAL MACHINE MONITORING > > USED MEMORY : 33554432 > NET_TX : 2015720310 > USED CPU: 431 > NET_RX : 559760171 > > VIRTUAL MACHINE TEMPLATE > > CPU=8.0 > DISK=[ > DISK_ID=0, > FILES=/cumulus/goldenimages/suse113/init.sh, > READONLY=no, > SOURCE=/cumulus/goldenimages/suse113/suse113.img, > TARGET=hda ] > FEATURES=[ > ACPI=yes ] > GRAPHICS=[ > LISTEN=0.0.0.0, > PORT=6904, > TYPE=vnc ] > MEMORY=32768 > NAME=one-1004 > NIC=[ > BRIDGE=br0, > IP=10.20.2.30, > MAC=02:00:0a:14:02:1e, > MODEL=virtio, > NETWORK=VNET202, > NETWORK_ID=3 ] > NIC=[ > BRIDGE=br1, > IP=10.21.2.30, > MAC=02:00:0a:15:02:1e, > MODEL=virtio, > NETWORK=VNET212, > NETWORK_ID=2 ] > OS=[ > ARCH=x86_64, > BOOT=hd ] > RAW=[ > DATA=" port=\"0\"/> path=\"/dev/pts/5\"/>", > TYPE=kvm ] > TEMPLATE_ID=1 > VCPU=8 > VMID=1004 > > VIRTUAL MACHINE HISTORY > > SEQHOSTNAME REASON STARTTIME PTIME >0 node-blc28 user 08/15 13:07:13 40 02:47:35 00 00:00:15 >1 node-cc08d5 none 09/24 15:54:48 00 17:40:19 00 00:00:00 >2 node-blc28 user 09/24 15:54:48 00 00:01:22 00 00:00:00 >3 node-cc08d5 none 09/24 15:56:10 00 17:38:57 00 00:00:00 >4 node-blc28 user 09/24 15:56:10 00 00:20:27 00 00:00:00 >5 node-blc07 none 09/24 16:16:37 00 17:18:30 00 00:00:00 >6 node-blc28 user 09/24 16:16:37 00 00:28:09 00 00:00:00 >7 node-blc02 none 09/24
[one-users] "Wrong state to perform action" when onevm livemigrate'ing a RUNNING machine ?
Hi, I have a working one-3.0 cluster, and for upgrading to 3.6 I need to move around VMs. The cluster has been up and running for about a year now. At least two VMs on one host are stubbornly refusing to live-migrate: oneadmin@cumulus:~$ onevm livemigrate 1004 55 [VirtualMachineDeploy] Wrong state to perform action As the machine is RUNNING, I can ssh into it, I am a bit at a loss here. Only strange thing is that current VM monitoring reports "Net_TX: -1" which happens since August 17th. Any idea ? sneumann: can you send onevm show of that machine to the mailing list? sneumann: there could be something wrong with the history and my pals may know what it is onehost list: ID NAME RVM TCPU FCPU ACPU TMEM FMEM AMEM STAT 55 node-blc02 3 2400 2393 0 94.6G 71.6G 46.6G on 37 node-blc28 2 2400766800 94.6G 48.7G 30.6G on ... onevm show 1004: VIRTUAL MACHINE 1004 INFORMATION ID : 1004 NAME: one-1004 USER: oneadmin GROUP : oneadmin STATE : ACTIVE LCM_STATE : RUNNING HOSTNAME: node-blc28 START TIME : 08/15 13:06:46 END TIME: - DEPLOY ID : one-1004 VIRTUAL MACHINE MONITORING USED MEMORY : 33554432 NET_TX : 2015720310 USED CPU: 431 NET_RX : 559760171 VIRTUAL MACHINE TEMPLATE CPU=8.0 DISK=[ DISK_ID=0, FILES=/cumulus/goldenimages/suse113/init.sh, READONLY=no, SOURCE=/cumulus/goldenimages/suse113/suse113.img, TARGET=hda ] FEATURES=[ ACPI=yes ] GRAPHICS=[ LISTEN=0.0.0.0, PORT=6904, TYPE=vnc ] MEMORY=32768 NAME=one-1004 NIC=[ BRIDGE=br0, IP=10.20.2.30, MAC=02:00:0a:14:02:1e, MODEL=virtio, NETWORK=VNET202, NETWORK_ID=3 ] NIC=[ BRIDGE=br1, IP=10.21.2.30, MAC=02:00:0a:15:02:1e, MODEL=virtio, NETWORK=VNET212, NETWORK_ID=2 ] OS=[ ARCH=x86_64, BOOT=hd ] RAW=[ DATA="", TYPE=kvm ] TEMPLATE_ID=1 VCPU=8 VMID=1004 VIRTUAL MACHINE HISTORY SEQHOSTNAME REASON STARTTIME PTIME 0 node-blc28 user 08/15 13:07:13 40 02:47:35 00 00:00:15 1 node-cc08d5 none 09/24 15:54:48 00 17:40:19 00 00:00:00 2 node-blc28 user 09/24 15:54:48 00 00:01:22 00 00:00:00 3 node-cc08d5 none 09/24 15:56:10 00 17:38:57 00 00:00:00 4 node-blc28 user 09/24 15:56:10 00 00:20:27 00 00:00:00 5 node-blc07 none 09/24 16:16:37 00 17:18:30 00 00:00:00 6 node-blc28 user 09/24 16:16:37 00 00:28:09 00 00:00:00 7 node-blc02 none 09/24 16:44:46 00 16:50:21 00 00:00:00 8 node-blc28 none - 268 07:35:0 00 00:00:00 oned.log: Tue Sep 25 09:35:37 2012 [ReM][D]: HostPoolInfo method invoked Tue Sep 25 09:35:37 2012 [ReM][D]: VirtualMachinePoolInfo method invoked Tue Sep 25 09:35:37 2012 [ReM][D]: AclInfo method invoked Tue Sep 25 09:35:50 2012 [ReM][D]: VirtualMachineInfo method invoked Tue Sep 25 09:35:50 2012 [ReM][D]: VirtualMachineDeploy method invoked Tue Sep 25 09:35:50 2012 [ReM][E]: [VirtualMachineDeploy] Wrong state to perform action 1004/vm.log: Tue Sep 25 09:30:22 2012 [VMM][D]: Monitor Information: CPU : 431 Memory: 33554432 Net_TX: -1 Net_RX: -1 -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
[one-users] Manually correcting ACPU ?
