Re: Verified fix for Bug 4137

Поиск
Список
Период
Сортировка
От Simon Riggs
Тема Re: Verified fix for Bug 4137
Дата
Msg-id 1210344736.4268.653.camel@ebony.site
обсуждение исходный текст
Ответ на Re: Verified fix for Bug 4137  ("Heikki Linnakangas" <heikki@enterprisedb.com>)
Список pgsql-patches
On Fri, 2008-05-09 at 15:37 +0100, Heikki Linnakangas wrote:
> Simon Riggs wrote:
> >       if (restartWALFileName)
> >       {
> > +         /*
> > +          * Don't do cleanup if the restartWALFileName provided
> > +          * is later than the xlog file requested. This is an error
> > +          * and we must not remove these files from archive.
> > +          * This shouldn't happen, but better safe than sorry.
> > +          */
> > +         if (strcmp(restartWALFileName, nextWALFileName) > 0)
> > +             return false;
> > +
> >           strcpy(exclusiveCleanupFileName, restartWALFileName);
> >           return true;
> >       }
>
> I committed this sanity check into pg_standy, though it really shouldn't
> happen, but it just occurred to me that the most likely reason for that
> to happen is probably that the user has screwed up his restore_command
> line, mixing up the  %p and %r arguments. Should we make that an error
> instead of just not doing the cleanup?

You can't explicitly throw a pgsql error in pg_standby, so the best we
can do is get the file requested if it exists. If the file is the wrong
one then recovery will throw the error. As long as we didn't delete
anything when that happens we can just correct the mistake and retry.

--
  Simon Riggs
  2ndQuadrant  http://www.2ndQuadrant.com


В списке pgsql-patches по дате отправления:

Предыдущее
От: Tom Lane
Дата:
Сообщение: Re: [GENERAL] pgbench not setting scale size correctly?
Следующее
От: Tom Lane
Дата:
Сообщение: Re: New flex warnings