OpenVZ Forum


Home » Mailing lists » Users » Service restart on host system kills same service inside all VEs
Service restart on host system kills same service inside all VEs [message #44050] Fri, 11 November 2011 12:53 Go to next message
antoney is currently offline  antoney
Messages: 12
Registered: January 2011
Location: Ukraine
Junior Member
From: *parallels.com
Hello, guys.

Everything is in subject. I'm experiencing this problem on both my openvz
hosts. For httpd it happens from time to time, for nrpe - always...
Tell please where should I look.
In /proc/user_beancounters "failcnt" is 0 for all options.

Anton
Re: Service restart on host system kills same service insideall VEs [message #44051 is a reply to message #44050] Fri, 11 November 2011 13:20 Go to previous messageGo to next message
Stoyan Stoyanov is currently offline  Stoyan Stoyanov
Messages: 11
Registered: November 2011
Junior Member
From: *parallels.com
Hi man,

I was used nrpe inside vz containers and on the hn from along time ago
and never see behavior like this. Is there some customizations in your
nrpe service start/stop scripts or something, I think there is the
place too look at why this happens. ... I think this situation is not
openvz related in specific, correct me if im wrong

best regs




On Nov 11, 2011, at 2:53 PM, Anton Zaytsev wrote:

> Hello, guys.
>
> Everything is in subject. I'm experiencing this problem on both my
> openvz hosts. For httpd it happens from time to time, for nrpe -
> always...
> Tell please where should I look.
> In /proc/user_beancounters "failcnt" is 0 for all options.
>
> Anton
>
> <ATT00001.txt>

Stoyan Stoyanov
Core System Administrator




CONFIDENTIAL
The information contained in this email and any attachment is
confidential. It is intended only for the named addressee(s). If you
are not the named addressee(s) please notify the sender immediately
and do not disclose, copy or distribute the contents to any other
person other than the intended addressee(s).
Re: Service restart on host system kills same service inside all VEs [message #44054 is a reply to message #44050] Fri, 11 November 2011 13:31 Go to previous messageGo to next message
Gary Wallis is currently offline  Gary Wallis
Messages: 15
Registered: July 2011
Junior Member
From: *parallels.com
Anton Zaytsev wrote:
> Hello, guys.
>
> Everything is in subject. I'm experiencing this problem on both my
> openvz hosts. For httpd it happens from time to time, for nrpe - always...
> Tell please where should I look.
> In /proc/user_beancounters "failcnt" is 0 for all options.
>
Your service restart script X (/etc/init.d/X) is probably killing
processes it should not.

E.g. host system kill will kill processes "inside" the containers.

Cheers,
Gary
Re: Service restart on host system kills same service insideall VEs [message #44056 is a reply to message #44051] Fri, 11 November 2011 13:40 Go to previous messageGo to next message
antoney is currently offline  antoney
Messages: 12
Registered: January 2011
Location: Ukraine
Junior Member
From: *parallels.com
Thanks for quick reply.


> Is there some customizations in your nrpe service start/stop scripts or
> something

no, all scripts are default. This happens for httpd too.

I suppose it's openvz related. Seems like some misunderstanding between
host and ve...
Systems are centos 4 and 5, latest kernels from repo.

On Fri, Nov 11, 2011 at 3:20 PM, Stoyan Stoyanov
<s.stoianov@maxtelecom.bg>wrote:

> Hi man,
>
> I was used nrpe inside vz containers and on the hn from along time ago and
> never see behavior like this. Is there some customizations in your nrpe
> service start/stop scripts or something, I think there is the place too
> look at why this happens. ... I think this situation is not openvz related
> in specific, correct me if im wrong
>
> best regs
>
>
>
>
>
> On Nov 11, 2011, at 2:53 PM, Anton Zaytsev wrote:
>
> Hello, guys.
>>
>> Everything is in subject. I'm experiencing this problem on both my openvz
>> hosts. For httpd it happens from time to time, for nrpe - always...
>> Tell please where should I look.
>> In /proc/user_beancounters "failcnt" is 0 for all options.
>>
>> Anton
>>
>> <ATT00001.txt>
>>
>
> Stoyan Stoyanov
> Core System Administrator
>
>
>
>
> CONFIDENTIAL
> The information contained in this email and any attachment is
> confidential. It is intended only for the named addressee(s). If you are
> not the named addressee(s) please notify the sender immediately and do not
> disclose, copy or distribute the contents to any other person other than
> the intended addressee(s).
>
>
Re: Service restart on host system kills same serviceinsideall VEs [message #44057 is a reply to message #44056] Fri, 11 November 2011 13:52 Go to previous messageGo to next message
Stoyan Stoyanov is currently offline  Stoyan Stoyanov
Messages: 11
Registered: November 2011
Junior Member
From: *parallels.com
LIke Gary says, the host system just kills the ct's proceses (may be
in a killall `procesname` style :)))
so the thing that u should take care is the procces's start/stop scripts

