On Wed, 17 May 2006, Tom Lane wrote:
I have a lurking feeling that the still-hypothetical connection between
archiver and foreground operations might come into operation at pg_clog
page boundaries (which require emitting XLOG events) --- that is, every
32K transactions something special happens. The time delay between
archiver wedging and foreground wedging would then correspond to how
long it took the XID counter to reach the next 32K multiple. (Jeff,
what transaction rate do you see on that server --- is that a plausible
delay for some thousands of transactions to pass?)
This is just a guess, but if you check the archives for Chris K-L's
out-of-disk-space server meltdown a year or three ago, you'll see
something similar.
This sounds interesting and I'll go have a look for this thread in the
archives.
It looks like it's running about 3,000 transactions per minute right now
(around 10pm). I got that by doing the following...tell me if this is
reasonable:
db3:~ $ echo select xact_commit from pg_stat_database where datname =
\'vb_web\'\; | psql vb_web ; sleep 60 ; echo select xact_commit from
pg_stat_database where datname = \'vb_web\'\; | psql vb_web
xact_commit
-------------
806325
(1 row)
xact_commit
-------------
809414
(1 row)
So, doing a little subtraction, I got 3,089 transactions for the 60 second
period I ran it. I would guess it's quite a bit more during the day. That
means it would hit 32k transactions around 10-11 minutes...does that fit in
with our timing from above? (Tom, I'm not sure which timing you're
using.)
--
Jeff Frost, Owner <jeff@xxxxxxxxxxxxxxxxxxxxxx>
Frost Consulting, LLC http://www.frostconsultingllc.com/
Phone: 650-780-7908 FAX: 650-649-1954
