You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@cloudstack.apache.org by Pearl d'Silva <pe...@shapeblue.com> on 2021/04/09 12:57:27 UTC

Re: [PROPOSAL] Force option and/or command to the detach ISO operation for running VMs

I've seen this issue happen on VMware.
+1 on having a force option to the attach/detach iso operations, thus giving control to the users as opposed to hard-coding the answer in the code.

Regards,
Pearl
________________________________
From: Nicolas Vazquez <Ni...@shapeblue.com>
Sent: Tuesday, March 30, 2021 4:06 AM
To: users@cloudstack.apache.org <us...@cloudstack.apache.org>
Subject: Re: [PROPOSAL] Force option and/or command to the detach ISO operation for running VMs

Hi Alireza,

Yes - I have seen the same. In my tests that VM question was the one with ID 'msg.cdromdisconnect.locked' to which CloudStack always answered "no" to that question and there was no option to control which answer to send.

I also noticed that the answering that question is only handled in the Vmware codebase, but not on other hypervisors codebase.


Regards,

Nicolas Vazquez

________________________________
From: Alireza Eskandari <as...@gmail.com>
Sent: Monday, March 29, 2021 3:40 PM
To: users@cloudstack.apache.org <us...@cloudstack.apache.org>
Subject: Re: [PROPOSAL] Force option and/or command to the detach ISO operation for running VMs

Hi Nicolas
AFAIK in these situations ESXi will raise a question about detaching iso
from VM.
I have faced with similar problem in this issue:
https://github.com/apache/cloudstack/issues/4808
I got the same error in vCenter. Before waiting for timeout, if you click
on the refresh button in the top right corner of the web gui, you can see
an unanswered question is raised for the VM.
Regards

On Mon, Mar 29, 2021 at 7:34 PM Nicolas Vazquez <
Nicolas.Vazquez@shapeblue.com> wrote:

> Hi all,
>
> I have observed a non-consistent issue on the detach ISO operation for
> running VMs. Some guest OS refuse to release an attached ISO even though
> the ISO is not mounted, making the detach ISO operation fail if the VM is
> running. For example, a CentOS 7 running VM on Vmware 6.5 fails to detach
> an ISO with error: 'Connection control operation failed for disk 'ide0:0'
> if the VM is running'. However, there is no issue if the VM is stopped.
>
> As a workaround executing an 'eject' command on the guest VM before the
> detach operation solves the problem. The detach ISO operation in CloudStack
> could be extended to support a 'force' option - which could perform a
> similar workaround when the VM is running.
>
> I would be interested to know if you have experimented similar issues on
> different hypervisors or could be some hypervisor specific version issue.
>
>
> Regards,
>
> Nicolas Vazquez
>
> Nicolas.Vazquez@shapeblue.com
> www.shapeblue.com<http://www.shapeblue.com>
> 3 London Bridge Street,  3rd floor, News Building, London  SE1 9SGUK
> @shapeblue
>
>
>
>

Nicolas.Vazquez@shapeblue.com
www.shapeblue.com<http://www.shapeblue.com>
3 London Bridge Street,  3rd floor, News Building, London  SE1 9SGUK
@shapeblue




pearl.dsilva@shapeblue.com 
www.shapeblue.com
3 London Bridge Street,  3rd floor, News Building, London  SE1 9SGUK
@shapeblue
  
 


Re: [PROPOSAL] Force option and/or command to the detach ISO operation for running VMs

Posted by Rohit Yadav <ro...@shapeblue.com>.
+1 makes sense - good approach to maintain backward compatibility but have an option to force eject. Does kvm/xenserver/xcp also support force ejecting/detaching of ISOs like VMware?


Regards.

________________________________
From: Harikrishna Patnala <Ha...@shapeblue.com>
Sent: Monday, April 12, 2021 11:30
To: users@cloudstack.apache.org <us...@cloudstack.apache.org>; dev@cloudstack.apache.org <de...@cloudstack.apache.org>
Subject: Re: [PROPOSAL] Force option and/or command to the detach ISO operation for running VMs

Yes, it's better to keep an optional flag in the API params to decide on force detach of ISO instead of hardcoding force in the final detach operation. Thats gives flexibility to users and also keeps backward compatibility.

