fuel-dev team mailing list archive
-
fuel-dev team
-
Mailing list archive
-
Message #00455
Re: Ubuntu HA failures on powerful servers
Any update on this bug? Anyone else reproduced it on hardware?
On Thu, Feb 13, 2014 at 5:36 PM, Sergey Vasilenko
<svasilenko@xxxxxxxxxxxx>wrote:
> Bogdan, error happens not in puppet layer.
> The essence of happens is this:
> we can commit change of CIB (with newly created resource) and request
> status. Commit is async.
> In returned status we can not found newly created resource just now.
> But after cluster sync we found resource.
>
>
> On Thu, Feb 13, 2014 at 5:29 PM, Bogdan Dobrelya <bdobrelia@xxxxxxxxxxxx>wrote:
>
>> On 02/12/2014 05:41 PM, Sergey Vasilenko wrote:
>> > Bogdan, that's not enough. We should check after each operation, thah
>> > this operation finished. Not only the mythical "ready" status.
>>
>> Standard exists? method for puppet provider should ensure that. It is
>> also missed for commit & shadow.
>>
>> >
>> > OSCI promise, that tomorrow we will have PCS package on ISO.
>> > I will rewrite part of resources for using this tool. It contains
>> > sync.modes for some operations.
>> >
>> >
>> > On Wed, Feb 12, 2014 at 2:17 PM, Mike Scherbakov
>> > <mscherbakov@xxxxxxxxxxxx <mailto:mscherbakov@xxxxxxxxxxxx>> wrote:
>> >
>> > Thanks for acting on this quickly, Bogdan.
>> >
>> > So let's wait now while QA folks reproduce the issue, and then try
>> > to apply a patch?
>> >
>> >
>> > On Wed, Feb 12, 2014 at 2:15 PM, Bogdan Dobrelya
>> > <bdobrelia@xxxxxxxxxxxx <mailto:bdobrelia@xxxxxxxxxxxx>> wrote:
>> >
>> > On 02/12/2014 11:45 AM, Bogdan Dobrelya wrote:
>> > > On 02/12/2014 11:36 AM, Gleb Galkin wrote:
>> > >> On 12.02.2014 13:12, Mike Scherbakov wrote:
>> > >>> Fuelers,
>> > >>> that's the patch to 4.0/stable which I've got from
>> > >>> Gleb: http://paste.openstack.org/show/64570/
>> > >>> He did it together with Sergey to make the installation pass
>> > on Ubuntu
>> > >>> on powerful nodes. The issue is not reproducible on VMs.
>> > >>>
>> > >>> Gleb, Sergey - do we have a bug report about this?
>> > >> Sure we have it here
>> > <https://bugs.launchpad.net/fuel/+bug/1279211>
>> > >>
>> > >>
>> > >>> Nastya - I'm looking forward to see if it is reproducible
>> on our
>> > >>> hardware lab.
>> > >>>
>> > >>> Also, hacks with "sleep 15" are not something we want in
>> master,
>> > >>> obviously... any ideas how it should be solved properly?
>> > >
>> > > Correct me please If I'm wrong, could it be related to the
>> missed
>> > > self.class.block_until_ready call for cs_commit provider?
>> > >
>> > > (just grep for block_until_ready in corosync module to
>> clarify its
>> > > use cases)
>> >
>> > I believe, the https://review.openstack.org/#/c/72911/ should
>> > fix it.
>> > Need to verify.
>> >
>> > >
>> > >>>
>> > >>> Thanks,
>> > >>> --
>> > >>> Mike Scherbakov
>> > >>> #mihgen
>> > >>
>> > >>
>> > >>
>> > >> --
>> > >> Best Regards,
>> > >> Gleb Galkin
>> > >> OpenStack Deployment Engineer
>> > >>
>> > >> Mirantis Inc.
>> > >> www.mirantis.com <http://www.mirantis.com>
>> > >>
>> > >>
>> > >>
>> > >
>> > >
>> >
>> >
>> > --
>> > Best regards,
>> > Bogdan Dobrelya,
>> > Skype #bogdando_at_yahoo.com <http://bogdando_at_yahoo.com>
>> > Irc #bogdando
>> >
>> >
>> >
>> >
>> > --
>> > Mike Scherbakov
>> > #mihgen
>> >
>> > --
>> > Mailing list: https://launchpad.net/~fuel-dev
>> > Post to : fuel-dev@xxxxxxxxxxxxxxxxxxx
>> > <mailto:fuel-dev@xxxxxxxxxxxxxxxxxxx>
>> > Unsubscribe : https://launchpad.net/~fuel-dev
>> > More help : https://help.launchpad.net/ListHelp
>> >
>> >
>>
>>
>> --
>> Best regards,
>> Bogdan Dobrelya,
>> Skype #bogdando_at_yahoo.com
>> Irc #bogdando
>>
>
>
--
Mike Scherbakov
#mihgen
Follow ups
References