Re: New sessions on a database to be dropped consume 100% cpu - Mailing list pgsql-bugs

From Feike Steenbergen
Subject Re: New sessions on a database to be dropped consume 100% cpu
Date
Msg-id CAK_s-G3SP71oabtJ2F_Wt-WEUk6YwNF2W+VRO8KHk57bDwfJ1Q@mail.gmail.com
Whole thread Raw
In response to Re: New sessions on a database to be dropped consume 100% cpu  (Tom Lane <tgl@sss.pgh.pa.us>)
Responses Re: New sessions on a database to be dropped consume 100% cpu  (Peter Geoghegan <pg@bowt.ie>)
List pgsql-bugs
> If you could install amcheck and see what it thinks about those indexes,
> that would be interesting.

I created a backup (using pg_basebackup) from this cluster and had a look:

name_to_be_dropped=# select bt_index_check('pg_catalog.pg_depend_depender_index');
ERROR:  cross page item order invariant violated for index "pg_depend_depender_index"
DETAIL:  Last item on page tid=(255,138) page lsn=33A5/F47BC3B0.

> Of course, if you're just wishing you could get rid of the database,
> you could forcibly kill those processes and move on.

Unfortunately other databases are still serving a useful purpose on the same cluster
and it seems my only recourse would be restart the whole cluster, as I currently can't
kill them using SIGTERM.

Thanks,

Feike

pgsql-bugs by date:

Previous
From: Thomas Munro
Date:
Subject: Re: BUG #15520: PAM authentication + domain socket -> DNS query forsymbolic hostname [local]
Next
From: Peter Eisentraut
Date:
Subject: Re: BUG #15520: PAM authentication + domain socket -> DNS query forsymbolic hostname [local]