+1

Regards,
Harikrishna
________________________________
From: Pearl d'Silva <pe...@shapeblue.com>
Sent: Friday, April 9, 2021 6:27 PM
To: users@cloudstack.apache.org <us...@cloudstack.apache.org>; dev@cloudstack.apache.org <de...@cloudstack.apache.org>
Subject: Re: [PROPOSAL] Force option and/or command to the detach ISO operation for running VMs

I've seen this issue happen on VMware.
+1 on having a force option to the attach/detach iso operations, thus giving control to the users as opposed to hard-coding the answer in the code.

Regards,
Pearl
________________________________
From: Nicolas Vazquez <Ni...@shapeblue.com>
Sent: Tuesday, March 30, 2021 4:06 AM
To: users@cloudstack.apache.org <us...@cloudstack.apache.org>
Subject: Re: [PROPOSAL] Force option and/or command to the detach ISO operation for running VMs

Hi Alireza,

Yes - I have seen the same. In my tests that VM question was the one with ID 'msg.cdromdisconnect.locked' to which CloudStack always answered "no" to that question and there was no option to control which answer to send.

I also noticed that the answering that question is only handled in the Vmware codebase, but not on other hypervisors codebase.


Regards,

Nicolas Vazquez

________________________________
From: Alireza Eskandari <as...@gmail.com>
Sent: Monday, March 29, 2021 3:40 PM
To: users@cloudstack.apache.org <us...@cloudstack.apache.org>
Subject: Re: [PROPOSAL] Force option and/or command to the detach ISO operation for running VMs

Hi Nicolas
AFAIK in these situations ESXi will raise a question about detaching iso
from VM.
I have faced with similar problem in this issue:
https://github.com/apache/cloudstack/issues/4808
I got the same error in vCenter. Before waiting for timeout, if you click
on the refresh button in the top right corner of the web gui, you can see
an unanswered question is raised for the VM.
Regards

On Mon, Mar 29, 2021 at 7:34 PM Nicolas Vazquez <
Nicolas.Vazquez@shapeblue.com> wrote:

> Hi all,
>
> I have observed a non-consistent issue on the detach ISO operation for
> running VMs. Some guest OS refuse to release an attached ISO even though
> the ISO is not mounted, making the detach ISO operation fail if the VM is
> running. For example, a CentOS 7 running VM on Vmware 6.5 fails to detach
> an ISO with error: 'Connection control operation failed for disk 'ide0:0'
> if the VM is running'. However, there is no issue if the VM is stopped.
>
> As a workaround executing an 'eject' command on the guest VM before the
> detach operation solves the problem. The detach ISO operation in CloudStack
> could be extended to support a 'force' option - which could perform a
> similar workaround when the VM is running.
>
> I would be interested to know if you have experimented similar issues on
> different hypervisors or could be some hypervisor specific version issue.
>
>
> Regards,
>
> Nicolas Vazquez
>
> Nicolas.Vazquez@shapeblue.com
> www.shapeblue.com<http://www.shapeblue.com>
> 3 London Bridge Street,  3rd floor, News Building, London  SE1 9SGUK
> @shapeblue
>
>
>
>

Nicolas.Vazquez@shapeblue.com
www.shapeblue.com<http://www.shapeblue.com>
3 London Bridge Street,  3rd floor, News Building, London  SE1 9SGUK
@shapeblue




pearl.dsilva@shapeblue.com
www.shapeblue.com<http://www.shapeblue.com>
3 London Bridge Street,  3rd floor, News Building, London  SE1 9SGUK
@shapeblue




Harikrishna.Patnala@shapeblue.com
www.shapeblue.com<http://www.shapeblue.com>
3 London Bridge Street,  3rd floor, News Building, London  SE1 9SG
@shapeblue




rohit.yadav@shapeblue.com 
www.shapeblue.com
3 London Bridge Street,  3rd floor, News Building, London  SE1 9SGUK
@shapeblue
  
 


Re: [PROPOSAL] Force option and/or command to the detach ISO operation for running VMs

Posted by Rohit Yadav <ro...@shapeblue.com>.
+1 makes sense - good approach to maintain backward compatibility but have an option to force eject. Does kvm/xenserver/xcp also support force ejecting/detaching of ISOs like VMware?


