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

On Sun, Jun 12, 2011 at 3:01 PM, Noah Misch <noah(at)leadboat(dot)com> wrote:
> I fully agree.  That said, if this works on the standby, we may as well also use
> it opportunistically on the master, to throttle bloat.

As long as the performance cost is de minimis, I agree.

>> At any rate, if taking a cleanup lock on the right-linked page on the
>> standby is sufficient to fix the problem, that seems like a far
>> superior solution in any case.  Presumably the frequency of someone
>> having a pin on that particular page will be far lower than any
>> matching based on XID or heavyweight locks.  And the vast majority of
>> such pins should disappear before the startup process feels obliged to
>> get out its big hammer.
>
> Yep; looks promising.
>
> Does such a thing have a chance of being backpatchable?  I think the chances
> start slim and fall almost to zero on account of the difficulty of avoiding a
> WAL format change.

I can't see back-patching it. Maybe people would feel it was worth
considering if we were getting legions of complaints about this
problem, but so far you're the only one. In any case, back-patching a
WAL format change is a complete non-starter -- we can't go making
minor versions non-interoperable.

> Assuming that conclusion, I do think it's worth starting
> with something simple, even if it means additional bloat on the master in the
> wal_level=hot_standby + vacuum_defer_cleanup_age / hot_standby_feedback case.
> In choosing those settings, the administrator has taken constructive steps to
> accept master-side bloat in exchange for delaying recovery conflict.  What's
> your opinion?

I'm pretty disinclined to go tinkering with 9.1 at this point, too.

--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company