Hi, we have a mid-sized cluster running ONE-3.0 on top of Ubuntu 11.10. We have several cases where the ACPU calculation is in an inconsistent state. ID NAME RVM TCPU FCPU ACPU TMEM FMEM AMEM STAT 32 node-blc17 3 2400164 -1600 94.6G 27.8G -35014 on It should be TCPU - ( sum of all running VCPU )*100 on this host, and definitely not negative... Note that the onehost list shows STAT=on, while onehost show (see below) still shows confusingly the error from earlier this afternoon. Question(s): How Do I manually sanitize the ACPU count on this host ? We're using mysql. I only found the "Database Schema for Accounting", but no documentation for the "normal" schema. In which table do I have to change what ? It would be a good feature if a re-calculation of ACPU could be triggered, based on all VMs currently allocated to this host. Any idea how that could have happened in first place, so we can avoid it in the future ? Yours, Steffen On Thu, 2011-07-07 at 10:22 +0200, Carlos Martín Sánchez wrote: > Subject: > Re: [one-users] VMs in PENDING > [was: Fixing up USED / FREECPU ?] ... > Maybe these kind of inconsistencies could be managed by the new onedb > command, we'll consider adding this feature for next releases. HOST 32 INFORMATION ID: 32 NAME : node-blc17 STATE : MONITORED IM_MAD: im_kvm VM_MAD: vmm_kvm TM_MAD: tm_shared HOST SHARES MAX MEM : 99202928 USED MEM (REAL) : 70086068 USED MEM (ALLOCATED) : 134217728 MAX CPU : 2400 USED CPU (REAL) : 1411 USED CPU (ALLOCATED) : 4000 RUNNING VMS : 3 MONITORING INFORMATION ARCH=x86_64 CLUSTER=default CPUSPEED=2300 ERROR=[ MESSAGE="Error monitoring host 32 : MONITOR FAILURE 32 - ", TIMESTAMP="Thu Mar 22 15:51:48 2012" ] FREECPU=988.8 FREEMEMORY=29116860 HOSTNAME=node-blc17 HYPERVISOR=kvm MODELNAME="AMD Opteron(tm) Processor 6176" NETRX=0 NETTX=0 TOTALCPU=2400 TOTALMEMORY=99202928 USEDCPU=1411.2 USEDMEMORY=70086068 -- IPB Halle AG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] How to get information of actual VMs from inside a cluster
Hi, On Wed, 2011-12-14 at 18:24 +0900, Akihiko Ota wrote: > I intend to build VM clusters for high-performance computing by > OpenNebula. We have a "virtual" HPC cluster, using the SGE, and the nodes register themselves as SGE hosts during bootup. > But, in order to submit MPI job, it is necessary to know actual > compute-node (hostname, IP address, cores, etc...). > The number of VM of cluster can change on-demand by OpenNebula. > But maybe each VM cannot know change of member of cluster. Another possibility would be to run ganglia on the virtual nodes, and query ganglia for the running hosts available for the MPI jobs. Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] OpenNebula ageing effect?
Hi, On Fri, 2011-11-04 at 20:01 +, Vinícius Vielmo Cogo wrote: > Which can be the possible causes? What DB are you using ? SQLite or MySQL ? There are a few DB queries, although I'd be surprised if they amounted to 140sec. For mysql, you could turn on query logging, and run some queries manually to check. Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] Configurable location for $ONE_LOCATION/var/one.db ?
Thanks for your reply, so I know it is not just me. On Wed, 2011-10-26 at 10:20 -0400, Robert Parrott wrote: ... > consider this a defect, though I haven't filed a ticket on it. Done: http://dev.opennebula.org/issues/943 Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
[one-users] Configurable location for $ONE_LOCATION/var/one.db ?
