Re: Creating a Template from VM and then VM's from that template

[Muhammad Adeel Zahid]

Hi Dag,


Unfortunately, I deleted everything (the template and the VMs) and now starting 
afresh. Will post if I have problems.


Regards

Adeel


From: Dag Sonstebo 
Sent: Thursday, October 19, 2017 4:17:24 AM
To: users@cloudstack.apache.org
Subject: Re: Creating a Template from VM and then VM's from that template

Could you post up the contents of your ifcfg-eth0 on the ISO VM and the 
template created VM? (Please don’t attach files since they are stripped by the 
mailing list).

The other thing I noticed just now is the documentation doesn’t mention 
cleaning your DHCP cache files – this also needs done on the source VM.

Regards,
Dag Sonstebo
Cloud Architect
ShapeBlue
 S: +44 20 3603 0540  | dag.sonst...@shapeblue.com | http://www.shapeblue.com 
 | Twitter:@ShapeBlue 



On 18/10/2017, 22:19, "Muhammad Adeel Zahid" <16030...@lums.edu.pk> wrote:

Hi Dag,


Thanks for the response. I created the first VM from ISO image. Then I 
followed the instructions in the link you posted but still having same problem. 
Any suggestions



Regards

Adeel


From: Dag Sonstebo 
Sent: Wednesday, October 18, 2017 9:08:15 PM
To: users@cloudstack.apache.org
Subject: Re: Creating a Template from VM and then VM's from that template

Hi Adeel,

You need to tidy up the first source VM before you create a template from 
this – it will most likely still have udev rules and hard coded MAC addresses 
still in it. This means when the first VM is created from the template the 
networking can’t find the NICs which were in the original VM and networking 
fails.

More information in 
http://docs.cloudstack.apache.org/projects/cloudstack-administration/en/4.9/templates.html

Regards,
Dag Sonstebo
Cloud Architect
ShapeBlue

On 18/10/2017, 15:59, "Muhammad Adeel Zahid" <16030...@lums.edu.pk> wrote:

Hello,


I have installed apache cloudstack suite on two of my machine. First 
machine acts both as management server and as hypervisor (kvm). My second 
machine acts only as hypervisor(kvm). I first created a VM, say vm1, from 
centOS ISO and added some software that I need new VM's to have. I, then, 
created a template, say t1,  from vm1 so that I can create new VMs using this 
template. But when I created a new VM say vm2 from t1 , it does not assume any 
IP address for eth0 interface. It only has loopback interface address.

The reason seems to be ifcfg-eth0 file in 
/etc/sysconfig/network-scripsts/ directory.

This file is exactly the same in vm1 and vm2 including HWADDR and UUID 
entries. I presume this is the reason vm2 is not being able to get an IP 
address. Any Suggestions? Am I doing it wrong way?


Regards

Adeel




dag.sonst...@shapeblue.com
www.shapeblue.com
53 Chandos Place, Covent Garden, London  WC2N 4HSUK
@shapeblue






dag.sonst...@shapeblue.com
www.shapeblue.com
53 Chandos Place, Covent Garden, London  WC2N 4HSUK
@shapeblue

My server running Cloudstack-management fails to shutdown properly

[Muhammad Adeel Zahid]

Hello,


I am running cloudstack on Cent OS 6 server. When I try to shut down this 
machine, it hangs. It fails to stop the cloudstack-management but it surpasses 
this point. Furthermore, during shutdown, machine hangs with this message


Unmounting NFS filesystem: unmount.nfs: /mnt/some-long-uuid: device is busy


After showing above message, it hangs and I have to power off the system.


How can I get around this problem.


Regards

Adeel

Re: Creating a Template from VM and then VM's from that template

[Dag Sonstebo]

No problem Adeel – in short you follow the same procedure whenever you create a 
template – whether this is for CloudStack – or for any other non-CloudStack 
hosted hypervisors. You are looking to strip out any NIC mac addresses and udev 
rules from your source VM, as well as any DHCP lease files. You also 
prepopulate with any SSH keys or root passwords you want your end users to 
utilise.

You’ll find a ton of guides for this – see a couple below (I’m not the owner of 
these):

https://access.redhat.com/documentation/en-US/Red_Hat_Enterprise_Virtualization/3.0/html/Evaluation_Guide/Evaluation_Guide-Create_RHEL_Template.html
 
https://banck.net/2016/04/creating-vmware-centos-7-template/ 

Regards,
Dag Sonstebo
Cloud Architect
ShapeBlue

On 19/10/2017, 08:55, "Muhammad Adeel Zahid" <16030...@lums.edu.pk> wrote:

Hi Dag,


Unfortunately, I deleted everything (the template and the VMs) and now 
starting afresh. Will post if I have problems.


Regards

Adeel


From: Dag Sonstebo 
Sent: Thursday, October 19, 2017 4:17:24 AM
To: users@cloudstack.apache.org
Subject: Re: Creating a Template from VM and then VM's from that template

Could you post up the contents of your ifcfg-eth0 on the ISO VM and the 
template created VM? (Please don’t attach files since they are stripped by the 
mailing list).

The other thing I noticed just now is the documentation doesn’t mention 
cleaning your DHCP cache files – this also needs done on the source VM.