Regards.

________________________________
From: Harikrishna Patnala <Ha...@shapeblue.com>
Sent: Monday, April 12, 2021 11:30
To: users@cloudstack.apache.org <us...@cloudstack.apache.org>; dev@cloudstack.apache.org <de...@cloudstack.apache.org>
Subject: Re: [PROPOSAL] Force option and/or command to the detach ISO operation for running VMs

Yes, it's better to keep an optional flag in the API params to decide on force detach of ISO instead of hardcoding force in the final detach operation. Thats gives flexibility to users and also keeps backward compatibility.

+1

Regards,
Harikrishna
________________________________
From: Pearl d'Silva <pe...@shapeblue.com>
Sent: Friday, April 9, 2021 6:27 PM
To: users@cloudstack.apache.org <us...@cloudstack.apache.org>; dev@cloudstack.apache.org <de...@cloudstack.apache.org>
Subject: Re: [PROPOSAL] Force option and/or command to the detach ISO operation for running VMs

I've seen this issue happen on VMware.
+1 on having a force option to the attach/detach iso operations, thus giving control to the users as opposed to hard-coding the answer in the code.

Regards,
Pearl
________________________________
From: Nicolas Vazquez <Ni...@shapeblue.com>
Sent: Tuesday, March 30, 2021 4:06 AM
To: users@cloudstack.apache.org <us...@cloudstack.apache.org>
Subject: Re: [PROPOSAL] Force option and/or command to the detach ISO operation for running VMs

Hi Alireza,

Yes - I have seen the same. In my tests that VM question was the one with ID 'msg.cdromdisconnect.locked' to which CloudStack always answered "no" to that question and there was no option to control which answer to send.

I also noticed that the answering that question is only handled in the Vmware codebase, but not on other hypervisors codebase.


Regards,

Nicolas Vazquez

________________________________
From: Alireza Eskandari <as...@gmail.com>
Sent: Monday, March 29, 2021 3:40 PM
To: users@cloudstack.apache.org <us...@cloudstack.apache.org>
Subject: Re: [PROPOSAL] Force option and/or command to the detach ISO operation for running VMs

Hi Nicolas
AFAIK in these situations ESXi will raise a question about detaching iso
from VM.
I have faced with similar problem in this issue:
https://github.com/apache/cloudstack/issues/4808
I got the same error in vCenter. Before waiting for timeout, if you click
on the refresh button in the top right corner of the web gui, you can see
an unanswered question is raised for the VM.
Regards

On Mon, Mar 29, 2021 at 7:34 PM Nicolas Vazquez <
Nicolas.Vazquez@shapeblue.com> wrote:

> Hi all,
>
> I have observed a non-consistent issue on the detach ISO operation for
> running VMs. Some guest OS refuse to release an attached ISO even though
> the ISO is not mounted, making the detach ISO operation fail if the VM is
> running. For example, a CentOS 7 running VM on Vmware 6.5 fails to detach
> an ISO with error: 'Connection control operation failed for disk 'ide0:0'
> if the VM is running'. However, there is no issue if the VM is stopped.
>
> As a workaround executing an 'eject' command on the guest VM before the
> detach operation solves the problem. The detach ISO operation in CloudStack
> could be extended to support a 'force' option - which could perform a
> similar workaround when the VM is running.
>
> I would be interested to know if you have experimented similar issues on
> different hypervisors or could be some hypervisor specific version issue.
>
>
> Regards,
>
> Nicolas Vazquez
>
> Nicolas.Vazquez@shapeblue.com
> www.shapeblue.com<http://www.shapeblue.com>
> 3 London Bridge Street,  3rd floor, News Building, London  SE1 9SGUK
> @shapeblue
>
>
>
>

Nicolas.Vazquez@shapeblue.com
www.shapeblue.com<http://www.shapeblue.com>
3 London Bridge Street,  3rd floor, News Building, London  SE1 9SGUK
@shapeblue




pearl.dsilva@shapeblue.com
www.shapeblue.com<http://www.shapeblue.com>
3 London Bridge Street,  3rd floor, News Building, London  SE1 9SGUK
@shapeblue