Hi, I have to move one.db (which is currently living on an NFS volume, which is dangerous according to the docs[1]) to a local filesystem on the head node. Where do I specify the path to one.db ? The documentation [2] does not mention further arguments in oned.conf DB section DB = [ backend = "sqlite" ]. The only reference to a variable DB location I found was DB: sqlite:///var/one/oneacct.db in $ONE_LOCATION/etc/acctd.conf but nothing for one.db. Do I need to hack the source or switch to MySQL ? Serving $ONE_LOCATION from a central storage/NAS might be a common scenario, so this could be a fairly important issue. Yours, Steffen [1] http://opennebula.org/documentation:rel3.0:sfs [2] http://opennebula.org/documentation:archives:rel2.2:oned_conf -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] cancel live migration in progress
Hi, On Thu, 2011-10-20 at 10:48 +0200, samuel wrote: > After almost 2 days, the migration has succeed: I didn't read the beginning of the thread, but just wanted to share my observation: I also had a VM that didn't live-migrate immediately. This was a 4 VCPU VM running 4 compute (and memory) intensive processes, and I got impatient after even just 10 minutes. Apparently, the migration had a tough time sync'ing the permanently changing memory to the new node. Once I had ssh'ed in, I used kill -STOP to pause these processes. Then I saw the blinking cursor freeze for a fraction of a second, and could resume the processes via kill -CONT again. Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] OpenNbula 3.0 migration: problem with onedb
Hi, On Tue, 2011-10-18 at 16:41 +0200, Ruben Diez wrote: > We had used the /usr/share/one/install_gems script to install gems > Never the less seems that some of them need the corresponded devel c > library, and fails to install Could you file a bug report at http://dev.opennebula.org/projects/opennebula/issues so the developers can add checks to install_gems for successful compilation, and track once that's fixed ? Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] OpenNbula 3.0 migration: problem with onedb
On Tue, 2011-10-18 at 16:00 +0200, Ruben Diez wrote: > Seems to fail to find a function Perhaps any library is needed?? > We think all the ruby gems are installed, as the documentation explain I had exactly the same problem with the SQLite DB, when upgrading from 2.1 -> 3.0, and it was missing a gem for SQLite. Sure you have all *MySQL* prerequisites ? Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] Bacula and OpenNebula
Hi, On Mon, 2011-10-17 at 12:09 +0100, Richard Palmer wrote: > > Have you tried the XML output format of onevm ? > Ah perfect. Although I don't fancy trying to parse XML in a shell script, > will move to perl/python. Could you get away with xmllint, xpath or some xslt and/on the command line ? Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] Bacula and OpenNebula
On Mon, 2011-10-17 at 11:45 +0100, Richard Palmer wrote: ... > I'm trying to write a suspend/resume script for bacula to enable > it to backup (running) virtual machine images. I'd like to be able > to map from a fixed server name to the dynamic vm_id, but can't see > any obvious way to do this. Any suggestions ? Have you tried the XML output format of onevm ? Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] VM stays in MIGRATE when trying to livemigrate it
On Wed, 2011-08-31 at 21:40 +0200, Adnan Pasic wrote: ... > However, now when I try to migrate or livemigrate any deployed VM, > this VM stays in SAVE or MIGRATE respectfully. Nothing else happens. For live-migration: Is the VM quite busy ? During live-migrate the RAM is synch'ed over to the new node, until nothing changes anymore. So if 1) your network is slow-ish (say, 100Mbit) and/or 2) you have one or many processes running within the VM which alter RAM, then the sync'ing can't keep up with the changes, hence the migration never finished. => ssh into the VM, "kill -STOP pids" the most active processes. You might observe that your cursor freezes for a second, then that was the migration (optionally: verify RUNNING), and "kill -CONT pids". Of course, it could also be something else Yours, Steffen -- IPB Halle AG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] OpenNebula 3.0: Lot of execution expired
Hi, On Mon, 2011-08-29 at 12:33 +0200, Jonathan Rouzaud-Cornabas wrote: > I use MySQL now and it helps to fix lot of the execution expired. > Also by tweaking NFS, I was able to make disappear all the remaining > execution expired. Thanks for the feedback ! Glad it worked out, we'll add a number of nodes soon, so we'll start to tune right from the start ;-) Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] Why to use Opennebula?
Hi, On Mon, 2011-08-29 at 07:52 +, Juan José Fuentes wrote: > I'm back from holidays and I don't remember ? > why I should use Opennebula instead of Xen (that it also allows live > migration) or someother products... What for ? Unless you lay out the use case, the answers can only point into the blue! > I don't know if this is the appropiate forum basically yes. > excuse me if it isn't, but actually I just want to justify to my boss the use > of this product. > Also I want to ask to the community if someone is using Opennebula in > production environments. > We are thinking about providing our products as SaaS, > and one way could be creating our own private cloud, I was wondering if > Opennebula could meet those needs. Depending on your needs, you might consider the professional C12G support. On our smallish cluster for research related simulations we don't need it, but if some business relies on it, you might be better off with some professionals for support cases. Yours, Steffen -- IPB Halle AG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] OpenNebula 3.0: Lot of execution expired
Hi, On Thu, 2011-08-18 at 14:35 +0200, Jonathan Rouzaud-Cornabas wrote: > I have setup a cluster of 15 nodes (24 cores / 48Gb of RAM per node). > I use OpenNebula 3.0 Beta with NFS and KVM. ONE has been deployed on much larger installations, but IIRC there is some performance tuning document in the knowledgebase. > My issue is that I received a lot of "execution expired" when I use > onetemplate create and or instantiate but also any other opennebula > command. For example: > # onevm list > execution expired You are using the sqlite backend ? If so, is the one.db file located on an NFS share ? It might be better if it is local. You could also migrate to MySQL. Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] Sunstone question/trouble
Hi, On Tue, 2011-07-12 at 17:45 -0400, Robert Schweikert wrote: > OK, now I see the problem. This depends on the sunstone-server running > as the oneadmin user with the HOME environment variable being set > properly by the shell. Why isn't HOME set correctly ? If it is run via su oneadmin -l -s /bin/bash -c 'one start' all should be fine ? > I think this is not correct, The code should parse /etc/password for the No! User code should not read /etc/passwd itself. What about other means like NIS, LDAP, AD, ... ? Use getent instead, which handles everything in /etc/nsswitch automagically. Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] VMs in PENDING [was: Fixing up USED / FREECPU ?]
