Re: [ADMIN] Error seen when vacuuming pg_largeobject table

2004-01-28 Thread Chris White (cjwhite)
Tom, It happened again. After doing the reindex, I did a vacuum full on the pg_largeobject table. This time I did not get the TUPLE INDEX error but the server terminated at the end of the vacuum command. The backend is still running even though I get the error message: aesop=# vacuum full verbose

Re: [ADMIN] How to determine which file contains which block

2004-01-28 Thread Tom Lane
Frank Way <[EMAIL PROTECTED]> writes: > After finding the ctid in (block,tuple) format, how do you know which > file in all the subdirectories under /usr/local/pgsql/data/base > contains that block? The subdirectories are named after the OIDs of their databases (pg_database.oid column). Individua

Re: [ADMIN] Error seen when vacuuming pg_largeobject table

2004-01-28 Thread Gaetano Mendola
Chris White (cjwhite) wrote: > Sorry no core files. The system is running with cores turned off. Next > time I will turn on cores prior to trying to debug this. This is not the first time that a usefull core go wasted. Happen to me twice till I realized to modify the script /etc/init.d/postgres i

Re: [ADMIN] Error seen when vacuuming pg_largeobject table

2004-01-28 Thread Chris White (cjwhite)
Sorry no core files. The system is running with cores turned off. Next time I will turn on cores prior to trying to debug this. Chris -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Tom Lane Sent: Wednesday, January 28, 2004 11:52 AM To: [EMAIL PROTECTED]

[ADMIN] How to determine which file contains which block

2004-01-28 Thread Frank Way
Hi all, I've recently had the opportunity to use pg_filedump to try and find some corruption in one of my database tables (7.3.4). If found the following description of how to find the block/tuple and to use pg_filedump: http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&threadm=20030922162322.E1

[ADMIN] log file: showing disconnects as well as connects?

2004-01-28 Thread Mark Harrison
Here's a quickie script to watch a particular session: tail -f $LOG | awk ' /connection received: host='$1'/ { pid = $3 } $3 == pid { print }' Is there something in the log file which will indicate a disconnect as well? Many TIA, Mark ---(end of broadcast)--

Re: [ADMIN] Error seen when vacuuming pg_largeobject table

2004-01-28 Thread Tom Lane
> Got it to happen again after a power fail. Here is the logs this time > with verbose. Hm, those backend crashes should have left core dumps; can you get a stack trace from the core? regards, tom lane ---(end of broadcast)-

Re: [ADMIN] Error seen when vacuuming pg_largeobject table

2004-01-28 Thread Chris White (cjwhite)
Tom, Got it to happen again after a power fail. Here is the logs this time with verbose. bash-2.05b$ ./postgres -D /t/data/sql -P -O aesop POSTGRES backend interactive interface $Revision: 1.245.2.2 $ $Date: 2002/02/27 23:17:01 $ backend> reindex index pg_largeobject_loid_pn_index; backend> b