On Jun 7, 12:44 pm, adrian.kla...@gmail.com (Adrian Klaver) wrote:
> On 06/07/2012 06:57 AM, Dinsdale wrote:
> > Hi there. I am new to Postgresql but very familiar with RDBMS. We are
> > running 8.4 on windows serer 2008 r2. I was updating a record through
> > pgAdmin and was being an a$$ and pressed the run button multiple times
> > and then the record just disappeared. Update query:
> >
> > update jaas_login set password=md5(login_name || ':' || realm_name ||
> > ':Password123'), has_logged_in = false, modified=now(),
> > failed_attempts=0 where login_name like 'CompanyAdmin@%';
>
> So just to be clear the above was what was run multiple times?
>
> >
> > I cannot find the record at all so tried to re-insert it and the query
> > just runs and runs and runs.
>
> Select * from jaas_login where login_name like CompanyAdmin@%'; shows
> nothing?
>
> What is your INSERT query?
> >
> > We thought we had backups of this DB but all we have are dump files
> > and I really really don't want to have to restore if I can avoid it. I
> > find it hard to believe that a db as advanced as postgresql will just
> > "lose" data and there has to be some explaination for where the record
> > is.
>
> I tend to doubt it just disappeared.
> >
> > Any help would really save my butt. Thanks
> >
> > Dinsdale
> >
>
> --
> Adrian Klaver
> adrian.kla...@gmail.com
>
> --
> Sent via pgsql-general mailing list (pgsql-gene...@postgresql.org)
> To make changes to your subscription:http://www.postgresql.org/mailpref/pgsql-general
I tried every select query I could think of including with no where
clause and simply searching manually for the record but it was just
not there. I had someone else verify this as well. I didn't think it
could just disappear either. We finally decided to use the dump file
and re-create the database but just before that I tried a backup and
got an error about a missing attribute on relid xxxx. Someone with
more Postgres experience said that the table was corrupted and i
should just rebuild, so we did.
Please note that I CANNOT reproduce this in our QA! I think maybe the
issue noted by steve below may be the problem.