Regards,
Dag Sonstebo
Cloud Architect
ShapeBlue
 S: +44 20 3603 0540  | dag.sonst...@shapeblue.com | 
http://www.shapeblue.com  | Twitter:@ShapeBlue 



On 18/10/2017, 22:19, "Muhammad Adeel Zahid" <16030...@lums.edu.pk> wrote:

Hi Dag,


Thanks for the response. I created the first VM from ISO image. Then I 
followed the instructions in the link you posted but still having same problem. 
Any suggestions



Regards

Adeel


From: Dag Sonstebo 
Sent: Wednesday, October 18, 2017 9:08:15 PM
To: users@cloudstack.apache.org
Subject: Re: Creating a Template from VM and then VM's from that 
template

Hi Adeel,

You need to tidy up the first source VM before you create a template 
from this – it will most likely still have udev rules and hard coded MAC 
addresses still in it. This means when the first VM is created from the 
template the networking can’t find the NICs which were in the original VM and 
networking fails.

More information in 
http://docs.cloudstack.apache.org/projects/cloudstack-administration/en/4.9/templates.html

Regards,
Dag Sonstebo
Cloud Architect
ShapeBlue

On 18/10/2017, 15:59, "Muhammad Adeel Zahid" <16030...@lums.edu.pk> 
wrote:

Hello,


I have installed apache cloudstack suite on two of my machine. 
First machine acts both as management server and as hypervisor (kvm). My second 
machine acts only as hypervisor(kvm). I first created a VM, say vm1, from 
centOS ISO and added some software that I need new VM's to have. I, then, 
created a template, say t1,  from vm1 so that I can create new VMs using this 
template. But when I created a new VM say vm2 from t1 , it does not assume any 
IP address for eth0 interface. It only has loopback interface address.

The reason seems to be ifcfg-eth0 file in 
/etc/sysconfig/network-scripsts/ directory.

This file is exactly the same in vm1 and vm2 including HWADDR and 
UUID entries. I presume this is the reason vm2 is not being able to get an IP 
address. Any Suggestions? Am I doing it wrong way?


Regards

Adeel




dag.sonst...@shapeblue.com
www.shapeblue.com
53 Chandos Place, Covent Garden, London  WC2N 4HSUK
@shapeblue






dag.sonst...@shapeblue.com
www.shapeblue.com
53 Chandos Place, Covent Garden, London  WC2N 4HSUK
@shapeblue






dag.sonst...@shapeblue.com 
www.shapeblue.com
53 Chandos Place, Covent Garden, London  WC2N 4HSUK
@shapeblue
  
 

Re: My server running Cloudstack-management fails to shutdown properly

[Dag Sonstebo]

Hi Adeel,

This looks like you have an NFS share mounted *on* (rather than *from*) the 
management server – which is something unique in your environment. It could be 
you use autofs for your home folder?
Either way the error is fairly standard – the file system thinks you have a 
file open on the NFS share, it therefore doesn’t allow you to unmount it 
cleanly, leading to the shutdown hang.
To troubleshoot just do a “mount” and see what share/folder is mounted to your 
NFS server.


If this was on your KVM host it would make more sense – since you will be using 
NFS for your secondary storage the KVM host will have this mounted at some 
point, and it may not unmount this cleanly if it thinks there are still files 
in use. If so you need to shut down the KVM cloudstack agent and libvirt before 
unmounting the share.

Regards,
Dag Sonstebo
Cloud Architect
ShapeBlue

On 19/10/2017, 09:00, "Muhammad Adeel Zahid" <16030...@lums.edu.pk> wrote:

Hello,


I am running cloudstack on Cent OS 6 server. When I try to shut down this 
machine, it hangs. It fails to stop the cloudstack-management but it surpasses 
this point. Furthermore, during shutdown, machine hangs with this message


Unmounting NFS filesystem: unmount.nfs: /mnt/some-long-uuid: device is busy


After showing above message, it hangs and I have to power off the system.


How can I get around this problem.


Regards

Adeel




dag.sonst...@shapeblue.com 
www.shapeblue.com
53 Chandos Place, Covent Garden, London  WC2N 4HSUK
@shapeblue
  
 

Re: AW:https Repo links don't work

[Benjamin Naber]

Hi Sudhansu,

where these "systemvm" directorys located ?
cant find them on systemvm template vhd. 

