Re: Hot Standby, release candidate?

Simon Riggs <simon@2ndquadrant.com>

From: Simon Riggs <simon@2ndQuadrant.com>
To: Peter Eisentraut <peter_e@gmx.net>
Cc: pgsql-hackers@postgresql.org
Date: 2009-12-17T10:08:36Z
Lists: pgsql-hackers
On Thu, 2009-12-17 at 12:01 +0200, Peter Eisentraut wrote:
> On sön, 2009-12-13 at 19:20 +0000, Simon Riggs wrote:
> > Barring resolving a few points and subject to even more testing, this
> > is the version I expect to commit to CVS on Wednesday.
> 
> So it's Thursday now.  Please keep us updated on the schedule, as we
> need to decide when to wrap alpha3 and whether to reopen the floodgates
> for post-CF commits.

I've been looking at a semaphore deadlock problem reported by Hiroyuki
Yamada. It's a serious issue, though luckily somewhat restricted in
scope.

I don't think its wise to rush in with a solution, since that involves
some work with semaphores and I could easily make that area less stable
by acting too quickly.

What I will do now is put in a restriction on lock waits in Hot Standby,
which will only happen for Alpha3. That avoids the deadlock issue in an
easy and safe way, though it is heavy handed and I aim to replace it
fairly soon, following discussion and testing.

-- 
 Simon Riggs           www.2ndQuadrant.com