On 12/23/12 3:17 PM, Simon Riggs wrote:
> If that last change was the cause, then its caused within VACUUM. I'm
> running a thrash test with autovacuums set much more frequently but
> nothing yet.
I am not very suspicious of that VACUUM change; just pointed it out for
completeness sake.
> Are you by any chance running with unlogged tables? There was a change
> there recently, something around checkpoint IIRC. Can you set
> checkpoints more frequent also?
This is straight pgbench, modifying only scale, client count, workers,
and duration. No unlogged tables involved.
The crash is so intermittent that I'm trying not to change anything
until I get more productive output out of one. The last run I kicked
off has been chugging for 14 hours without a hiccup yet.
--
Greg Smith 2ndQuadrant US greg@2ndQuadrant.com Baltimore, MD
PostgreSQL Training, Services, and 24x7 Support www.2ndQuadrant.com