Harikrishna.Patnala@shapeblue.com
www.shapeblue.com<http://www.shapeblue.com>
3 London Bridge Street,  3rd floor, News Building, London  SE1 9SG
@shapeblue




rohit.yadav@shapeblue.com 
www.shapeblue.com
3 London Bridge Street,  3rd floor, News Building, London  SE1 9SGUK
@shapeblue
  
 


Re: [PROPOSAL] Force option and/or command to the detach ISO operation for running VMs

Posted by Harikrishna Patnala <Ha...@shapeblue.com>.
Yes, it's better to keep an optional flag in the API params to decide on force detach of ISO instead of hardcoding force in the final detach operation. Thats gives flexibility to users and also keeps backward compatibility.

+1

Regards,
Harikrishna
________________________________
From: Pearl d'Silva <pe...@shapeblue.com>
Sent: Friday, April 9, 2021 6:27 PM
To: users@cloudstack.apache.org <us...@cloudstack.apache.org>; dev@cloudstack.apache.org <de...@cloudstack.apache.org>
Subject: Re: [PROPOSAL] Force option and/or command to the detach ISO operation for running VMs

I've seen this issue happen on VMware.
+1 on having a force option to the attach/detach iso operations, thus giving control to the users as opposed to hard-coding the answer in the code.

Regards,
Pearl
________________________________
From: Nicolas Vazquez <Ni...@shapeblue.com>
Sent: Tuesday, March 30, 2021 4:06 AM
To: users@cloudstack.apache.org <us...@cloudstack.apache.org>
Subject: Re: [PROPOSAL] Force option and/or command to the detach ISO operation for running VMs

Hi Alireza,

Yes - I have seen the same. In my tests that VM question was the one with ID 'msg.cdromdisconnect.locked' to which CloudStack always answered "no" to that question and there was no option to control which answer to send.

I also noticed that the answering that question is only handled in the Vmware codebase, but not on other hypervisors codebase.


Regards,

Nicolas Vazquez

________________________________
From: Alireza Eskandari <as...@gmail.com>
Sent: Monday, March 29, 2021 3:40 PM
To: users@cloudstack.apache.org <us...@cloudstack.apache.org>
Subject: Re: [PROPOSAL] Force option and/or command to the detach ISO operation for running VMs

Hi Nicolas
AFAIK in these situations ESXi will raise a question about detaching iso
from VM.
I have faced with similar problem in this issue:
https://github.com/apache/cloudstack/issues/4808
I got the same error in vCenter. Before waiting for timeout, if you click
on the refresh button in the top right corner of the web gui, you can see
an unanswered question is raised for the VM.
Regards

On Mon, Mar 29, 2021 at 7:34 PM Nicolas Vazquez <
Nicolas.Vazquez@shapeblue.com> wrote:

> Hi all,
>
> I have observed a non-consistent issue on the detach ISO operation for
> running VMs. Some guest OS refuse to release an attached ISO even though
> the ISO is not mounted, making the detach ISO operation fail if the VM is
> running. For example, a CentOS 7 running VM on Vmware 6.5 fails to detach
> an ISO with error: 'Connection control operation failed for disk 'ide0:0'
> if the VM is running'. However, there is no issue if the VM is stopped.
>
> As a workaround executing an 'eject' command on the guest VM before the
> detach operation solves the problem. The detach ISO operation in CloudStack
> could be extended to support a 'force' option - which could perform a
> similar workaround when the VM is running.
>
> I would be interested to know if you have experimented similar issues on
> different hypervisors or could be some hypervisor specific version issue.
>
>
> Regards,
>
> Nicolas Vazquez
>
> Nicolas.Vazquez@shapeblue.com
> www.shapeblue.com<http://www.shapeblue.com>
> 3 London Bridge Street,  3rd floor, News Building, London  SE1 9SGUK
> @shapeblue
>
>
>
>

Nicolas.Vazquez@shapeblue.com
www.shapeblue.com<http://www.shapeblue.com>
3 London Bridge Street,  3rd floor, News Building, London  SE1 9SGUK
@shapeblue




pearl.dsilva@shapeblue.com
www.shapeblue.com<http://www.shapeblue.com>
3 London Bridge Street,  3rd floor, News Building, London  SE1 9SGUK
@shapeblue




Harikrishna.Patnala@shapeblue.com 
www.shapeblue.com
3 London Bridge Street,  3rd floor, News Building, London  SE1 9SG
@shapeblue
  
 


Re: [PROPOSAL] Force option and/or command to the detach ISO operation for running VMs

Posted by Harikrishna Patnala <Ha...@shapeblue.com>.
Yes, it's better to keep an optional flag in the API params to decide on force detach of ISO instead of hardcoding force in the final detach operation. Thats gives flexibility to users and also keeps backward compatibility.

