Re: race condition in sync rep

On Sat, Mar 26, 2011 at 3:00 PM, Robert Haas <robertmhaas(at)gmail(dot)com> wrote:
> On Mar 26, 2011, at 6:16 AM, Simon Riggs <simon(at)2ndquadrant(dot)com> wrote:
>> That seems very unlikely even with fsync=off in a real config where we
>> have network path to consider.
>>
>> Your definition of a "nasty" race condition seems off.
>>
>> I've added code for you.
>
> Your skepticism seems out of place. I actually hit this problem in testing. We could debate how realistic my test setup was, but why?  It is not our policy - and bad practice in general - to ship code with race conditions.  The code you added is not "for me"; it's "to fix a race condition".

A race that will not be lost with normal and sensible settings and
results in a slight performance degradation only in a contrived worse
case scenario. There is no question that it caused any other problem.

> Thanks for the quick fix.

--
 Simon Riggs                   http://www.2ndQuadrant.com/
 PostgreSQL Development, 24x7 Support, Training & Services