Hi, On Wed, 2011-07-06 at 16:25 +0200, Steffen Neumann wrote: ... > That host surely was running with USEDCPU 1280 at some time, > but I had to restart oned because of an I/O error on one.db, > so I guess the book keeping is inconsistent because currently > there are only two VMs running. > > So I am back to my original question: how do I tell ONE > the *correct* USEDCPU / FREECPU ? Delete / Add the whole host again ? > Or is there anything less invasive I can do ? Since I *need* that host to accept VMs now, I migrated all VMs I could see away from it, and wanted to delete the whole host, but I got oneadmin@cumulus:~/var$ onehost delete 12 Host still has associated VMs, aborting delete. So this again revolves around: how do I purge this non-existing cold dead VM out of ONE's memory ? Since the one* tools can't figure out what's running there, I will have to manually modify one.db ? Which tables in there ? Or did I miss the point ? Yours, Steffen oneadmin@cumulus:~/var$ onehost show 12 -x 12 node-blc02 0 im_kvm vmm_kvm tm_nfs 1308659636 default 12 0 8388608 100 0 66108092 1600 0 34700700 320 0 31407392 1280 1 -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] VMs in PENDING [was: Fixing up USED / FREECPU ?]
Hi, On Wed, 2011-07-06 at 15:46 +0200, Ruben S. Montero wrote: > If you look at the xml info for node–blc02, it reports 2 running vms > and a free cpu of 320 (you are requesting 400). Yes, that's why I originally asked about fixing up FREECPU, because those two VMs (the second one I started while writing the mail, it has just one CPU and was to check that node-blc02 is _generally_ working) together only have 5 CPUs, so a 16 core host should be able to run more than that. That host surely was running with USEDCPU 1280 at some time, but I had to restart oned because of an I/O error on one.db, so I guess the book keeping is inconsistent because currently there are only two VMs running. So I am back to my original question: how do I tell ONE the *correct* USEDCPU / FREECPU ? Delete / Add the whole host again ? Or is there anything less invasive I can do ? Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] VMs in PENDING [was: Fixing up USED / FREECPU ?]
Hi, I am not so sure anymore where the problem of my VMs in PENDING is rooted, so I'll give some more information that were asked in previous posts about stuck in pending. I try to deploy VMID 568 with (V)CPU=4, and 8192M RAM, and I'd *think* it should deploy nicely on host node-blc02, on that host there is only VMID=506 with 2 CPUs running. Any clues how to debug this ? Yours, Steffen ID USER NAME STAT CPU MEMHOSTNAMETIME 437 oneadmin sus113_8 runn 176 32G node-blc09 53 17:59:04 438 oneadmin sus113_8 runn 228 32G node-blc10 53 17:52:50 506 oneadmin sus113_2 runn 121 8G node-blc02 14 17:54:19 507 oneadmin sus113_2 runn 1 8G node-blc09 14 17:54:16 508 oneadmin sus113_4 runn 1 16G node-blc09 14 17:02:01 509 oneadmin sus113_4 runn 1 16G node-blc09 14 17:01:53 510 oneadmin sus113_2 runn 64 8G node-blc10 14 16:56:20 511 oneadmin sus113_2 runn 3 8G node-blc10 14 16:56:18 558 oneadmin ubuntulu runn 398 8G node-blc01 01 17:04:45 559 oneadmin ubuntulu runn 398 8G node-blc01 01 17:03:43 560 oneadmin ubuntulu runn 398 8G node-blc01 01 17:02:39 561 oneadmin ubuntulu runn 398 8G node-blc01 01 17:01:42 562 oneadmin ubuntulu runn 400 8G node-blc10 01 17:00:39 =>568 oneadmin ubuntulu pend 0 0K 00 00:02:53 ID NAME CLUSTER RVM TCPU FCPU ACPUTMEMFMEM STAT 6 node-blc10default4 1600 1258 20 63G 54G on 7 node-blc09default4 1600 1276 -200 63G 46.2G on 8 node-blc01default4 1600800 80 63G 38.3G on 10 node-cc09d5 default0800197800 126.2G 107.2G on 11 node-cc09d6 default0800 65800 126.2G 90.2G on =>12 node-blc02default1 1600320 1400 63G 33.1G on 13 node-cc08d5 default0800203800 31.5G 13.6G off VIRTUAL MACHINE 568 INFORMATION ID : 568 NAME : ubuntulucid STATE : PENDING LCM_STATE : LCM_INIT START TIME : 07/06 07:52:01 END TIME : - DEPLOY ID: : VIRTUAL MACHINE MONITORING NET_TX : 0 NET_RX : 0 USED MEMORY: 0 USED CPU : 0 VIRTUAL MACHINE TEMPLATE CPU=4 DISK=[ BUS=virtio, DISK_ID=0, FILES=/cumulus/goldenimages/ubuntulucid/init.sh /cumulus/goldenimages/ubuntulucid/id_dsa.pub, READONLY=no, SOURCE=/cumulus/goldenimages/ubuntulucid/ubuntulucid.img, TARGET=hda ] DISK=[ BUS=virtio, DISK_ID=1, FORMAT=ext4, READONLY=no, SIZE=1024, TARGET=hdb, TYPE=fs ] FEATURES=[ ACPI=yes ] GRAPHICS=[ LISTEN=0.0.0.0, PORT=6468, TYPE=vnc ] MEMORY=8192 NAME=ubuntulucid NIC=[ BRIDGE=br0, IP=10.20.1.75, MAC=02:00:0a:14:01:4b, MODEL=virtio, NETWORK=VNET201, NETWORK_ID=0 ] NIC=[ BRIDGE=br1, IP=10.21.1.75, MAC=02:00:0a:15:01:4b, MODEL=virtio, NETWORK=VNET211, NETWORK_ID=1 ] OS=[ ARCH=x86_64, BOOT=hd ] RAW=[ DATA=, TYPE=kvm ] VCPU=4 VMID=568 sched.log: v This is the node-blc02 Wed Jul 6 09:26:06 2011 [HOST][D]: Discovered Hosts (enabled): 6 7 8 12 Wed Jul 6 09:26:06 2011 [VM][D]: Pending virtual machines : 568 569 Wed Jul 6 09:26:06 2011 [RANK][W]: No rank defined for VM Wed Jul 6 09:26:06 2011 [RANK][W]: No rank defined for VM Wed Jul 6 09:26:06 2011 [SCHED][I]: Select hosts PRI HID --- Virtual Machine: 568 Virtual Machine: 569 12 node-blc02 0 im_kvm vmm_kvm tm_nfs 1308659636 default 12 0 16777216 300 0 66108092 1600 0 34700700 320 0 31407392 1280 2 -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