Kind regards

Benjamin
 
> Sudhansu Sahu  hat am 18. Oktober 2017 um 09:34 
> geschrieben:
> 
> 
> Its overridden by jvm argument -Djavax.net.ssl.trustStore 
> (/usr/local/cloud/systemvm/secstorage.sh).
> 
> Thanks
> Sudhansu
> 
> On 10/17/17, 2:34 PM, "Benjamin Naber"  wrote:
> 
> Hi Sudhansu,
> 
> ty for Bug Reporting. You know by which script the realhostip.keystore 
> will be overwriten ?
> 
> I cant find the path or file in the basic systemvm template.
> 
> Kind regards
> 
> benjamin
> 
> > Sudhansu Sahu  hat am 17. Oktober 2017 um 
> 10:51 geschrieben:
> > 
> > I have created below defect for this. 
> https://issues.apache.org/jira/browse/CLOUDSTACK-10112
> > 
> > Sent from my iPhone
> > 
> > On 17-Oct-2017, at 1:56 PM, Benjamin Naber 
> mailto:benjamin.na...@coders-area.de>> wrote:
> > 
> > Hi Sudhansu,
> > 
> > that fix worked for me. thanks for your help.
> > 
> > Kind regards
> > 
> > Benjamin
> > Sudhansu Sahu 
> mailto:sudhansu.s...@accelerite.com>> hat am 
> 17. Oktober 2017 um 10:03 geschrieben:
> > 
> > For realhostip.keystore the password is ‘vmops.com’. 
> and for java keystore the password is ‘changeit’.
> > 
> > Thanks
> > Sudhansu Sahu
> > 
> > On 10/17/17, 1:29 PM, "Benjamin Naber" 
> mailto:benjamin.na...@coders-area.de>> wrote:
> > 
> > Hi Sudhansu,
> > 
> > ty for the tip ! do you know the default password for the keystore on 
> the secondary storage vm ?
> > 
> > Kind regards
> > 
> > Benjamin
> > 
> > Sudhansu Sahu 
> mailto:sudhansu.s...@accelerite.com>> hat am 
> 17. Oktober 2017 um 06:25 geschrieben:
> > 
> > In SSVM default java keystore is overriden by "realhostip.keystore" and 
> does not have necessary certificates.
> > 
> > Import all java cacerts to 
> ‘/usr/local/cloud/systemvm/certs/realhostip.keystore’.
> > 
> > keytool -importkeystore –srckeystore -destkeystore 
> /usr/local/cloud/systemvm/certs/realhostip.keystore
> > 
> > Thanks
> > Sudhansu Sahu
> > 
> > On 10/16/17, 9:25 PM, "Rafael Weingärtner" 
> mailto:raf...@autonomiccs.com.br>> wrote:
> > 
> > Well, if everything is there, then you should not see this error of
> > unable to build the certification path.
> > 
> > Then, it would be a good idea to check which cacerts Java is using.
> > 
> > On 10/16/2017 1:54 PM, Benjamin Naber wrote:
> > 
> > Hi Rafael,
> > 
> > i exported the Keystore to textfile and checked fingerprints.
> > 
> > root CAs also in the Keystore.
> > 
> > Kind Regards
> > 
> > Benjamin
> > 
> > Rafael Weingärtner 
> mailto:raf...@autonomiccs.com.br>> hat am 16. 
> Oktober 2017 um 17:45 geschrieben:
> > 
> > How did you check the certificates? Did you list and checked by name? Or
> > by fingerprint?
> > 
> > Also, did you check if the root CA was there as well? For instance, the
> > Let's encrypt root CA is "DST Root CA X3" (the CA that signs let's
> > encrypt CA's certificate)
> > 
> > On 10/16/2017 1:42 PM, Benjamin Naber wrote:
> > 
> > Hi Rafael,
> > 
> > ive allready checked the Java Keystore. All Certificates are included.
> > 
> > Im using ACS 4.10
> > 
> > Kind Regards
> > 
> > Benjamin
> > 
> > Benjamin Naber 
> mailto:benjamin.na...@coders-area.de>> hat am 
> 16. Oktober 2017 um 17:26 geschrieben:
> > 
> > Hi Rafael,
> > 
> > Currently no ssl Backlund works. Also COMODO Certificates don't work.
> > 
> > Kind Regards
> > 
> > Benjamin
> > 
> > Von meinem Huawei-Mobiltelefon gesendet
> > 
> >  Originalnachricht 
> > Betreff: Re: https Repo links don't work
> > Von: Rafael Weingärtner 
> mailto:raf...@autonomiccs.com.br>>
> > An: users@cloudstack.apache.org
> > Cc:
> > 
> > Probably the let´s encrypt CA is not in your java keystore. You will
> > need to add it.
> > 
> > On 10/16/2017 1:13 PM, Benjamin Naber - NETFORMIC GmbH wrote:
> > 
> > Hi together,
> > 
> > i have an issue deploying templates with https Repo Links.
> > 
> > When i create a global template like debian stretch netinstall and the
> > ISO file is located on a https backlink
> > 
> (https://cdimage.debian.org/debian-cd/current/amd64/iso-cd/debian-9.2.1-amd64-netinst.iso)
> > 
> .
> > 
> > the following error occours:
> > 
> > sun.security.validator.ValidatorException: PKIX path building failed:
> > sun.security.provider.certpath.SunCertPathBuilderException: u

Problem with diskpath after (failed?) Snapshot-Consolidation? (VMWare)

[Markus Kosubek]

Hello List,

we experience a repeating issue in our environment.

- VMWare Vcenter Server: 6.0.0
- cloudstack 4.7.1

In not yet known circumstances* after Snapshot-Consolidation** in the 
Hypervisor, the change of disk-configuration is not recognized by 
cloudstack.


Example:
- Vm runs on a snapshot for a certain period of time, let's say: 
i-99-500-VM-0002.vmdk, like written in cloudstack-db (column 'path' 
in table 'volumes')


- *Consolidation happens*

- VM stops (there might be a second vmware-based problem here because 
this shouldn't happen!)


- VM's root-disk in vmware is now: i-99-500-VM.vmdk

- but cloudstack-db still has path 'i-99-500-VM-0002.vmdk'

- VM won't start again, because cloudstack does not find root-disk.

We can fix this then with updating the db manually, but this is of 
course not what we want long-term.


Does anybody of you have experience with issues of that kind?

Any hints would greatly be appreciated.

Regards
Markus

* probably because consolidation takes too long so that cloudstack 
thinks this task timed out, but in reality on vmware-site, the task 
succeeded at the end


** this happens, if snapshots of a vm exist after a certain period of 
time, i believe. Not quite sure about the trigger for that

Re: Problem with diskpath after (failed?) Snapshot-Consolidation? (VMWare)

[Rene Moser]

Hi Markus

Yes,

We (VMware 5.5 and CS 4.5) experience also such kind of issues related
to path.

