Обсуждение: Column pgl.transaction does not exist in 8.3 beta 4

Just upgraded from 8.2.5 to 8.3b4 on Windows and after reimporting my database I noticed the following messages are
showingup sporadically in the server  logs: 

2007-12-07 11:56:17 CST ERROR:  column pgl.transaction does not exist at character 171
2007-12-07 11:56:17 CST STATEMENT:  SELECT (SELECT datname FROM pg_database WHERE oid = pgl.database) AS dbname,
pgl.relation::regclassAS class, pg_get_userbyid(pg_stat_get_backend_userid(svrid)) as user, pgl.transaction,
pg_stat_get_backend_pid(svrid)AS pid, pgl.mode, pgl.granted, pg_stat_get_backend_activity(svrid) AS current_query,
pg_stat_get_backend_activity_start(svrid)AS query_start FROM pg_stat_get_backend_idset() svrid, pg_locks pgl WHERE
pgl.pid= pg_stat_get_backend_pid(svrid) ORDER BY pid; 

Looks like it should be referencing pgl.transactionid instead?  Anyone have any idea as to what is generating this
statement?

Regards,

Shelby Cain




      ____________________________________________________________________________________
Never miss a thing.  Make Yahoo your home page.
http://www.yahoo.com/r/hs

Shelby Cain <alyandon@yahoo.com> writes:
> Just upgraded from 8.2.5 to 8.3b4 on Windows and after reimporting my
> database I noticed the following messages are showing up sporadically
> in the server logs:
> 2007-12-07 11:56:17 CST ERROR:  column pgl.transaction does not exist at character 171

> Looks like it should be referencing pgl.transactionid instead?

Most likely you want pg_locks.virtualtransaction, instead --- that's
the closest match to the old behavior of pg_locks.transaction.

> Anyone have any idea as to what is generating this statement?

Nope.

            regards, tom lane

> ------- Original Message -------
> From: Tom Lane <tgl@sss.pgh.pa.us>
> To: Shelby Cain <alyandon@yahoo.com>
> Sent: 08/12/07, 01:21:08
> Subject: Re: [GENERAL] Column pgl.transaction does not exist in 8.3 beta 4
>
> Most likely you want pg_locks.virtualtransaction, instead --- that's
> the closest match to the old behavior of pg_locks.transaction.
>
> > Anyone have any idea as to what is generating this statement?

Sounds like a now-fixed bug in pgAdmin 1.8.0's server status monitor.

Regards, Dave