[one-users] Fixing up USED / FREECPU ?
Hi, due to some problems (NFS connectivity) earlier on, ONE's idea of free CPU on some nodes differs from reality and new VMs are stuck PENDING. E.g. I find from onehost show node-blc02 TOTALCPU=1600 USEDCPU=1280.0 whereas I know from onevm list that there is only a single KVM running: onevm list | grep node-blc02 506 oneadmin suse113 runn 121 8G node-blc02 13 01:18:38 So USEDCPU should really be about 200 or so. Question: How do I "fix" ONE's idea about the USED/FREE CPU for individual hosts ? As there are still VMs, rebooting is not really an option. Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] Unable to Ping Migrated VM
Hi, On Mon, 2011-06-06 at 11:59 +0300, Koushik Annapureddy wrote: > Can any one help in debugging this problem. First idea is bridge setup, or maybe firewall rules. 1) Does ping work when you migrate back ? 2) Can you deploy a VM directly to node2 ? (You may have to disable node1 so that subsequent onevm create go directly to node2) Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] live-/migration not working, uknown error
Hi, On Mon, 2011-05-30 at 13:02 +0200, Adnan Pasic wrote: > machines normally. The only problem I’, still having is the Live > Migration! ... > Is it possible that I somehow messed up the TLS encryption? I CHeck out the list archives, in opennebula-2.2 you need to change qmeu to qemu+ssh in one of the scripts. If you don't find it, ask again. Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] VM starting fail
Hi, On Tue, 2011-05-24 at 09:23 +0200, genna...@email.it wrote: > prolog, error getting driver tm_ss This is surely a typo in the mail ^ you did specify tm_ssh ? Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] Open Nebula, NFS and requirements of disk space?
Ah, now I get the idea: the golden images are on the headnode, exported to the worker nodes via NFS, and then they're cloned onto the local worker disk drives with 36GB. Did I get that right ? Now, speed-wise the only benefit would be if NFS gets the image to the worker nodes faster over the network than copying via ssh / scp. Which, I think, is only a few percent, if at all. The images definitely need to be copied fully, because each node needs its own copy. If all used the same image, the filesystem would go up in smoke if two nodes write to it without knowing about each other... So yes, it might be possible, but IMHO does not make a lot of sense. Or did I misunderstand anything ? Yours, Steffen On Mon, 2011-05-16 at 15:25 +0200, Anders Branderud wrote: > He thinks it is a faster solution - i.e. with NFS all of the image > doesn't necessarily need to be copied from the front end to the worker > node, but only the parts that the user needs. Thus it can be more > efficient. > "SSH Non shared" requires the image file to be cloned to all VM's, > which in many cases is unnecessary and isn't efficient. > > > Regards, Anders > > 2011/5/16 Steffen Neumann > On Mon, 2011-05-16 at 13:48 +0200, Anders Branderud wrote: > > Hello! > > Thanks Steffen! > > > > Yes, I am aware of the "Non-Shared - SSH"-solution. > > However, the person that I am doing my thesis for, want it > to be done > > with NFS. > > What does he want on NFS ? > > Gruss, > > Steffen > > -- > IPB HalleAG Massenspektrometrie & > Bioinformatik > Dr. Steffen Neumann http://www.IPB-Halle.DE > Weinberg 3 http://msbi.bic-gh.de > 06120 Halle Tel. +49 (0) 345 5582 - 1470 > +49 (0) 345 5582 - 0 > sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 > > > > > > > -- > Kind Regards, Anders Branderud > > [Personal blog] bloganders.blogspot.com : Logical reasons - based on > scientific premises - for the existence of a Creator and that He > hasn't left His sapient creatures without an Instruction Manual - > Torah ['books of Moses'] - to ascertain, and aspire to, His purpose. > > [Company] Anders Branderud IT Solutions - www.abitsolutions.org > -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] allocating IP for a VM.
