Thread: relation 71478240 deleted while still in use on 8.1
I got : 2007-03-21 20:56:17 CET ERROR: relation 71478240 deleted while still in use 2007-03-21 20:56:17 CET ERROR: relation 71478240 deleted while still in use 2007-03-21 20:56:17 CET ERROR: relation 71478240 deleted while still in use on 8.1.8 my production server here. some inserts/updates, selects, and pg_dump was running. there's a chance cluster was running in http://dupa.privatepaste.com/ d317KYXt5n It is repeatable, if I run pg_dump. Guess that has something to do, with slowing down db server - and some race cond occurs. Any ideas ? should I add it as a bug ? or perhaps someone already knows about that kinda issue -- Grzegorz Jaskiewicz C/C++ freelance for hire
Grzegorz Jaskiewicz wrote: > I got : > > 2007-03-21 20:56:17 CET ERROR: relation 71478240 deleted while still in > use > 2007-03-21 20:56:17 CET ERROR: relation 71478240 deleted while still in > use > 2007-03-21 20:56:17 CET ERROR: relation 71478240 deleted while still in > use > > on 8.1.8 my production server here. > some inserts/updates, selects, and pg_dump was running. > there's a chance cluster was running in > http://dupa.privatepaste.com/d317KYXt5n > > It is repeatable, if I run pg_dump. Guess that has something to do, with > slowing down db server - and some race cond occurs. > Any ideas ? > should I add it as a bug ? or perhaps someone already knows about that > kinda issue Which transaction gets the error, pg_dump or the one you copied to that web page? Is there anything else running? How often does it happen, always? Can you reduce it to a self-contained test case that doesn't require dblink? -- Heikki Linnakangas EnterpriseDB http://www.enterprisedb.com
Grzegorz Jaskiewicz <gj@pointblue.com.pl> writes:
> should I add it as a bug ?
Only if you can reproduce it in 8.2 --- what it looks like to me is a
relcache-opening race condition, which is (believed) fixed in 8.2 and
is not feasible to fix in older branches.
regards, tom lane
On Mar 21, 2007, at 11:25 PM, Tom Lane wrote: > Grzegorz Jaskiewicz <gj@pointblue.com.pl> writes: >> should I add it as a bug ? > > Only if you can reproduce it in 8.2 okie dokie, I am moving the DB onto 8.2 next week (120M of rows, and 8.2 does sorting much faster). -- Grzegorz Jaskiewicz C/C++ freelance for hire
okay, I got it. The main reason behind it - is that I do drop table in transaction. Every 10 minutes. So during that period, when 'replication' is running - the thing becomes unstable, and this error can appear. -- Grzegorz Jaskiewicz C/C++ freelance for hire