On Nov 11, 2011, at 3:40 PM, Anton Zaytsev wrote:

> Thanks for quick reply.
>
> Is there some customizations in your nrpe service start/stop scripts
> or something
> no, all scripts are default. This happens for httpd too.
>
> I suppose it's openvz related. Seems like some misunderstanding
> between host and ve...
> Systems are centos 4 and 5, latest kernels from repo.
>
> On Fri, Nov 11, 2011 at 3:20 PM, Stoyan Stoyanov <s.stoianov@maxtelecom.bg
> > wrote:
> Hi man,
>
> I was used nrpe inside vz containers and on the hn from along time
> ago and never see behavior like this. Is there some customizations
> in your nrpe service start/stop scripts or something, I think there
> is the place too look at why this happens. ... I think this
> situation is not openvz related in specific, correct me if im wrong
>
> best regs
>
>
>
>
>
> On Nov 11, 2011, at 2:53 PM, Anton Zaytsev wrote:
>
> Hello, guys.
>
> Everything is in subject. I'm experiencing this problem on both my
> openvz hosts. For httpd it happens from time to time, for nrpe -
> always...
> Tell please where should I look.
> In /proc/user_beancounters "failcnt" is 0 for all options.
>
> Anton
>
> <ATT00001.txt>
>
> Stoyan Stoyanov
> Core System Administrator
>
>
>
>
> CONFIDENTIAL
> The information contained in this email and any attachment is
> confidential. It is intended only for the named addressee(s). If you
> are not the named addressee(s) please notify the sender immediately
> and do not disclose, copy or distribute the contents to any other
> person other than the intended addressee(s).
>
>
> <ATT00001.txt>

Stoyan Stoyanov
Core System Administrator




CONFIDENTIAL
The information contained in this email and any attachment is
confidential. It is intended only for the named addressee(s). If you
are not the named addressee(s) please notify the sender immediately
and do not disclose, copy or distribute the contents to any other
person other than the intended addressee(s).
Re: Service restart on host system kills same service insideall VEs [message #44059 is a reply to message #44056] Fri, 11 November 2011 14:04 Go to previous messageGo to next message
antoney is currently offline  antoney
Messages: 12
Registered: January 2011
Location: Ukraine
Junior Member
From: *parallels.com
Ok, nrpe init script does "killproc nrpe" and its killing all processes
named nrpe, inside containers too. But almost all of init scripts do so.
So modify all of them? Looks not so pretty:)

On Fri, Nov 11, 2011 at 3:40 PM, Anton Zaytsev <anton.zajtsev@gmail.com>wrote:

> Thanks for quick reply.
>
>
>> Is there some customizations in your nrpe service start/stop scripts or
>> something
>
> no, all scripts are default. This happens for httpd too.
>
> I suppose it's openvz related. Seems like some misunderstanding between
> host and ve...
> Systems are centos 4 and 5, latest kernels from repo.
>
> On Fri, Nov 11, 2011 at 3:20 PM, Stoyan Stoyanov <s.stoianov@maxtelecom.bg
> > wrote:
>
>> Hi man,
>>
>> I was used nrpe inside vz containers and on the hn from along time ago
>> and never see behavior like this. Is there some customizations in your nrpe
>> service start/stop scripts or something, I think there is the place too
>> look at why this happens. ... I think this situation is not openvz related
>> in specific, correct me if im wrong
>>
>> best regs
>>
>>
>>
>>
>>
>> On Nov 11, 2011, at 2:53 PM, Anton Zaytsev wrote:
>>
>> Hello, guys.
>>>
>>> Everything is in subject. I'm experiencing this problem on both my
>>> openvz hosts. For httpd it happens from time to time, for nrpe - always...
>>> Tell please where should I look.
>>> In /proc/user_beancounters "failcnt" is 0 for all options.
>>>
>>> Anton
>>>
>>> <ATT00001.txt>
>>>
>>
>> Stoyan Stoyanov
>> Core System Administrator
>>
>>
>>
>>
>> CONFIDENTIAL
>> The information contained in this email and any attachment is
>> confidential. It is intended only for the named addressee(s). If you are
>> not the named addressee(s) please notify the sender immediately and do not
>> disclose, copy or distribute the contents to any other person other than
>> the intended addressee(s).
>>
>>
>>
Re: Service restart on host system kills same serviceinsideall VEs [message #44060 is a reply to message #44059] Fri, 11 November 2011 14:46 Go to previous messageGo to next message
Stoyan Stoyanov is currently offline  Stoyan Stoyanov
Messages: 11
Registered: November 2011
Junior Member
From: *parallels.com
Agree :) it's not so pretty.. but every process from every container
is a seen process on the host

On Nov 11, 2011, at 4:04 PM, Anton Zaytsev wrote:

> Ok, nrpe init script does "killproc nrpe" and its killing all
> processes named nrpe, inside containers too. But almost all of init
> scripts do so.
> So modify all of them? Looks not so pretty:)
>
> On Fri, Nov 11, 2011 at 3:40 PM, Anton Zaytsev <anton.zajtsev@gmail.com
> > wrote:
> Thanks for quick reply.
>
> Is there some customizations in your nrpe service start/stop scripts
> or something
> no, all scripts are default. This happens for httpd too.
>
> I suppose it's openvz related. Seems like some misunderstanding
> between host and ve...
> Systems are centos 4 and 5, latest kernels from repo.
>
> On Fri, Nov 11, 2011 at 3:20 PM, Stoyan Stoyanov <s.stoianov@maxtelecom.bg
> > wrote:
> Hi man,
>
> I was used nrpe inside vz containers and on the hn from along time
> ago and never see behavior like this. Is there some customizations
> in your nrpe service start/stop scripts or something, I think there
> is the place too look at why this happens. ... I think this
> situation is not openvz related in specific, correct me if im wrong
>
> best regs
>
>
>
>
>
> On Nov 11, 2011, at 2:53 PM, Anton Zaytsev wrote:
>
> Hello, guys.
>
> Everything is in subject. I'm experiencing this problem on both my
> openvz hosts. For httpd it happens from time to time, for nrpe -
> always...
> Tell please where should I look.
> In /proc/user_beancounters "failcnt" is 0 for all options.
>
> Anton
>
> <ATT00001.txt>
>
> Stoyan Stoyanov
> Core System Administrator
>
>
>
>
> CONFIDENTIAL
> The information contained in this email and any attachment is
> confidential. It is intended only for the named addressee(s). If you
> are not the named addressee(s) please notify the sender immediately
> and do not disclose, copy or distribute the contents to any other
> person other than the intended addressee(s).
>
>
>
> <ATT00001.txt>

Stoyan Stoyanov
Core System Administrator




CONFIDENTIAL
The information contained in this email and any attachment is
confidential. It is intended only for the named addressee(s). If you
are not the named addressee(s) please notify the sender immediately
and do not disclose, copy or distribute the contents to any other
person other than the intended addressee(s).
Re: Service restart on host system kills same service insideall VEs [message #44061 is a reply to message #44059] Fri, 11 November 2011 15:37 Go to previous messageGo to next message
Tim Small is currently offline  Tim Small
Messages: 24
Registered: April 2011
Junior Member
From: *parallels.com
On 11/11/11 14:04, Anton Zaytsev wrote:
> Ok, nrpe init script does "killproc nrpe" and its killing all processes
> named nrpe, inside containers too. But almost all of init scripts do so.
> So modify all of them? Looks not so pretty:)
>

Use Debian? I've not seen this broken behaviour with any of Debian's
init scripts...

