Open Stack

Prod is based on nginx. It is configured to forward dynamic requests to
uwsgi process, thus nginx itself does not consume much memory.

The config is following:
server {
listen   80;
server_name stackalytics.com;
location / {
        uwsgi_pass  unix:///tmp/stackalytics.sock;
        include     uwsgi_params;
}
location  /static/ {
        alias  /opt/stack/stackalytics/stackalytics/dashboard/static/;
}
}

--Ilya

2015-11-09 4:56 GMT+03:00 Paul Belanger <pabelanger at redhat.com>:

> On Fri, Nov 06, 2015 at 06:59:15PM +0300, Ilya Shakhat wrote:
> > I confirm that 8Gb is not enough. Currently the production is 16Gb - 3.5
> > for memcached, 3.5 for every uwsgi process and up to 2 may be consumed by
> > the processor.
> >
> Out of curiosity, are you using nginx or apache?  I've currently setup
> Apache,
> but might test nginx tomorrow. If you are using nginx how much RAM is
> consumed
> by each process.
>
> > --Ilya
> >
> > 2015-11-06 17:27 GMT+03:00 Paul Belanger <pabelanger at redhat.com>:
> >
> > > On Mon, Nov 02, 2015 at 08:25:19AM -0500, Paul Belanger wrote:
> > > > On Thu, Oct 29, 2015 at 07:14:00AM +0300, Ilya Shakhat wrote:
> > > > > Hi Paul,
> > > > >
> > > > > Using existing dump makes the cold start much more pleasant. Full
> data
> > > > > reload usually needed only when some incompatible changes are
> > > introduced. I
> > > > > mark such commits with the warning note and bump release number. To
> > > help
> > > > > you with the initial setup I can grab the dump from production and
> > > share it
> > > > > with you. The raw size is about 2G, hope it compressible.
> > > > >
> > > > Great. I'm sure if you jump into #openstack-infra we can find a
> > > > location to upload it too.
> > > >
> > > > > This week I'm at the summit. If you are not, we can schedule a chat
> > > next
> > > > > week, e.g. on Thu would be ok.
> > > > >
> > > > Yup, that works for me too. Again, I'll be idling in #openstack-infra
> > > when you
> > > > have free time.
> > > >
> > > > > Thanks,
> > > > > Ilya
> > > > >
> > > > > 2015-10-28 18:20 GMT+03:00 Paul Belanger <pabelanger at redhat.com>:
> > > > >
> > > > > > Greetings Ilya,
> > > > > >
> > > > > > I was hoping you could help save me some time doing the initial
> > > import of
> > > > > > stackalytics.o.o by using a stackalytics-dump from your
> instance? If
> > > I
> > > > > > understand correctly, this is how you currently do backups /
> > > persistent
> > > > > > data
> > > > > > from memcached?
> > > > > >
> > > > > > Additionally, maybe we could setup some time next week to chat
> about
> > > how to
> > > > > > currently do maintenance on stackalytics.org, so I can include
> that
> > > > > > functionality
> > > > > > into puppet-stackalytics[1]. Things like how often you dump
> memcached
> > > > > > data, if
> > > > > > any existing scripts exists, current backup / restore procedures.
> > > > > >
> > > > > > PB
> > > > > >
> > > > > > [1] https://github.com/openstack-infra/puppet-stackalytics
> > > > > >
> > > >
> > > Just a follow up to this thread, we've successfully merged 2 patches
> that
> > > now
> > > allow -infra to launch stackalytics.o.o[1][2].  At the moment, we are
> > > holding
> > > off on doing so because we'd like to confirm with you the current
> system
> > > requirements for stackalytics.com.  In my testing I found 8GB not to
> be
> > > enough
> > > RAM for importing the stats and running uwsgi.  We are considering
> > > launching a
> > > node with 15GB of RAM however would like to hear back from you first.
> > >
> > > Also, do you mind connecting with on of the infra-roots to discuss an
> > > upload
> > > location for your current memcached dump?
> > >
> > > ---
> > > [1] https://review.openstack.org/#/c/190765/
> > > [2] https://review.openstack.org/#/c/190708/
> > >
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-infra/attachments/20151109/8e3ff2bf/attachment.html>