Hi, On Fri, 2011-05-13 at 11:14 +0700, Khoa Nguyen wrote: > Thank you for your reply? How to prevent the other DHCP server from > answering to my VMs? I suggest you talk to your network administrator on that. We have our ONE nodes on a private 10.x.x.x network, so the other DHCP doesn't see the DHCPREQUESTS in first place. Another way is that you talk to the admin of the other DHCP server, and negotiate that they ignore a certain MAC prefix, which you can define for all your VMs. We used dnsmasq btw, because it also handles DNS. Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
[one-users] Resubmit from sunstone ?
Hi, I have a sunstone question: I just performed a "onevm resubmit 428 429 430" because they had FAILed, how can I do the same with sunstone ? The menue with the VM actions does not contain "resubmit" ? Is that on purpose ? Or am I blind somewhere ? Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
[one-users] VMs still reported as running although host is down
Hi, I shut down some hosts (had to, there were NFS issues) and didn't delete the VMs beforehand. I am a bit surprised that onehost list still shows the node as "on": node02 default2 1600 1280 1200 63G 58.1G on and that onevm list also reports the VMs as running: 386 oneadmin suse113 runn 22 8G node02 08 06:51:45 1) Why doesn't onehost report "err" if it can't execute the probes on the shutdown node ? 2) Why doesn't onevm report "unknown" or whatever if it can't reach the not-running VM on the not-running node ? Should I open (one or two ?) tickets ? Yours, Steffen -- IPB Halle AG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] Feature request: Locking of VM, images, nets...
Hi, I think coming versions of ONE will have better authentication + permission handling. Then you'd simply not allow "normal" users to do dangerous stuff. Steffen On Wed, 2011-05-11 at 10:58 +0200, Giovanni Toraldo wrote: > On 11/05/2011 10:52, Luis M. Carril wrote: > > I think that would be great if a set of locking commands to the > > interface are implemented to avoid deleting the wrong virtual machine > > due to a typing error. > > > > Something like: onevm lock vm_iddenys all operations one the > > virtual machine vm_id until a onevm unlock vm_id is done > > I don't think that it will be a solution to the problem you surfaced: > people will start to use unlock command without keeping attention too. > > Maybe a confirmation for disruptive commands will be better (Are you > really sure to delete VM nn? y/n) > > Bye. > > ___ > Users mailing list > Users@lists.opennebula.org > http://lists.opennebula.org/listinfo.cgi/users-opennebula.org -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] allocating IP for a VM.
Hi, On Wed, 2011-05-11 at 15:57 +0700, Khoa Nguyen wrote: > But I don't know to do > Add entries to your DHCP server binding the MAC address OpenNebula will ... > If i can't configure DHCP server, can i do? No. We have a little script (pre-)generating a part of /etc/dnsmasq.conf. If you don't have control over your "official" DHCP, you could try to setup your own for the ONE cluster, and somehow prevent the other DHCP server from answering to your VMs. Yours, Steffen dhcp-host=02:00:0a:14:01:08, net:cluster, 10.20.1.8, vubuntu008 dhcp-host=02:00:0a:15:01:08, net:nfs, 10.21.1.8, vubuntu008-nfs dhcp-host=02:00:0a:14:02:08, net:cluster, 10.20.2.8, vsuse008 dhcp-host=02:00:0a:15:02:08, net:nfs, 10.21.2.8, vsuse008-nfs dhcp-host=02:00:0a:14:01:09, net:cluster, 10.20.1.9, vubuntu009 dhcp-host=02:00:0a:15:01:09, net:nfs, 10.21.1.9, vubuntu009-nfs dhcp-host=02:00:0a:14:02:09, net:cluster, 10.20.2.9, vsuse009 dhcp-host=02:00:0a:15:02:09, net:nfs, 10.21.2.9, vsuse009-nfs dhcp-host=02:00:0a:14:01:0a, net:cluster, 10.20.1.10, vubuntu010 dhcp-host=02:00:0a:15:01:0a, net:nfs, 10.21.1.10, vubuntu010-nfs dhcp-host=02:00:0a:14:02:0a, net:cluster, 10.20.2.10, vsuse010 dhcp-host=02:00:0a:15:02:0a, net:nfs, 10.21.2.10, vsuse010-nfs dhcp-host=02:00:0a:14:01:0b, net:cluster, 10.20.1.11, vubuntu011 dhcp-host=02:00:0a:15:01:0b, net:nfs, 10.21.1.11, vubuntu011-nfs dhcp-host=02:00:0a:14:02:0b, net:cluster, 10.20.2.11, vsuse011 dhcp-host=02:00:0a:15:02:0b, net:nfs, 10.21.2.11, vsuse011-nfs -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] Beginner questions: Opennebula 2.2.1
Hi, one add-on question: On Fri, 2011-05-06 at 12:15 +0200, Tino Vazquez wrote: ... > Shutdown represents a more clean approach, the VM is cleanly > shut down, and the image saved if marked so. Is that atomic ? Or under file locking control ? If I have some place, say /srv/could/goldenimage/ubuntu.img where my disk image file is stored, does shutdown then overwrite the existing ubuntu.img, and what happens if during the XX seconds of writing back a new VM is created from ubuntu.img ? Unless "save" is atomic, the new VM will fail on a half-written mix of old and new. Or did I miss anything ? What happens if three VMs created off ubuntu.img shutdown simultanously ? Will each write its changes back into that single ubuntu.img ? Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] Error executing "econe-server start"
Hi, Check the message: Port busy, please shutdown the service or move econe server port. Which port did you change ? The 8880 ? Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] error at GPG key installation
