Re: On-the-fly index tuple deletion vs. hot_standby

On Fri, 2010-12-03 at 21:43 +0200, Heikki Linnakangas wrote:
> On 29.11.2010 08:10, Noah Misch wrote:
> > I have a hot_standby system and use it to bear the load of various reporting
> > queries that take 15-60 minutes each.  In an effort to avoid long pauses in
> > recovery, I set a vacuum_defer_cleanup_age constituting roughly three hours of
> > the master's transactions.  Even so, I kept seeing recovery pause for the
> > duration of a long-running query.  In each case, the culprit record was an
> > XLOG_BTREE_DELETE arising from on-the-fly deletion of an index tuple.  The
> > attached test script demonstrates the behavior (on HEAD); the index tuple
> > reclamation conflicts with a concurrent "SELECT pg_sleep(600)" on the standby.
> >
> > Since this inserting transaction aborts, HeapTupleSatisfiesVacuum reports
> > HEAPTUPLE_DEAD independent of vacuum_defer_cleanup_age.  We go ahead and remove
> > the index tuples.  On the standby, btree_xlog_delete_get_latestRemovedXid does
> > not regard the inserting-transaction outcome, so btree_redo proceeds to conflict
> > with snapshots having visibility over that transaction.  Could we correctly
> > improve this by teaching btree_xlog_delete_get_latestRemovedXid to ignore tuples
> > of aborted transactions and tuples inserted and deleted within one transaction?

@Noah Easily the best bug reported submitted in a long time. Thanks.

> Seems reasonable. HeapTupleHeaderAdvanceLatestRemovedXid() will need 
> similar treatment. Actually, btree_xlog_delete_get_latestRemovedXid() 
> could just call HeapTupleHeaderAdvanceLatestRemoveXid().

Yes, it applies to other cases also. Thanks for the suggestion.

Fix committed. Please double-check my work, committed early since I'm
about to jump on a plane.

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