I would say that it's also broken for non-openvz hosts - I've previously
run several separate copies of apache and ssh (each with their own init
scripts) on a Debian host before, and restarting one hasn't killed the
others...

Open bugs against Redhat/CentOS? I'd guess this problem will continue
with LXC in the future too...

Tim.

--
South East Open Source Solutions Limited
Registered in England and Wales with company number 06134732.
Registered Office: 2 Powell Gardens, Redhill, Surrey, RH1 1TQ
VAT number: 900 6633 53 http://seoss.co.uk/ +44-(0)1273-808309
Re: Service restart on host system kills same serviceinsideall VEs [message #44063 is a reply to message #44060] Fri, 11 November 2011 15:59 Go to previous messageGo to next message
antoney is currently offline  antoney
Messages: 12
Registered: January 2011
Location: Ukraine
Junior Member
From: *parallels.com
Ok, I've decided to just leave it as is. It's production systems, service
restarts are not so often.
So only solution for this issue now is to use host system just as host.

Tim,
Migration to Debian will be not so simple.
Yes, I think it's needed to send bug report to CentOS/Redhat packaging. I
surprised that no one has mentioned this before.

On Fri, Nov 11, 2011 at 4:46 PM, Stoyan Stoyanov
<s.stoianov@maxtelecom.bg>wrote:

> Agree :) it's not so pretty.. but every process from every container is a
> seen process on the host
>
>
> On Nov 11, 2011, at 4:04 PM, Anton Zaytsev wrote:
>
> Ok, nrpe init script does "killproc nrpe" and its killing all processes
>> named nrpe, inside containers too. But almost all of init scripts do so.
>> So modify all of them? Looks not so pretty:)
>>
>> On Fri, Nov 11, 2011 at 3:40 PM, Anton Zaytsev <anton.zajtsev@gmail.com>
>> wrote:
>> Thanks for quick reply.
>>
>> Is there some customizations in your nrpe service start/stop scripts or
>> something
>> no, all scripts are default. This happens for httpd too.
>>
>> I suppose it's openvz related. Seems like some misunderstanding between
>> host and ve...
>> Systems are centos 4 and 5, latest kernels from repo.
>>
>> On Fri, Nov 11, 2011 at 3:20 PM, Stoyan Stoyanov <
>> s.stoianov@maxtelecom.bg> wrote:
>> Hi man,
>>
>> I was used nrpe inside vz containers and on the hn from along time ago
>> and never see behavior like this. Is there some customizations in your nrpe
>> service start/stop scripts or something, I think there is the place too
>> look at why this happens. ... I think this situation is not openvz related
>> in specific, correct me if im wrong
>>
>> best regs
>>
>>
>>
>>
>>
>> On Nov 11, 2011, at 2:53 PM, Anton Zaytsev wrote:
>>
>> Hello, guys.
>>
>> Everything is in subject. I'm experiencing this problem on both my openvz
>> hosts. For httpd it happens from time to time, for nrpe - always...
>> Tell please where should I look.
>> In /proc/user_beancounters "failcnt" is 0 for all options.
>>
>> Anton
>>
>> <ATT00001.txt>
>>
>> Stoyan Stoyanov
>> Core System Administrator
>>
>>
>>
>>
>> CONFIDENTIAL
>> The information contained in this email and any attachment is
>> confidential. It is intended only for the named addressee(s). If you are
>> not the named addressee(s) please notify the sender immediately and do not
>> disclose, copy or distribute the contents to any other person other than
>> the intended addressee(s).
>>
>>
>>
>> ______________________________**_________________
>> Users mailing list
>> Users@openvz.org
>> https://openvz.org/mailman/**listinfo/users<https://openvz.org/mailman/listinfo/users>
>>
>>
>>
>> <ATT00001.txt>
>>
>
> Stoyan Stoyanov
> Core System Administrator
>
>
>
>
> CONFIDENTIAL
> The information contained in this email and any attachment is
> confidential. It is intended only for the named addressee(s). If you are
> not the named addressee(s) please notify the sender immediately and do not
> disclose, copy or distribute the contents to any other person other than
> the intended addressee(s).
>
>
Re: Service restart on host system kills same serviceinsideall VEs [message #44064 is a reply to message #44063] Fri, 11 November 2011 16:42 Go to previous messageGo to next message
antoney is currently offline  antoney
Messages: 12
Registered: January 2011
Location: Ukraine
Junior Member
From: *parallels.com
>
> Yes, I think it's needed to send bug report to CentOS/Redhat packaging. I
> surprised that no one has mentioned this before.

