Aleksandar Ivanisevic | 2 May 09:27 2011
X-Face
Picon

Re: glusterfs/moosefs/etc as openvz backend?

Toby Burress <tdb@...> writes:

> You have to disable ACLs, and don't squash root, but other than that it
> seems to work fine.  It's not a distributed file system, but my guess
> is if the file system itself is working properly, a VE will probably
> run on top of it with little fuss.  There might be issues with POSIX
> semantics, though, that these file systems (which I've never used)
> may not honor.

Thats what I'm trying to do too, and those were exactly my worries. I
know that it might work with NFS, but I want something more flexible,
so I'm looking into distributed filesystems.

The problem is that I'm not satisfied with daily or hourly rsnapshot
backup, I want to have the replica as close to the original as
possible. We now replicate using DRBD but it introduces high latency
and it becomes really messy once you have more than 2 nodes (ganeti
might help with this, but it was pretty imature on RHEL last time I
looked). 

One alternative is also lsyncd but the real deal would be a real
distributed file system where everything happens in the background.
knawnd | 3 May 14:17 2011
Picon

custom location for OS template

Hi!

Almost a month ago (on April 4th) I had posted a message [1] to 
Discussion thread on OpenVZ forum but I still haven't got reply. So I am 
trying to ask in OpenVZ mail list.
--------------- start quotation ---------------
Is there any way to specify a location for os templates? The default one 
is written in /etc/vz/vz.conf as
template=/vz/template

but I would like to have a possibility to specify the custom one e.g. 
with absolute path:
$ vzctl create --ostemplate /vz/tmpl/centos-5-x86.tar.gz <CTID>

with corresponding value for ostemplate variable in VE config:
ostemplate="/vz/tmpl/centos-5-x86.tar.gz"

I think that way would provide more flexibility.

Should I submit a feature request on that?
--------------- end quotation ---------------

Thanks!
Nikolay.

[1] 
http://forum.openvz.org/index.php?t=msg&goto=42369&&srch=os+template+location#msg_42369 
shule ney | 3 May 21:58 2011
Picon

HELP:How can one VM interrupt another running VM ??

Hi all:
I was wondering how can I use one VM to interrupt another running VM, is there any schedular surpport this function, or any control I can use to implement it. For OpenVZ, is it possible to achieve this? 
_______________________________________________
Users mailing list
Users@...
https://openvz.org/mailman/listinfo/users
Steven Crothers | 3 May 22:03 2011
Picon

Re: HELP:How can one VM interrupt another running VM ??

Why does this request sound malicious?

On Tue, May 3, 2011 at 3:58 PM, shule ney <neyshule@...> wrote:
> Hi all:
> I was wondering how can I use one VM to interrupt another running VM, is
> there any schedular surpport this function, or any control I can use
> to implement it. For OpenVZ, is it possible to achieve this?
> _______________________________________________
> Users mailing list
> Users@...
> https://openvz.org/mailman/listinfo/users
>
>

--

-- 
Steven Crothers
steven.crothers@...
shule ney | 3 May 22:08 2011
Picon

Re: HELP:How can one VM interrupt another running VM ??

I can't quite understand you, why you mean it's "malicious"?

2011/5/3 Steven Crothers <steven.crothers-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
Why does this request sound malicious?

On Tue, May 3, 2011 at 3:58 PM, shule ney <neyshule-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
> Hi all:
> I was wondering how can I use one VM to interrupt another running VM, is
> there any schedular surpport this function, or any control I can use
> to implement it. For OpenVZ, is it possible to achieve this?
> _______________________________________________
> Users mailing list
> Users-GEFAQzZX7r8dnm+yROfE0A@public.gmane.org
> https://openvz.org/mailman/listinfo/users
>
>



--
Steven Crothers
steven.crothers-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org

_______________________________________________
Users mailing list
Users-GEFAQzZX7r8dnm+yROfE0A@public.gmane.org
https://openvz.org/mailman/listinfo/users

_______________________________________________
Users mailing list
Users@...
https://openvz.org/mailman/listinfo/users
Daniel Pittman | 3 May 22:12 2011
Picon

Re: HELP:How can one VM interrupt another running VM ??

On Tue, May 3, 2011 at 12:58, shule ney <neyshule@...> wrote:

> I was wondering how can I use one VM to interrupt another running VM, is
> there any schedular surpport this function, or any control I can use
> to implement it. For OpenVZ, is it possible to achieve this?

Not as such, no, but: OpenVZ has a single scheduler across all threads
/ processes in all VEs.  So, if you imagine the scheduler as working
on a single system with all those tasks the normal rules of process
scheduling apply.

Which means that if you set your higher priority stuff to a high nice
level, or real-time priority, it will take precedence over another VE
that runs lower priority stuff.

Finally, if you assign appropriate CPU shares to the VEs you can
ensure that they will get all the time divided appropriately at a high
level.

Regards,
    Daniel

PS: Native Linux doesn't support this either. :)

--

-- 
⎋ Puppet Labs Developer – http://puppetlabs.com
✉ Daniel Pittman <daniel@...>
✆ Contact me via gtalk, email, or phone: +1 (503) 893-2285
♲ Made with 100 percent post-consumer electrons
shule ney | 3 May 22:27 2011
Picon

Re: HELP:How can one VM interrupt another running VM ??

Thanks Daniel, but it's not interrupt for that thread/process, the nice value just set higher priority, what I want to do is like stop the current running process in one VM, context swiching to another process in another VM, when this process finishes, switch back to the original VM, of course this preempting VM has higher nice value. Is it possible?  