As a result "under certain conditions", it can result in a broken VM or
a VM reset to template. We just experienced this today morning with one
of the VMs.

We added a monitoring for this state "path out of sync" and manually
change the path in the DB but I agree, not a real solution.

Regards
René


On 10/19/2017 11:23 AM, Markus Kosubek wrote:
> Hello List,
> 
> we experience a repeating issue in our environment.
> 
> - VMWare Vcenter Server: 6.0.0
> - cloudstack 4.7.1
> 
> In not yet known circumstances* after Snapshot-Consolidation** in the
> Hypervisor, the change of disk-configuration is not recognized by
> cloudstack.
> 
> Example:
> - Vm runs on a snapshot for a certain period of time, let's say:
> i-99-500-VM-0002.vmdk, like written in cloudstack-db (column 'path'
> in table 'volumes')
> 
> - *Consolidation happens*
> 
> - VM stops (there might be a second vmware-based problem here because
> this shouldn't happen!)
> 
> - VM's root-disk in vmware is now: i-99-500-VM.vmdk
> 
> - but cloudstack-db still has path 'i-99-500-VM-0002.vmdk'
> 
> - VM won't start again, because cloudstack does not find root-disk.
> 
> We can fix this then with updating the db manually, but this is of
> course not what we want long-term.
> 
> Does anybody of you have experience with issues of that kind?
> 
> Any hints would greatly be appreciated.
> 
> Regards
> Markus
> 
> * probably because consolidation takes too long so that cloudstack
> thinks this task timed out, but in reality on vmware-site, the task
> succeeded at the end
> 
> ** this happens, if snapshots of a vm exist after a certain period of
> time, i believe. Not quite sure about the trigger for that
> 

Re: My server running Cloudstack-management fails to shutdown properly

[Muhammad Adeel Zahid]

Hi Dag,


Thanks for pointing out that I am using primary and secondary storage on 
machine running the management server. I was just trying to delete the primary 
storage but management UI shows me an error message showing me the following


Cannot delete image store with active templates backup!


Please note that I have deleted all the instances that were using storage.


Furthermore, I was working to set up new primary and secondary storage on kvm 
host. I did directory setup and put the right info in /etc/hosts file but I was 
unable to download the system vm template using following command

/usr/share/cloudstack-common/scripts/storage/secondary/cloud-install-sys-tmplt \
-m /secondary \
-u 
http://cloudstack.apt-get.eu/systemvm/4.6/systemvm64template-4.6.0-kvm.qcow2.bz2
 \
-h kvm -F

The reason of failure, obviously, is that I don't have mysql on hypervisor 
machine
How can I resolve these problems?


Regards
Adeel




From: Dag Sonstebo 
Sent: Thursday, October 19, 2017 1:19:54 PM
To: users@cloudstack.apache.org
Subject: Re: My server running Cloudstack-management fails to shutdown properly

Hi Adeel,

This looks like you have an NFS share mounted *on* (rather than *from*) the 
management server – which is something unique in your environment. It could be 
you use autofs for your home folder?
Either way the error is fairly standard – the file system thinks you have a 
file open on the NFS share, it therefore doesn’t allow you to unmount it 
cleanly, leading to the shutdown hang.
To troubleshoot just do a “mount” and see what share/folder is mounted to your 
NFS server.


If this was on your KVM host it would make more sense – since you will be using 
NFS for your secondary storage the KVM host will have this mounted at some 
point, and it may not unmount this cleanly if it thinks there are still files 
in use. If so you need to shut down the KVM cloudstack agent and libvirt before 
unmounting the share.

Regards,
Dag Sonstebo
Cloud Architect
ShapeBlue

On 19/10/2017, 09:00, "Muhammad Adeel Zahid" <16030...@lums.edu.pk> wrote:

Hello,


I am running cloudstack on Cent OS 6 server. When I try to shut down this 
machine, it hangs. It fails to stop the cloudstack-management but it surpasses 
this point. Furthermore, during shutdown, machine hangs with this message


Unmounting NFS filesystem: unmount.nfs: /mnt/some-long-uuid: device is busy


After showing above message, it hangs and I have to power off the system.


How can I get around this problem.


Regards

Adeel




dag.sonst...@shapeblue.com
www.shapeblue.com
53 Chandos Place, Covent Garden, London  WC2N 4HSUK
@shapeblue

Re: My server running Cloudstack-management fails to shutdown properly

[Muhammad Adeel Zahid]

Is there a way, I can delete everything in cloudstack at once and start afresh?


From: Muhammad Adeel Zahid <16030...@lums.edu.pk>
Sent: Thursday, October 19, 2017 3:10:56 PM
To: users@cloudstack.apache.org
Subject: Re: My server running Cloudstack-management fails to shutdown properly

Hi Dag,


Thanks for pointing out that I am using primary and secondary storage on 
machine running the management server. I was just trying to delete the primary 
storage but management UI shows me an error message showing me the following


Cannot delete image store with active templates backup!


Please note that I have deleted all the instances that were using storage.


Furthermore, I was working to set up new primary and secondary storage on kvm 
host. I did directory setup and put the right info in /etc/hosts file but I was 
unable to download the system vm template using following command

/usr/share/cloudstack-common/scripts/storage/secondary/cloud-install-sys-tmplt \
-m /secondary \
-u 
http://cloudstack.apt-get.eu/systemvm/4.6/systemvm64template-4.6.0-kvm.qcow2.bz2
 \
-h kvm -F

The reason of failure, obviously, is that I don't have mysql on hypervisor 
machine
How can I resolve these problems?


