Re: XMin Hot Standby Feedback patch

On Wed, 2011-02-16 at 11:25 +0900, Fujii Masao wrote:
> On Wed, Feb 16, 2011 at 6:15 AM, Simon Riggs <simon(at)2ndquadrant(dot)com> wrote:
> > On Tue, 2011-02-15 at 12:15 -0500, Robert Haas wrote:
> >> Looks pretty good to me, though I haven't tested it. I like some of
> >> the safety valves you put in there, but I don't understand this part
> >
> > Reworked logic covering all feedback, plus tests, plus docs.
> >
> > Last comments before commit please.
>
> When I started the standby with hot_standby = off and hot_standby_feedback = on,
> I got the following assertion error.
>
> -----------------
> sby LOG: streaming replication successfully connected to primary
> TRAP: FailedAssertion("!(((result) >= ((TransactionId) 3)))", File:
> "procarray.c", Line: 1027)
> act LOG: unexpected EOF on standby connection
> sby LOG: WAL receiver process (PID 17572) was terminated by signal 6: Aborted
> sby LOG: terminating any other active server processes
> -----------------

Thanks

> vacuum_defer_cleanup_age on the *standby* should not affect the
> feedback xid.

Not sure, will think some more.

> VACUUM TABLE on the *primary* doesn't use the feedback xid at all.
> Is this intentional?

Yes, I was in the middle of fixing that.

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