Re: [GENERAL] large object does not exist after pg_migrator

Jamie Fox wrote:
> Hi -
> After what seemed to be a normal successful pg_migrator migration from 8.3.7
> to 8.4.0, in either link or copy mode, vacuumlo fails on both our production
> and qa databases:
>
> Jul 1 11:17:03 db2 postgres[9321]: [14-1] LOG: duration: 175.563 ms
> statement: DELETE FROM vacuum_l WHERE lo IN (SELECT "xml_data" FROM
> "public"."xml_user")
> Jul 1 11:17:03 db2 postgres[9321]: [15-1] ERROR: large object 17919608
> does not exist
> Jul 1 11:17:03 db2 postgres[9321]: [16-1] ERROR: current transaction is
> aborted, commands ignored until end of transaction block
>
> I migrated our qa database using pg_dump/pg_restore and vacuumlo has no
> problem with it. When I try querying the two databases for large objects
> manually I see the same error in the one that was migrated with pg_migrator:
>
> select loread(lo_open(xml_data,262144),1073741819) from xml_user where id =
> '10837246';
> ERROR: large object 24696063 does not exist
> SQL state: 42704
>
> I can also see that the pg_largeobject table is different, in the pg_restore
> version the Rows (estimated) is 316286 and Rows (counted) is the same, in
> the pg_migrator version the Rows (counted) is only 180507.
>
> Any advice on what I might look for to try and track down this problem?
> pg_restore on our production database takes too long so it would be really
> nice to use pg_migrator instead.

[ Email moved to hackers list.]

Wow, I didn't test large objects specifically, and I am confused why
there would be a count discrepancy. I will need to do some research
unless someone else can guess about the cause.

--
Bruce Momjian <bruce(at)momjian(dot)us> http://momjian.us
EnterpriseDB http://enterprisedb.com

+ If your life is a hard drive, Christ can be your backup. +