I mean this bug)

On Fri, Nov 11, 2011 at 5:59 PM, Anton Zaytsev <anton.zajtsev@gmail.com>wrote:

> Ok, I've decided to just leave it as is. It's production systems, service
> restarts are not so often.
> So only solution for this issue now is to use host system just as host.
>
> Tim,
> Migration to Debian will be not so simple.
> Yes, I think it's needed to send bug report to CentOS/Redhat packaging. I
> surprised that no one has mentioned this before.
>
> On Fri, Nov 11, 2011 at 4:46 PM, Stoyan Stoyanov <s.stoianov@maxtelecom.bg
> > wrote:
>
>> Agree :) it's not so pretty.. but every process from every container is a
>> seen process on the host
>>
>>
>> On Nov 11, 2011, at 4:04 PM, Anton Zaytsev wrote:
>>
>> Ok, nrpe init script does "killproc nrpe" and its killing all processes
>>> named nrpe, inside containers too. But almost all of init scripts do so.
>>> So modify all of them? Looks not so pretty:)
>>>
>>> On Fri, Nov 11, 2011 at 3:40 PM, Anton Zaytsev <anton.zajtsev@gmail.com>
>>> wrote:
>>> Thanks for quick reply.
>>>
>>> Is there some customizations in your nrpe service start/stop scripts or
>>> something
>>> no, all scripts are default. This happens for httpd too.
>>>
>>> I suppose it's openvz related. Seems like some misunderstanding between
>>> host and ve...
>>> Systems are centos 4 and 5, latest kernels from repo.
>>>
>>> On Fri, Nov 11, 2011 at 3:20 PM, Stoyan Stoyanov <
>>> s.stoianov@maxtelecom.bg> wrote:
>>> Hi man,
>>>
>>> I was used nrpe inside vz containers and on the hn from along time ago
>>> and never see behavior like this. Is there some customizations in your nrpe
>>> service start/stop scripts or something, I think there is the place too
>>> look at why this happens. ... I think this situation is not openvz related
>>> in specific, correct me if im wrong
>>>
>>> best regs
>>>
>>>
>>>
>>>
>>>
>>> On Nov 11, 2011, at 2:53 PM, Anton Zaytsev wrote:
>>>
>>> Hello, guys.
>>>
>>> Everything is in subject. I'm experiencing this problem on both my
>>> openvz hosts. For httpd it happens from time to time, for nrpe - always...
>>> Tell please where should I look.
>>> In /proc/user_beancounters "failcnt" is 0 for all options.
>>>
>>> Anton
>>>
>>> <ATT00001.txt>
>>>
>>> Stoyan Stoyanov
>>> Core System Administrator
>>>
>>>
>>>
>>>
>>> CONFIDENTIAL
>>> The information contained in this email and any attachment is
>>> confidential. It is intended only for the named addressee(s). If you are
>>> not the named addressee(s) please notify the sender immediately and do not
>>> disclose, copy or distribute the contents to any other person other than
>>> the intended addressee(s).
>>>
>>>
>>>
>>> ______________________________**_________________
>>> Users mailing list
>>> Users@openvz.org
>>> https://openvz.org/mailman/**listinfo/users<https://openvz.org/mailman/listinfo/users>
>>>
>>>
>>>
>>> <ATT00001.txt>
>>>
>>
>> Stoyan Stoyanov
>> Core System Administrator
>>
>>
>>
>>
>> CONFIDENTIAL
>> The information contained in this email and any attachment is
>> confidential. It is intended only for the named addressee(s). If you are
>> not the named addressee(s) please notify the sender immediately and do not
>> disclose, copy or distribute the contents to any other person other than
>> the intended addressee(s).
>>
>>
Re: Service restart on host system kills same service insideall VEs [message #44065 is a reply to message #44059] Fri, 11 November 2011 20:42 Go to previous messageGo to next message
MailingListe is currently offline  MailingListe
Messages: 29
Registered: May 2008
Junior Member
From: *parallels.com
Zitat von Anton Zaytsev <anton.zajtsev@gmail.com>:

> Ok, nrpe init script does "killproc nrpe" and its killing all processes
> named nrpe, inside containers too. But almost all of init scripts do so.
> So modify all of them? Looks not so pretty:)

you should not run any services beside maybe SSH on the hardware node.
It is the "root" node and should only be used to manage the VEs. Every
problem you got with any service running at the HN will affect all
VEs. Use the HN only as hypervisor and you are done.

