[Gluster-infra] build.gluster.org downtime on 20th June

[Nigel Babu]

Other than Gerrit, the rest can be rebooted ruthlessly.

On 19-Jun-2017 23:16, "Michael Scherer" <mscherer at redhat.com> wrote:

> Le lundi 19 juin 2017 à 17:55 +0200, Michael Scherer a écrit :
> > Le lundi 19 juin 2017 à 17:12 +0530, Nigel Babu a écrit :
> > > Hello folks,
> > >
> > > We'll be having a short downtime for build.gluster.org on 20th June
> 2017
> > > (tomorrow).
> > >
> > > Date: 20th June 2017
> > > Time: 0230 UTC (2230 EDT / 0430 CEST / 0800 IST)
> > > Duration: 1h
> > >
> > > Jenkins will be in a quiet time from 1h before the outage where no new
> builds
> > > will be allowed to start.
> > >
> > > This downtime is to complete the installation of a required plugin.
> Though
> > > there is a 1h downtime window, I expect the actual outage to last
> about 20
> > > minutes at maximum.
> >
> > Since https://www.qualys.com/2017/06/19/stack-clash/stack-clash.txt just
> > have been published and it kinda mean "rebooting the whole infra", can
> > we combine do it at the same time ? (it might mean "wait until we have
> > Centos updated package", ie likely more than 24h) ?
>
> Ok, so upon reading the advisory a bit more, I do not see much reason to
> panic (selinux would contains most exploitation vectors I can think off,
> we strongly limited access, and the exploit aren't exactly silent nor
> fast), but I would still like to combine the 2 downtime if possible.
>
> I will also ruthlessly reboot the various builders during the week,
> unless people tell me not to do it.
> --
> Michael Scherer
> Sysadmin, Community Infrastructure and Platform, OSAS
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.gluster.org/pipermail/gluster-infra/attachments/20170620/e0f29a3c/attachment.html>