From: | Christoph Berg <christoph(dot)berg(at)credativ(dot)de> |
---|---|
To: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
Cc: | Kevin Grittner <Kevin(dot)Grittner(at)wicourts(dot)gov>, pgsql-hackers(at)postgresql(dot)org, Joachim Wieland <joe(at)mcknight(dot)de> |
Subject: | Re: Notify system doesn't recover from "No space" error |
Date: | 2012-07-02 15:55:38 |
Message-ID: | 20120702155538.GG8539@msgid.df7cb.de |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
Re: Tom Lane 2012-06-29 <8308(dot)1340933751(at)sss(dot)pgh(dot)pa(dot)us>
> in async.c that would try to create the missing page. This error will
> persist until the server is restarted (thus resetting the state of
> pg_notify), even if the underlying disk-full condition is cleared.
Thanks for digging into this and providing a patch. I've applied it to
9.1.4-2 here, and now the server recovers more gracefully from the
error: (sorry for the German locale, the error is ENOSPC)
2012-07-02 17:45:59 CEST FEHLER: konnte auf den Status von Transaktion 0 nicht zugreifen
2012-07-02 17:45:59 CEST DETAIL: Konnte nicht in Datei »pg_notify/0000« bei Position 32768 schreiben: Auf dem Gerät ist kein Speicherplatz mehr verfügbar.
2012-07-02 17:45:59 CEST ANWEISUNG: commit
A subsequent "notify foobar" command did succeed after clearing the
disk space.
What is still puzzling me is that the customer is repeatedly reporting
these issues, even after rebooting the system. I am not 100% sure at
which point the disk full situation was resolved, but at least it
sounds like that problem would reappear after restarts. I will need to
get more detailed reports on the order of events there to make sure.
Christoph
--
cb(at)df7cb(dot)de | http://www.df7cb.de/
From | Date | Subject | |
---|---|---|---|
Next Message | Tom Lane | 2012-07-02 16:03:55 | Re: Notify system doesn't recover from "No space" error |
Previous Message | David Fetter | 2012-07-02 15:49:05 | Re: [COMMITTERS] pgsql: Make walsender more responsive. |