2011/5/3 Daniel Pittman <daniel-zvVxMF7wGoXk1uMJSBkQmQ@public.gmane.org>
On Tue, May 3, 2011 at 12:58, shule ney <neyshule-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:

> I was wondering how can I use one VM to interrupt another running VM, is
> there any schedular surpport this function, or any control I can use
> to implement it. For OpenVZ, is it possible to achieve this?

Not as such, no, but: OpenVZ has a single scheduler across all threads
/ processes in all VEs.  So, if you imagine the scheduler as working
on a single system with all those tasks the normal rules of process
scheduling apply.

Which means that if you set your higher priority stuff to a high nice
level, or real-time priority, it will take precedence over another VE
that runs lower priority stuff.

Finally, if you assign appropriate CPU shares to the VEs you can
ensure that they will get all the time divided appropriately at a high
level.

Regards,
   Daniel

PS: Native Linux doesn't support this either. :)

--
⎋ Puppet Labs Developer – http://puppetlabs.com
✉ Daniel Pittman <daniel-zvVxMF7wGoWDGRHsOpWV0g@public.gmane.orgt>
✆ Contact me via gtalk, email, or phone: +1 (503) 893-2285
♲ Made with 100 percent post-consumer electrons

_______________________________________________
Users mailing list
Users-GEFAQzZX7r8dnm+yROfE0A@public.gmane.org
https://openvz.org/mailman/listinfo/users

_______________________________________________
Users mailing list
Users@...
https://openvz.org/mailman/listinfo/users
Scott Dowdle | 3 May 23:27 2011

Re: HELP:How can one VM interrupt another running VM ??

Greetings,

----- Original Message -----
> Thanks Daniel, but it's not interrupt for that thread/process, the
> nice value just set higher priority, what I want to do is like stop
> the current running process in one VM, context swiching to another
> process in another VM, when this process finishes, switch back to
> the original VM, of course this preempting VM has higher nice value.
> Is it possible?

No, that is NOT a built in feature of OpenVZ... and one would argue that it should not be a feature.

Having said that though, assuming you have root access to the host node running the containers, you could
setup some type of monitoring of the process in the container that communicates to the host node, and when
the host node sees a condition, you could chkcpoint the container, pass a message to the second container,
start up some process there, monitor it, and when done, restore from chkpoint the first container.

That would take some work.  The easiest way for monitoring would be checking for the presence or absence of a
file... and that should be easy since the host node can see the container's filesystems.

I don't know if that is what you want or not but if I had to have that feature, that's what I'd do.

TYL,
--

-- 
Scott Dowdle
704 Church Street
Belgrade, MT 59714
(406)388-0827 [home]
(406)994-3931 [work]
Daniel Pittman | 3 May 23:41 2011
Picon

Re: HELP:How can one VM interrupt another running VM ??

No, that is absolutely impossible.  The only mechanism for that Linux
provides for this is using RT scheduling and IPC, and those are not
really available to you.  What you are describing sounds very much
like cooperative multitasking, which Linux doesn't do at this level.

Are you sure you don't want, for example, a TCP socket passing data to
the "switch to" process, and then to block until the response comes
back?

Daniel

On Tue, May 3, 2011 at 13:27, shule ney <neyshule@...> wrote:
> Thanks Daniel, but it's not interrupt for that thread/process, the nice
> value just set higher priority, what I want to do is like stop the current
> running process in one VM, context swiching to another process in another
> VM, when this process finishes, switch back to the original VM, of
> course this preempting VM has higher nice value. Is it possible?
>
> 2011/5/3 Daniel Pittman <daniel@...>
>>
>> On Tue, May 3, 2011 at 12:58, shule ney <neyshule@...> wrote:
>>
>> > I was wondering how can I use one VM to interrupt another running VM, is
>> > there any schedular surpport this function, or any control I can use
>> > to implement it. For OpenVZ, is it possible to achieve this?
>>
>> Not as such, no, but: OpenVZ has a single scheduler across all threads
>> / processes in all VEs.  So, if you imagine the scheduler as working
>> on a single system with all those tasks the normal rules of process
>> scheduling apply.
>>
>> Which means that if you set your higher priority stuff to a high nice
>> level, or real-time priority, it will take precedence over another VE
>> that runs lower priority stuff.
>>
>> Finally, if you assign appropriate CPU shares to the VEs you can
>> ensure that they will get all the time divided appropriately at a high
>> level.
>>
>> Regards,
>>    Daniel
>>
>> PS: Native Linux doesn't support this either. :)
>>
>> --
>> ⎋ Puppet Labs Developer – http://puppetlabs.com
>> ✉ Daniel Pittman <daniel@...>
>> ✆ Contact me via gtalk, email, or phone: +1 (503) 893-2285
>> ♲ Made with 100 percent post-consumer electrons
>>
>> _______________________________________________
>> Users mailing list
>> Users@...
>> https://openvz.org/mailman/listinfo/users
>
>
> _______________________________________________
> Users mailing list
> Users@...
> https://openvz.org/mailman/listinfo/users
>
>

--

-- 
⎋ Puppet Labs Developer – http://puppetlabs.com
✉ Daniel Pittman <daniel@...>
✆ Contact me via gtalk, email, or phone: +1 (503) 893-2285
♲ Made with 100 percent post-consumer electrons
Anton Gorlov | 5 May 13:30 2011
Picon

Brocade FC/FCOE Adapter, model: Brocade-815 (bfa)

 From which version has the ovz kernel begun to support Brocade-815 
(bfa) adapter?

I'm also wondering whether the bug "BFA[error] HAL_ASSERT: Assertion 
failure:" has been fixed.

Gmane