Regards
Adeel




From: Dag Sonstebo 
Sent: Thursday, October 19, 2017 1:19:54 PM
To: users@cloudstack.apache.org
Subject: Re: My server running Cloudstack-management fails to shutdown properly

Hi Adeel,

This looks like you have an NFS share mounted *on* (rather than *from*) the 
management server – which is something unique in your environment. It could be 
you use autofs for your home folder?
Either way the error is fairly standard – the file system thinks you have a 
file open on the NFS share, it therefore doesn’t allow you to unmount it 
cleanly, leading to the shutdown hang.
To troubleshoot just do a “mount” and see what share/folder is mounted to your 
NFS server.


If this was on your KVM host it would make more sense – since you will be using 
NFS for your secondary storage the KVM host will have this mounted at some 
point, and it may not unmount this cleanly if it thinks there are still files 
in use. If so you need to shut down the KVM cloudstack agent and libvirt before 
unmounting the share.

Regards,
Dag Sonstebo
Cloud Architect
ShapeBlue

On 19/10/2017, 09:00, "Muhammad Adeel Zahid" <16030...@lums.edu.pk> wrote:

Hello,


I am running cloudstack on Cent OS 6 server. When I try to shut down this 
machine, it hangs. It fails to stop the cloudstack-management but it surpasses 
this point. Furthermore, during shutdown, machine hangs with this message


Unmounting NFS filesystem: unmount.nfs: /mnt/some-long-uuid: device is busy


After showing above message, it hangs and I have to power off the system.


How can I get around this problem.


Regards

Adeel




dag.sonst...@shapeblue.com
www.shapeblue.com
53 Chandos Place, Covent Garden, London  WC2N 4HSUK
@shapeblue

Re: My server running Cloudstack-management fails to shutdown properly

[Muhammad Adeel Zahid]

I faced too many problems when trying to delete the a zone and adding new one.

For deleting zone-1

I was able to successfully delete instances hosts, cluster, pod and region but 
I was not able to delete primary and secondary storage for the zone. When I 
tried deleting secondary storage, it tells me I can't do so because there are 
templates. When I try to delete templates, I get error message, "Failed to 
delete template/ISO". I saw in log files that this is because no system VM is 
running which is correct.


Adding Zone 2:

I added zone, region, pod, cluster, primary and secondary storage and one host. 
But now system vm is running probably because I get error when CS tries to run 
system vm for zone-1.


I am currently not able to find how to fix this


Regards

Adeel


From: Muhammad Adeel Zahid
Sent: Thursday, October 19, 2017 3:41:45 PM
To: users@cloudstack.apache.org
Subject: Re: My server running Cloudstack-management fails to shutdown properly


Is there a way, I can delete everything in cloudstack at once and start afresh?


From: Muhammad Adeel Zahid <16030...@lums.edu.pk>
Sent: Thursday, October 19, 2017 3:10:56 PM
To: users@cloudstack.apache.org
Subject: Re: My server running Cloudstack-management fails to shutdown properly

Hi Dag,


Thanks for pointing out that I am using primary and secondary storage on 
machine running the management server. I was just trying to delete the primary 
storage but management UI shows me an error message showing me the following


Cannot delete image store with active templates backup!


Please note that I have deleted all the instances that were using storage.


Furthermore, I was working to set up new primary and secondary storage on kvm 
host. I did directory setup and put the right info in /etc/hosts file but I was 
unable to download the system vm template using following command

/usr/share/cloudstack-common/scripts/storage/secondary/cloud-install-sys-tmplt \
-m /secondary \
-u 
http://cloudstack.apt-get.eu/systemvm/4.6/systemvm64template-4.6.0-kvm.qcow2.bz2
 \
-h kvm -F

The reason of failure, obviously, is that I don't have mysql on hypervisor 
machine
How can I resolve these problems?


Regards
Adeel




From: Dag Sonstebo 
Sent: Thursday, October 19, 2017 1:19:54 PM
To: users@cloudstack.apache.org
Subject: Re: My server running Cloudstack-management fails to shutdown properly

Hi Adeel,

This looks like you have an NFS share mounted *on* (rather than *from*) the 
management server – which is something unique in your environment. It could be 
you use autofs for your home folder?
Either way the error is fairly standard – the file system thinks you have a 
file open on the NFS share, it therefore doesn’t allow you to unmount it 
cleanly, leading to the shutdown hang.
To troubleshoot just do a “mount” and see what share/folder is mounted to your 
NFS server.


If this was on your KVM host it would make more sense – since you will be using 
NFS for your secondary storage the KVM host will have this mounted at some 
point, and it may not unmount this cleanly if it thinks there are still files 
in use. If so you need to shut down the KVM cloudstack agent and libvirt before 
unmounting the share.

Regards,
Dag Sonstebo
Cloud Architect
ShapeBlue

On 19/10/2017, 09:00, "Muhammad Adeel Zahid" <16030...@lums.edu.pk> wrote:

Hello,


I am running cloudstack on Cent OS 6 server. When I try to shut down this 
machine, it hangs. It fails to stop the cloudstack-management but it surpasses 
this point. Furthermore, during shutdown, machine hangs with this message


Unmounting NFS filesystem: unmount.nfs: /mnt/some-long-uuid: device is busy


After showing above message, it hangs and I have to power off the system.


How can I get around this problem.


