Re: ERROR: Index pg_toast_8443892_index is not a btree - Mailing list pgsql-hackers

From strk
Subject Re: ERROR: Index pg_toast_8443892_index is not a btree
Date
Msg-id 20031208172303.A57240@freek.keybit.net
Whole thread Raw
In response to Re: ERROR: Index pg_toast_8443892_index is not a btree  (Jan Wieck <JanWieck@Yahoo.com>)
Responses Re: ERROR: Index pg_toast_8443892_index is not a btree
Re: ERROR: Index pg_toast_8443892_index is not a btree
Re: ERROR: Index pg_toast_8443892_index is not a btree
List pgsql-hackers
JanWieck wrote:
> strk wrote:
> 
> > I get the following error when vacuuming a db or inserting
> > a big value in a column of a toastable datatype (GEOMETRY).
> > 
> >     ERROR:  Index pg_toast_8443892_index is not a btree
> > 
> > My last action has been killing a psql that was getting
> > mad about receiving too much input and beeping as hell
> > (readline issue ?).
> 
> You must have killed a lot more than your psql frontend to get that as a 
> result.

really... I hit ^C at the psql terminal and
kill -9 <psql_pid>

The only other reason I can thing about is the data type text
input function screwing pg internal pointers...

> 
> > 
> > After that, I put the insert query I was trying to feed to
> > psql in a file and sourced it... Bump! that error appeared.
> > 
> >     psql:B:477: ERROR:  Index pg_toast_8443892_index is not a btree
> > 
> > Line 477 is EOF...
> > 
> > Vacuum does not solve this (as you can see in the first error message)
> > 
> > Do you have any hint about how to fix this ? 
> 
> Try "reindex table <tablename>". If you really only lost that btree 
> index, that should do.

gis=# reindex table test; -- this is the table I was trying to insert into
WARNING:  table "test" wasn't reindexed
REINDEX
gis=# reindex table pg_toast_8443892; -- this was an assuption I made
ERROR:  Relation "pg_toast_8443892" does not exist
gis=# vacuum;
ERROR:  Index pg_toast_8443892_index is not a btree
gis=#

Where could this pg_toast_8443892_index reference be found ?

--strk;


pgsql-hackers by date:

Previous
From: Andrew Dunstan
Date:
Subject: Re: CVS HEAD compile failure
Next
From: Tom Lane
Date:
Subject: Re: CVS HEAD compile failure