Regards

Andreas
  • Attachment: smime.p7s
    (Size: 6.03KB, Downloaded 119 times)
Re: Service restart on host system kills same service insideall VEs [message #44066 is a reply to message #44065] Fri, 11 November 2011 22:50 Go to previous messageGo to next message
Sharp is currently offline  Sharp
Messages: 14
Registered: March 2011
Junior Member
From: *parallels.com
On Sat, Nov 12, 2011 at 12:42 AM, <lst_hoe02@kwsoft.de> wrote:
> you should not run any services beside maybe SSH on the hardware node. It is
> the "root" node and should only be used to manage the VEs.

That's bad advice. It's generally good idea to run things like nrpe
and munin on HN to monitor it's performance and status.

--
SY, Ilya A. Otyutskiy aka Sharp
Re: Service restart on host system kills same service insideall VEs [message #44070 is a reply to message #44066] Sat, 12 November 2011 19:24 Go to previous messageGo to next message
MailingListe is currently offline  MailingListe
Messages: 29
Registered: May 2008
Junior Member
From: *parallels.com
Zitat von "Ilya A. Otyutskiy" <sharp@thesharp.ru>:

> On Sat, Nov 12, 2011 at 12:42 AM, <lst_hoe02@kwsoft.de> wrote:
>> you should not run any services beside maybe SSH on the hardware node. It is
>> the "root" node and should only be used to manage the VEs.
>
> That's bad advice. It's generally good idea to run things like nrpe
> and munin on HN to monitor it's performance and status.

If this "things" or the start/stop scripts used tend to kill all kind
of processes not belonging to them it *is* a really bad idea to run
such "things" on the HN. You should only use software absolutely
needed and really trustworth at the HN. Do all other things in the VE.
This even apply monitor software.
If you do not agree, do not complain about the problems arise.

Regards

Andreas
  • Attachment: smime.p7s
    (Size: 6.03KB, Downloaded 117 times)
Re: Service restart on host system kills same service insideall VEs [message #44072 is a reply to message #44070] Sun, 13 November 2011 10:50 Go to previous messageGo to next message
Sharp is currently offline  Sharp
Messages: 14
Registered: March 2011
Junior Member
From: *parallels.com
On Sat, Nov 12, 2011 at 11:24 PM, <lst_hoe02@kwsoft.de> wrote:
> If this "things" or the start/stop scripts used tend to kill all kind of
> processes not belonging to them it *is* a really bad idea to run such
> "things" on the HN.

So you are suggesting to throw those useful "things" away instead of
patching their init-scripts and sending them to upstream?

> You should only use software absolutely needed and
> really trustworth at the HN. Do all other things in the VE. This even apply
> monitor software.

Maybe you will tell us how to monitor something like HN's disk space from VE? :)

--
SY, Ilya A. Otyutskiy aka Sharp
Re: Service restart on host system kills same serviceinsideall VEs [message #44078 is a reply to message #44063] Mon, 14 November 2011 09:57 Go to previous messageGo to next message
Aleksandar Ivanisevic is currently offline  Aleksandar Ivanisevic
Messages: 34
Registered: April 2011
Member
From: *parallels.com
Anton Zaytsev <anton.zajtsev@gmail.com>
writes:

> Ok, I've decided to just leave it as is. It's production systems, service
> restarts are not so often.
> So only solution for this issue now is to use host system just as host.
>
> Tim,
> Migration to Debian will be not so simple.
> Yes, I think it's needed to send bug report to CentOS/Redhat packaging. I
> surprised that no one has mentioned this before.

[...]

It gets mentioned every once in a while, but this is a general problem
with centos init scripts and/or most init scripts in general, they are
mostly written to assume a single instance of the service.