Regards

Adeel




dag.sonst...@shapeblue.com
www.shapeblue.com
53 Chandos Place, Covent Garden, London  WC2N 4HSUK
@shapeblue

Re: VPC across multiple hypevisors

[Chiradeep Vittal]

There is a possibility that you might lose the VR under some circumstances.
E.g., the VR is created on XS and it needs to be recreated ( XS server
dies) but there is no XS capacity left. This might have been fixed, but
best to test these corner cases out.

On Wed, Oct 18, 2017 at 10:35 PM Asanka Gunasekara  wrote:

> Hi All thanks for the reapplies, and Dag too mentioned this is possible.
>
> Thank you and Best Regards
>
> Asanka
>
> On 17 October 2017 at 20:51, Rafael Weingärtner  >
> wrote:
>
> > I believe it is. I have already designed and deployed a cloud environment
> > where we had XenServers and VMware ESXi servers in the same zone. We had
> > cases where the VR was in the VMware cluster and VMs in the XenServer and
> > everything was fine.
> >
> >
> > On 10/17/2017 12:34 PM, Daan Hoogland wrote:
> >
> >> Asanka,
> >> I've not seen that work but it might. I can't think of a hard impediment
> >> but having the two types of hypervisors in the same zone is required. If
> >> all else fails, another solution may be to split the tiers into separate
> >> networks and create a private gateway from the one to the other.
> >> hope you succeed, let us know
> >> Daan
> >>
> >> On Mon, Oct 16, 2017 at 8:24 AM, Asanka Gunasekara 
> >> wrote:
> >>
> >> Thanks for the reply Makrand
> >>>
> >>> We have a requirement, where we need to have Database on Vmware Esxi
> and
> >>> Application server On Xen. I am Applisuer using routing we can connect
> >>> these two together. But is it possible to create a single VPC combining
> >>> these two environments or just to extend the Database VPC to
> application
> >>> server.
> >>>
> >>> Thanks adn regards
> >>>
> >>> Asanka
> >>>
> >>> On 16 October 2017 at 11:47, Makrand  wrote:
> >>>
> >>> Hi,
> 
>  Can you be a bit more descriptive as to what exactly you want to do?
> 
>  Note:-
>  VPC is nothing but multi-tier network (A VPC acts as a container for
>  multiple isolated networks that can communicate with each other via
> its
>  virtual router.) So if you're having VMs running in multi-hypervisor
> 
> >>> setup,
> >>>
>  as long as they are connected to same VR that can work. I haven't
> setup
>  anything like this. Just a guess on my part.
> 
> 
> 
>  --
>  Makrand
> 
> 
>  On Mon, Oct 16, 2017 at 10:37 AM, Asanka Gunasekara  >
>  wrote:
> 
>  Hi Guys, is there a way to do a VPC across multiple hypervisors
> >
> > Thanks and Regards
> >
> > Asanka
> >
> >
> >>
> >>
> > --
> > Rafael Weingärtner
> >
> >
>

Re: VPC across multiple hypevisors

[Rafael Weingärtner]

This would happen if you force a hypervisor type for system vms. If you are
ok with any hypervisor type, then if all XenServer are down, ACS would try
to spin up another in other available host (even if this host uses a
different hypervisor)

On Thu, Oct 19, 2017 at 2:26 PM, Chiradeep Vittal 
wrote:

> There is a possibility that you might lose the VR under some circumstances.
> E.g., the VR is created on XS and it needs to be recreated ( XS server
> dies) but there is no XS capacity left. This might have been fixed, but
> best to test these corner cases out.
>
> On Wed, Oct 18, 2017 at 10:35 PM Asanka Gunasekara 
> wrote:
>
> > Hi All thanks for the reapplies, and Dag too mentioned this is possible.
> >
> > Thank you and Best Regards
> >
> > Asanka
> >
> > On 17 October 2017 at 20:51, Rafael Weingärtner <
> raf...@autonomiccs.com.br
> > >
> > wrote:
> >
> > > I believe it is. I have already designed and deployed a cloud
> environment
> > > where we had XenServers and VMware ESXi servers in the same zone. We
> had
> > > cases where the VR was in the VMware cluster and VMs in the XenServer
> and
> > > everything was fine.
> > >
> > >
> > > On 10/17/2017 12:34 PM, Daan Hoogland wrote:
> > >
> > >> Asanka,
> > >> I've not seen that work but it might. I can't think of a hard
> impediment
> > >> but having the two types of hypervisors in the same zone is required.
> If
> > >> all else fails, another solution may be to split the tiers into
> separate
> > >> networks and create a private gateway from the one to the other.
> > >> hope you succeed, let us know
> > >> Daan
> > >>
> > >> On Mon, Oct 16, 2017 at 8:24 AM, Asanka Gunasekara 
> > >> wrote:
> > >>
> > >> Thanks for the reply Makrand
> > >>>
> > >>> We have a requirement, where we need to have Database on Vmware Esxi
> > and
> > >>> Application server On Xen. I am Applisuer using routing we can
> connect
> > >>> these two together. But is it possible to create a single VPC
> combining
> > >>> these two environments or just to extend the Database VPC to
> > application
> > >>> server.
> > >>>
> > >>> Thanks adn regards
> > >>>
> > >>> Asanka
> > >>>
> > >>> On 16 October 2017 at 11:47, Makrand  wrote:
> > >>>
> > >>> Hi,
> > 
> >  Can you be a bit more descriptive as to what exactly you want to do?
> > 
> >  Note:-
> >  VPC is nothing but multi-tier network (A VPC acts as a container for
> >  multiple isolated networks that can communicate with each other via
> > its
> >  virtual router.) So if you're having VMs running in multi-hypervisor
> > 
> > >>> setup,
> > >>>
> >  as long as they are connected to same VR that can work. I haven't
> > setup
> >  anything like this. Just a guess on my part.
> > 
> > 
> > 
> >  --
> >  Makrand
> > 
> > 
> >  On Mon, Oct 16, 2017 at 10:37 AM, Asanka Gunasekara <
> asa...@nimbus.lk
> > >
> >  wrote:
> > 
> >  Hi Guys, is there a way to do a VPC across multiple hypervisors
> > >
> > > Thanks and Regards
> > >
> > > Asanka
> > >
> > >
> > >>
> > >>
> > > --
> > > Rafael Weingärtner
> > >
> > >
> >
>