Hi, that's probably not the right list for these questions, maybe better look for some CentOS specific forums. Yours, Steffen On Wed, 2011-04-27 at 11:36 +0530, soge.geo...@wipro.com wrote: > Hi, > > > Am installing OPENNEBULA for the first time in CENTOS,I have installed > Fedora Extras rebuild Repository and Misc Packages rebuild Repository. > By the help of http://centos.karan.org/ site, but for running GPG key > from centos terminals am getting an error. > Code and the error are the following > rpm --import http://centos.karan.org/RPM-GPG-KEY-karan.org.txt > > error:http://centos.karan.org/RPM-GPG-KEY-karan.org.txt:import read > failed(-1). > > > > Please kindly help me for debugging this error. > > > > Thanks and Regards, > > Soge George > > > > > > > Please do not print this email unless it is absolutely necessary. > > The information contained in this electronic message and any > attachments to this message are intended for the exclusive use of the > addressee(s) and may contain proprietary, confidential or privileged > information. If you are not the intended recipient, you should not > disseminate, distribute or copy this e-mail. Please notify the sender > immediately and destroy all copies of this message and any > attachments. > > WARNING: Computer viruses can be transmitted via email. The recipient > should check this email and any attachments for the presence of > viruses. The company accepts no liability for any damage caused by any > virus transmitted by this email. > > www.wipro.com > > ___ > Users mailing list > Users@lists.opennebula.org > http://lists.opennebula.org/listinfo.cgi/users-opennebula.org -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] Asking about onevm migrate command
Hi, On Wed, 2011-04-27 at 16:00 +0700, Hoang Dinh Phuc wrote: > I'm using OpenNebula 2.2, and 2 kvm server (Ubuntu 10.10). Can you be a bit more specific ? Do you have a shared NFS installation, or ssh to migrate images ? Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] Specify VM boot order
Hi, On Sun, 2011-04-24 at 15:07 -0500, Rangababu Chakravarthula wrote: > Does Opennebula expose Libvirt's VM border order option ? No. There was a thread on this list a few weeks ago, check the archives. > If not that would be a good feature to add I'd second that. Feel free to open a feature request on the http://dev.opennebula.org/ tracker. Do other hypervisors (XEN, etc) support the same concept ? Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] Opennebula Deltacloud Adaptor
Hi Sebastian, could you be more specific ? What error occurs ? It might be a git / firewall / proxy problem ? On Sun, 2011-04-24 at 13:05 +0100, Sebastian Neg wrote: > it's the same thing, i can not get the source ... > On Sun, Apr 24, 2011 at 12:43 PM, Damien Raude-Morvan > wrote: ... > Try with git.opennebula.org : > git clone git://git.opennebula.org/deltacloud-adaptor.git Works for me, Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] Virt-manager created Ubuntu image without an IP address
Hi, On Mon, 2011-04-18 at 15:31 +0200, Idafen Santana Pérez wrote: > interfaces (ifconfig -a): lo and eth1. But when I run the VM with > opennebula (onevm create template.one) with the same VM image as > before, it shows lo and eth2, and I have no clue about this behaviour. https://help.ubuntu.com/10.04/serverguide/C/network-configuration.html You have to disable (or fix) that persistent-net behaviour in Ubuntu. Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] Live migrate failed
Hi, I have a similar problem here, and I wonder whether there is a simpler solution, if you're in a confined and trusted environment. http://opennebula.org/documentation:rel2.2:kvmg says: OpenNebula uses libvirt's migration capabilities. More precisely, it uses the TCP protocol offered by libvirt. In order to configure the physical nodes, the following files have to be modified: /etc/libvirt/libvirtd.conf : Uncomment “listen_tcp = 1”. Security configuration is left to the admin's choice, file is full of useful comments to achieve a correct configuration. /etc/default/libvirt-bin : add -l option to libvirtd_opts under the hood one should call virsh --connect qemu+ssh:///system migrate --live one-0 qemu+ssh://node-blc09/system but jfontan just confirmed on IRC that remotes/vmm/kvm/kvmrc should have "export QEMU_PROTOCOL=qemu+ssh" but it is missing the +ssh bit in the current 2.2 release. Fix that both on the nodes in /var/tmp/one/vmm/kvm/kvmrc but also on the oned host, so that it also works if you add future hosts. Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] Sunstone in ubuntu
Hi, On Mon, 2011-04-11 at 19:21 +0800, Jeff wrote: > Thank you olivier. It seems that there is something wrong with the > option -H when using sunstone-server command .So I edited the > $ONE_LOCATION/bin/sunstone-server and replaced line ' HOST="127.0.0.1" > ' with 'HOST="myip"' ,then it worked. That's kind of weird. Maybe it > is a bug ? No, that's how you configure how a server should LISTEN, you can restrict that to interfaces, and 127.0.0.1 would be lo instead of eth0. You can also set HOST="0.0.0.0" and it'll listen on all interfaces, IIRC. Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] Authentication ldap and authdb issues
Hi, On Mon, 2011-04-11 at 22:54 +0530, Madhurranjan Mohaan wrote: > I am trying to use the ldap add on to integrate the setup with ldap I have some followup questions: 1) The documentation suggests that LDAP integration was created for the 2.0 release: documentation:archives:rel2.0:ldap I assume it still (is supposed to) work with 2.2 ? 2) The user should add its credentials to $ONE_AUTH file (usually $HOME/.one/one_auth) in this fashion: user_dn:plain:user_password Does that mean the users have to put their passwds in plain text into their homedir ?? This would be a gaping security hole. Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] Sunstone: how detach cdrom ?