+1

Regards,
Harikrishna
________________________________
From: Pearl d'Silva <pe...@shapeblue.com>
Sent: Friday, April 9, 2021 6:27 PM
To: users@cloudstack.apache.org <us...@cloudstack.apache.org>; dev@cloudstack.apache.org <de...@cloudstack.apache.org>
Subject: Re: [PROPOSAL] Force option and/or command to the detach ISO operation for running VMs

I've seen this issue happen on VMware.
+1 on having a force option to the attach/detach iso operations, thus giving control to the users as opposed to hard-coding the answer in the code.

Regards,
Pearl
________________________________
From: Nicolas Vazquez <Ni...@shapeblue.com>
Sent: Tuesday, March 30, 2021 4:06 AM
To: users@cloudstack.apache.org <us...@cloudstack.apache.org>
Subject: Re: [PROPOSAL] Force option and/or command to the detach ISO operation for running VMs

Hi Alireza,

Yes - I have seen the same. In my tests that VM question was the one with ID 'msg.cdromdisconnect.locked' to which CloudStack always answered "no" to that question and there was no option to control which answer to send.

I also noticed that the answering that question is only handled in the Vmware codebase, but not on other hypervisors codebase.


Regards,

Nicolas Vazquez

________________________________
From: Alireza Eskandari <as...@gmail.com>
Sent: Monday, March 29, 2021 3:40 PM
To: users@cloudstack.apache.org <us...@cloudstack.apache.org>
Subject: Re: [PROPOSAL] Force option and/or command to the detach ISO operation for running VMs

Hi Nicolas
AFAIK in these situations ESXi will raise a question about detaching iso
from VM.
I have faced with similar problem in this issue:
https://github.com/apache/cloudstack/issues/4808
I got the same error in vCenter. Before waiting for timeout, if you click
on the refresh button in the top right corner of the web gui, you can see
an unanswered question is raised for the VM.
Regards

On Mon, Mar 29, 2021 at 7:34 PM Nicolas Vazquez <
Nicolas.Vazquez@shapeblue.com> wrote:

> Hi all,
>
> I have observed a non-consistent issue on the detach ISO operation for
> running VMs. Some guest OS refuse to release an attached ISO even though
> the ISO is not mounted, making the detach ISO operation fail if the VM is
> running. For example, a CentOS 7 running VM on Vmware 6.5 fails to detach
> an ISO with error: 'Connection control operation failed for disk 'ide0:0'
> if the VM is running'. However, there is no issue if the VM is stopped.
>
> As a workaround executing an 'eject' command on the guest VM before the
> detach operation solves the problem. The detach ISO operation in CloudStack
> could be extended to support a 'force' option - which could perform a
> similar workaround when the VM is running.
>
> I would be interested to know if you have experimented similar issues on
> different hypervisors or could be some hypervisor specific version issue.
>
>
> Regards,
>
> Nicolas Vazquez
>
> Nicolas.Vazquez@shapeblue.com
> www.shapeblue.com<http://www.shapeblue.com>
> 3 London Bridge Street,  3rd floor, News Building, London  SE1 9SGUK
> @shapeblue
>
>
>
>

Nicolas.Vazquez@shapeblue.com
www.shapeblue.com<http://www.shapeblue.com>
3 London Bridge Street,  3rd floor, News Building, London  SE1 9SGUK
@shapeblue




pearl.dsilva@shapeblue.com
www.shapeblue.com<http://www.shapeblue.com>
3 London Bridge Street,  3rd floor, News Building, London  SE1 9SGUK
@shapeblue




Harikrishna.Patnala@shapeblue.com 
www.shapeblue.com
3 London Bridge Street,  3rd floor, News Building, London  SE1 9SG
@shapeblue