-- 
Rafael Weingärtner

Iptables of Host machine blocks http traffic to Guest VM's

[Muhammad Adeel Zahid]

Hello,


Finally, I have created a template from centos 6.8 ISO with jdk 1.8 installed 
and a web application hosted. I can now create VM's from this templates and 
they work fine except one problem. The web applications in guest VMs created 
from template are only accessible from the host running the VMs. If I access 
them (web applications) from some other system on the same LAN they are not 
accessible until I turn off the iptables service on host machine. Is there an 
ip table rule that I can add to work around this problem not only for the 
existing VMs but for the VMs I will be creating on this host in the future?


Regards

Adeel

Re: Iptables of Host machine blocks http traffic to Guest VM's

[Muhammad Adeel Zahid]

One more finding. Even restarting the iptables service does the trick. Why is 
that? What can I do to avoid having to restart the iptables service?



Regards

Adeel


From: Muhammad Adeel Zahid <16030...@lums.edu.pk>
Sent: Friday, October 20, 2017 12:08:43 AM
To: users@cloudstack.apache.org
Subject: Iptables of Host machine blocks http traffic to Guest VM's

Hello,


Finally, I have created a template from centos 6.8 ISO with jdk 1.8 installed 
and a web application hosted. I can now create VM's from this templates and 
they work fine except one problem. The web applications in guest VMs created 
from template are only accessible from the host running the VMs. If I access 
them (web applications) from some other system on the same LAN they are not 
accessible until I turn off the iptables service on host machine. Is there an 
ip table rule that I can add to work around this problem not only for the 
existing VMs but for the VMs I will be creating on this host in the future?


Regards

Adeel

Re: Iptables of Host machine blocks http traffic to Guest VM's

[Rafael Weingärtner]

What type of deployment are you using?
Did you try configuring the ingress/egress rules for the network of the VM
you are creating?

On Thu, Oct 19, 2017 at 5:17 PM, Muhammad Adeel Zahid <16030...@lums.edu.pk>
wrote:

> One more finding. Even restarting the iptables service does the trick. Why
> is that? What can I do to avoid having to restart the iptables service?
>
>
>
> Regards
>
> Adeel
>
> 
> From: Muhammad Adeel Zahid <16030...@lums.edu.pk>
> Sent: Friday, October 20, 2017 12:08:43 AM
> To: users@cloudstack.apache.org
> Subject: Iptables of Host machine blocks http traffic to Guest VM's
>
> Hello,
>
>
> Finally, I have created a template from centos 6.8 ISO with jdk 1.8
> installed and a web application hosted. I can now create VM's from this
> templates and they work fine except one problem. The web applications in
> guest VMs created from template are only accessible from the host running
> the VMs. If I access them (web applications) from some other system on the
> same LAN they are not accessible until I turn off the iptables service on
> host machine. Is there an ip table rule that I can add to work around this
> problem not only for the existing VMs but for the VMs I will be creating on
> this host in the future?
>
>
> Regards
>
> Adeel
>
>


-- 
Rafael Weingärtner

Re: Iptables of Host machine blocks http traffic to Guest VM's

[Muhammad Adeel Zahid]

I am using KVM as hypervisor and for network configuration I followed the 
following link

http://docs.cloudstack.apache.org/projects/cloudstack-installation/en/4.6/qig.html


Except that my ip addresses are from 10.0.0.x series. And I do not know about 
ingress/egress rules and how to configure them.

Quick Installation Guide for CentOS 6 — Apache CloudStack 
...
docs.cloudstack.apache.org
High level overview of the process¶ This runbook will focus on building a 
CloudStack cloud using KVM on CentOS 6.5 with NFS storage on a flat layer-2 
network ...





From: Rafael Weingärtner 
Sent: Friday, October 20, 2017 12:19:49 AM
To: users@cloudstack.apache.org
Subject: Re: Iptables of Host machine blocks http traffic to Guest VM's

What type of deployment are you using?
Did you try configuring the ingress/egress rules for the network of the VM
you are creating?