Hi, On Tue, 2011-03-29 at 12:45 +0200, ignazio.cass...@provincia.torino.it > After installation the vm must boot from hd rather than cdrom but using > sunstone I cannot detach the cdrom or tell kvm to start vm from hd. and I'd like to ask a related question: if I installed from netboot via PXE, how to switch to booting from HD ? I think both questions boil down to "how do I specify a boot order analogous to the BIOS settings in a system" Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] config issue: can't swith "tm_ssh" mode to "tm_nfs" mode
On Fri, 2011-03-11 at 16:51 +0100, Olivier Sallou wrote: > Hi, > I have not tested it but will "one" allow to remove a host while still having > VM running on host? No, It'll complain and refuse. oneadmin@cumulus:~$ onehost delete 17 Host still has associated VMs, aborting delete. Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] VM with multiple CPUs in template ?
Hi, On Mon, 2011-02-28 at 13:56 +0100, Steffen Neumann wrote: > Question: How do I specify a dual/quad CPU in template.one ? I was able to specify a quad-CPU VM using virt-manager, and that's also used after deployment via ONE. The original question remains, is it possible to specify the number of processors afterwards in the VM template ? Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
[one-users] VM with multiple CPUs in template ?
Hi, Our applications benefit from SMP with multiple CPUs. Question: How do I specify a dual/quad CPU in template.one ? The documentation http://www.opennebula.org/documentation:rel2.0:template mentions fractions (e.g. 0.5) but a value of CPU=2.0 does not create a 2CPU system. I also tried two CPU=1.0 entries, also no luck. (We're running one-2.0 with KVM) Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] 2.2 release ?
Hi, On Mon, 2011-02-14 at 15:37 +0100, Zeeshan Ali Shah wrote: > what is the expected date for 2.2 release ? or where to access operation > centre beforehand ? I am unsure what you refer to by "operation center", but you can always go to http://dev.opennebula.org/projects/opennebula and check out the current source with git. You could also browse the code via http://dev.opennebula.org/projects/opennebula/repository and monitor the commit messages. Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] Load balancing hosts / "VM Consolidation Scheduler"
Hi, On Tue, 2011-02-08 at 11:27 +0100, Tino Vazquez wrote: ... > > 2) Is there a way to migrate VMs between hosts > > when load gets to high ? The VMs could become busy > > and saturate their host, or a few new hosts became available. > > Is live-migration a requirement for this ? > > You can use the default OpenNebula scheduler to achieve an evenly > distribution of resources (see the load-aware policy in [1]). You > won't get automatic migration out-of-the-box, possibles solutions are: > > * Extend the scheduler to implement a new migration policy > * Develop a script to monitor and eventually perform the migration If possible I'd like to avoid re-inventing the wheel, and I could imagine that my idea is of general interest. There is an early presentation http://www.slideshare.net/llorente/vm-management-for-greenit-data-centers which covers exactly what I've been looking for, especially the "VM Consolidation Scheduler" mentioned on p14. Is any code for this available ? Yours, Steffen -- IPB Halle AG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
[one-users] Load balancing hosts
Hi, I have a question about the scheduling in ONE: 1) Is there a way to migrate VMs off a host to another free host ? I know I can manually onevm migrate 56 13, but I had somehow expected that onehost 12 disable would clean out that host. 2) Is there a way to migrate VMs between hosts when load gets to high ? The VMs could become busy and saturate their host, or a few new hosts became available. Is live-migration a requirement for this ? I am fine with pointers into the fine manual, if the information is there it evaded me. Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
[one-users] passwordless ssh between onehosts
Hi, how is the passwordless ssh supposed to work ? I can ssh from head -> any node just fine, but "onevm migrate" requires passwordless ssh from node01 -> node02 so oneadmin@node01's id_rsa_pub needs to be in oneadmin@node02's authorized keys. This will explode with the number of hosts involved. Or did I miss something simpler ? Is that already handled ? Is the id_rsa_pub (supposed to be) the same on all hosts ? Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
[one-users] onevnet range network handing without range
Hi, from my old days I know about class A,B and C networks. That definition is also used in the TYPE = RANGED network definition. I see two problems there: 1) People might have different networks, e.g. e.g. a 192.168.0.0/23 with addresses from approx. 192.168.0.1 to 192.168.1.254 the onevnet definition should be as flexible. 2) In my 10.20.0.0/16 network, I want to have the physical hosts on 10.20.0.x and virtual machines on 10.20.1.y, essentially sharing the same subnet. onevnet is defined as class B 10.20.0.0, but it will happily provide a lease colliding with my (dhcp-assigned) physical hosts. The onevnet RANGE definition should allow a real range, similar to dnsmask's dhcp-range=10.20.0.10,10.20.0.250,1h Or did I miss anything ? Yours, Steffen -- IPB HalleAG Massenspektrometrie & Bioinformatik Dr. Steffen Neumann http://www.IPB-Halle.DE Weinberg 3 http://msbi.bic-gh.de 06120 Halle Tel. +49 (0) 345 5582 - 1470 +49 (0) 345 5582 - 0 sneumann(at)IPB-Halle.DE Fax. +49 (0) 345 5582 - 1409 ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