The solution is simply to run as little as possible on the host, and
to fix the scripts for the stuff you need to run. Painful, but
unavoidable. Maybe if you submit a patch to rhel, it could maybe make
it into a release in a couple of years ;)


--
Ti si arogantan, prepotentan i peglaš vlastitu frustraciju. -- Ivan
Tišljar, hr.comp.os.linux
Re: Re: Service restart on host system kills same serviceinsideall VEs [message #44080 is a reply to message #44078] Mon, 14 November 2011 16:46 Go to previous messageGo to next message
antoney is currently offline  antoney
Messages: 12
Registered: January 2011
Location: Ukraine
Junior Member
From: *parallels.com
I've submited the bug to RHEL.
https://bugzilla.redhat.com/show_bug.cgi?id=753834

Yes for now it's only to limit services on the host.

On Mon, Nov 14, 2011 at 11:57 AM, Aleksandar Ivanisevic <
aleksandar@ivanisevic.de> wrote:

> Anton Zaytsev <anton.zajtsev@gmail.com>
> writes:
>
> > Ok, I've decided to just leave it as is. It's production systems, service
> > restarts are not so often.
> > So only solution for this issue now is to use host system just as host.
> >
> > Tim,
> > Migration to Debian will be not so simple.
> > Yes, I think it's needed to send bug report to CentOS/Redhat packaging. I
> > surprised that no one has mentioned this before.
>
> [...]
>
> It gets mentioned every once in a while, but this is a general problem
> with centos init scripts and/or most init scripts in general, they are
> mostly written to assume a single instance of the service.
>
> The solution is simply to run as little as possible on the host, and
> to fix the scripts for the stuff you need to run. Painful, but
> unavoidable. Maybe if you submit a patch to rhel, it could maybe make
> it into a release in a couple of years ;)
>
>
> --
> Ti si arogantan, prepotentan i peglaš vlastitu frustraciju. -- Ivan
> Tišljar, hr.comp.os.linux
>
Re: Re: Service restart on host system kills same serviceinsideall VEs [message #44081 is a reply to message #44080] Mon, 14 November 2011 17:56 Go to previous messageGo to next message
Tim Small is currently offline  Tim Small
Messages: 24
Registered: April 2011
Junior Member
From: *parallels.com
On 14/11/11 16:46, Anton Zaytsev wrote:
> I've submited the bug to RHEL.
> https://bugzilla.redhat.com/show_bug.cgi?id=753834
>
> Yes for now it's only to limit services on the host.
>


To get this bug taken more seriously, I would be more concrete.

e.g. give the name of an init script which fails, and corresponding rpm
which it shipped with.

Give the example of a development server which runs multiple copies of
said service for multiple developers (with separate config/data/pid
files). Say that restarting the system service kills all copies.

paste in some shell history exhibiting this behaviour.

I'd also mention that Debian, Ubuntu etc. doesn't exhibit this behaviour
e.g. search for "and that they don't kill unfortunately-named user
processes" in http://www.debian.org/doc/debian-policy/ch-opersys.html

Cheers,

Tim.


--
South East Open Source Solutions Limited
Registered in England and Wales with company number 06134732.
Registered Office: 2 Powell Gardens, Redhill, Surrey, RH1 1TQ
VAT number: 900 6633 53 http://seoss.co.uk/ +44-(0)1273-808309
Re: Service restart on host system kills same serviceinsideall VEs [message #44087 is a reply to message #44080] Tue, 15 November 2011 09:43 Go to previous message
Aleksandar Ivanisevic is currently offline  Aleksandar Ivanisevic
Messages: 34
Registered: April 2011
Member
From: *parallels.com
Anton Zaytsev <anton.zajtsev@gmail.com>
writes:

> I've submited the bug to RHEL.
> https://bugzilla.redhat.com/show_bug.cgi?id=753834


Eh, you have filed it against EL4, which is going to be EOLed in 3 months ;)
Previous Topic: bug or feature?: ps -el on HN shows all processes, incl. those of VEs
Next Topic: Question About Debian Squeeze OVZ Kernel
Goto Forum:
  


Current Time: Mon Oct 23 00:38:11 GMT 2017