On Thu, Oct 19, 2017 at 5:17 PM, Muhammad Adeel Zahid <16030...@lums.edu.pk>
wrote:

> One more finding. Even restarting the iptables service does the trick. Why
> is that? What can I do to avoid having to restart the iptables service?
>
>
>
> Regards
>
> Adeel
>
> 
> From: Muhammad Adeel Zahid <16030...@lums.edu.pk>
> Sent: Friday, October 20, 2017 12:08:43 AM
> To: users@cloudstack.apache.org
> Subject: Iptables of Host machine blocks http traffic to Guest VM's
>
> Hello,
>
>
> Finally, I have created a template from centos 6.8 ISO with jdk 1.8
> installed and a web application hosted. I can now create VM's from this
> templates and they work fine except one problem. The web applications in
> guest VMs created from template are only accessible from the host running
> the VMs. If I access them (web applications) from some other system on the
> same LAN they are not accessible until I turn off the iptables service on
> host machine. Is there an ip table rule that I can add to work around this
> problem not only for the existing VMs but for the VMs I will be creating on
> this host in the future?
>
>
> Regards
>
> Adeel
>
>


--
Rafael Weingärtner

Re: Iptables of Host machine blocks http traffic to Guest VM's

[Rafael Weingärtner]

This process of enabling ingress/egress traffic is from a user perspective.
You should bear in mind that this is an IaaS orchestration system, and as
such the user that is allocating and consuming resources has to be able to
do such configurations.

The link you are following is from the installation manual. It presents the
requirements and basic design and installation guidelines. However, that
link is not meant to show how to use Apache CloudStack (ACS).

I do not know much user manuals, but you can find one here [1] that is
developed by Leaseweb. I find it quite detailed and will be worth for you
at the begging (learning how to get around ACS UI and API). The Ingress and
Egress rules that I am talking about are discussed in Step 3 of [1].

[1] https://kb.leaseweb.com/display/KB/Getting+Started%3A+CloudStack

On Thu, Oct 19, 2017 at 5:32 PM, Muhammad Adeel Zahid <16030...@lums.edu.pk>
wrote:

> I am using KVM as hypervisor and for network configuration I followed the
> following link
>
> http://docs.cloudstack.apache.org/projects/cloudstack-
> installation/en/4.6/qig.html
>
>
> Except that my ip addresses are from 10.0.0.x series. And I do not know
> about ingress/egress rules and how to configure them.
>
> Quick Installation Guide for CentOS 6 — Apache CloudStack ...<
> http://docs.cloudstack.apache.org/projects/cloudstack-installation/en/4.
> 6/qig.html>
> docs.cloudstack.apache.org
> High level overview of the process¶ This runbook will focus on building a
> CloudStack cloud using KVM on CentOS 6.5 with NFS storage on a flat layer-2
> network ...
>
>
>
>
> 
> From: Rafael Weingärtner 
> Sent: Friday, October 20, 2017 12:19:49 AM
> To: users@cloudstack.apache.org
> Subject: Re: Iptables of Host machine blocks http traffic to Guest VM's
>
> What type of deployment are you using?
> Did you try configuring the ingress/egress rules for the network of the VM
> you are creating?
>
> On Thu, Oct 19, 2017 at 5:17 PM, Muhammad Adeel Zahid <
> 16030...@lums.edu.pk>
> wrote:
>
> > One more finding. Even restarting the iptables service does the trick.
> Why
> > is that? What can I do to avoid having to restart the iptables service?
> >
> >
> >
> > Regards
> >
> > Adeel
> >
> > 
> > From: Muhammad Adeel Zahid <16030...@lums.edu.pk>
> > Sent: Friday, October 20, 2017 12:08:43 AM
> > To: users@cloudstack.apache.org
> > Subject: Iptables of Host machine blocks http traffic to Guest VM's
> >
> > Hello,
> >
> >
> > Finally, I have created a template from centos 6.8 ISO with jdk 1.8
> > installed and a web application hosted. I can now create VM's from this
> > templates and they work fine except one problem. The web applications in
> > guest VMs created from template are only accessible from the host running
> > the VMs. If I access them (web applications) from some other system on
> the
> > same LAN they are not accessible until I turn off the iptables service on
> > host machine. Is there an ip table rule that I can add to work around
> this
> > problem not only for the existing VMs but for the VMs I will be creating
> on
> > this host in the future?
> >
> >
> > Regards
> >
> > Adeel
> >
> >
>
>
> --
> Rafael Weingärtner
>



-- 
Rafael Weingärtner

Rsize / Wsize configuration for NFS share

[Ivan Kudryavtsev]

Hello, friends.

I'm planning to deploy new cluster with KVM and shared NFS storage soon.
Right now I already have such deploys that operate fine. Currently, I'm
trying to compare different storage options for new cluster between
- LVM+EXT4+HWRAID+BBU,
- LVM+EXT4+MDADM and
- ZFS.

HW setup is:
- 5 x Micron 5100PRO 1.9TB
- LSI9300-8i HBA or LSI9266-8i+BBU
- Cisco UCS server w/ 24x2.5" / 2xE5-2650 / 64GB RAM

I have got competitive performance results between them locally already,
but now I need to test over NFS. I'm pretty sure that first two options
will operate nice with ACS default NFS mount args (because I already have
such cases in prod), but ZFS is quite smart thing, so I started to
investigate how to change NFS client mount options and unfortunately
haven't succeed defining the proper place where cloudstack agent determines
how to mount share and what args to use. I read a lot of ZFS-related
articles and people write rsize/wsize affect quite much, so I wonder how to
instruct cloudstack agent to use specific rsize/wsize args to mount primary
storage.

Also, I haven't found in ACS archives mentions about ZFS NFS share, so
might be it's a bad case for ACS because of Qcow image format?, but I think
it could be a good one so want to test personally.

Any suggestions are welcome.

-- 
With best regards, Ivan Kudryavtsev
Bitworks Software, Ltd.
Cell: +7-923-414-1515
WWW: http://bitworks.software/