we've resynced slave and now everything is OK, thanks you for help.
And only one last question, where to read about this bug, because my colleges want to know, what happens.
On Fri, Apr 13, 2012 at 00:48, Merlin Moncure <mmoncure@xxxxxxxxx> wrote:
On Thu, Apr 12, 2012 at 9:20 AM, Ivan Evtuhovich <evtuhovich@xxxxxxxxx> wrote:
> Hello,Per some off-list conversation with Ivan, this is only happening on
>
> More then month ago we upgrade DB from 9.0 to 9.1 with pg_upgrade. Then we
> move DB to another server with standard pg streaming replication.
>
> Now we have two entries with the same primary key. And I do not know what to
> do.
>
> SELECT ctid, id from billing_invoices where id = 27362891;
>
> ctid | id
> --------------+----------
> (1112690,11) | 27362891
> (1112438,26) | 27362891
>
the standby. Ivan, what's the precise version of postgres you are
using? When you first went to hs/sr? I bet your clog files are out
of whack (and if so, probably upgrading to recent bugfix postgres and
standby resync is the correct course of action).
merlin
