Re: Scaling XLog insertion (was Re: Moving more work outside WALInsertLock)

Tom Lane <tgl@sss.pgh.pa.us>

From: Tom Lane <tgl@sss.pgh.pa.us>

To: Alvaro Herrera <alvherre@commandprompt.com>

Cc: Simon Riggs <simon@2ndquadrant.com>, Heikki Linnakangas <heikki.linnakangas@enterprisedb.com>, Fujii Masao <masao.fujii@gmail.com>, Jeff Janes <jeff.janes@gmail.com>, Robert Haas <robertmhaas@gmail.com>, Pg Hackers <pgsql-hackers@postgresql.org>

Date: 2012-03-07T15:04:19Z

Lists: pgsql-hackers

Alvaro Herrera <alvherre@commandprompt.com> writes:
> Just to keep things in perspective -- For a commit record to reach one
> megabyte, it would have to be a transaction that drops over 43k tables.
> Or have 64k smgr inval messages (for example, a TRUNCATE might send half
> a dozen of these messages). Or have 262k subtransactions.  Or
> combinations thereof.

> Now admittedly, a page is only 8 kB, so for a commit record to be "many
> pages long" (that is, >=3) it would require about 1500 smgr inval
> messages, or, say, about 250 TRUNCATEs (of permanent tables with at
> least one toastable field and at least one index).

What about the locks (if running hot-standby)?

> So they are undoubtely rare.  Not sure if as rare as Higgs bosons.

Even if they're rare, having a major performance hiccup when one happens
is not a side-effect I want to see from a patch whose only reason to
exist is better performance.

			regards, tom lane