Re: Why so many buildfarm errors with contacting "git.postgresql.org"?

On Thu, Dec 9, 2010 at 23:53, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> Magnus Hagander <magnus@hagander.net> writes:
>> On Wed, Dec 8, 2010 at 00:06, Tom Lane <tgl@sss.pgh.pa.us> wrote:
>>> Is there any difference between the network connectivity of
>>> git.postgresql.org and the old anoncvs server?
>
>> Yes, they are in completely different datacenters.
>
>> I had a discussoin with Stefan a couple of days ago about this, and
>> the current estimate is that we're simply hitting the bandwidth limit
>> of where it is now, because it now takes so much more traffic. We have
>> some space on another machine that we can move the VM to, so we'll be
>> looking at doing that when things have calmed down a bit after getting
>> back from PGDay.EU. This will cause some short downtime as DNS
>> switches over, so we'll post a note on exactly when we plan to do it,
>> once it's planned.
>
> Sounds like a plan.  Thanks.
>
> (BTW, since it's just a read-only clone of master, couldn't you avoid
> downtime by duplicating the VM and running two in parallel until the DNS
> change propagates fully?  Or are you just thinking it's not worth the
> trouble?)

It's not just that. It also runs git hosting for a bunch of projects,
including but certainly not limited to pgadmin and slony, where it is
the master.


--
 Magnus Hagander
 Me: http://www.hagander.net/
 Work: http://www.redpill-linpro.com/