We have used the jabber dispatcher since december. We soon discovered that our XMPP-server implementation (Openfire) sometimes freeze. It does not crash completely, but takes forever to answer connections. When this freeze happens and NAV tries to send an alert via jabber, the whole alertqueue clogs up and no alerts gets sent at all - no email, no sms, nothing. When the jabber-daemon is restarted all the backed up alerts are sent right away.
(This of course happened during the holidays and we have disabled the jabber dispatcher ever since.)
Is it possible to add a timeout or something to jabber_dispatcher.py to prevent this to happen?
--Ingeborg
Not entirely related, but I might as well mention it..
I hope the developers can add some kind of logics to supress queued SMS messages in the alert queue after restarting smsd after a crash. We've had situations where smsd has pushed out 100-200 SMS messages after a restart, often more than 20 messages to one mobile phone.
Are there any modifications planned to maybe supress queued up SMS messages after a crash?
Best regards,
Vidar S
-----Original message----- Fra: Ingeborg Hellemo [mailto:ingeborg.hellemo@uit.no] Sendt: 25. januar 2010 12:46 Til: nav-users@uninett.no Emne: Jabber dispatcher clogs the whole alertqueue
We have used the jabber dispatcher since december. We soon discovered that our XMPP-server implementation (Openfire) sometimes freeze. It does not crash completely, but takes forever to answer connections. When this freeze happens and NAV tries to send an alert via jabber, the whole alertqueue clogs up and no alerts gets sent at all - no email, no sms, nothing. When the jabber-daemon is restarted all the backed up alerts are sent right away.
(This of course happened during the holidays and we have disabled the jabber dispatcher ever since.)
Is it possible to add a timeout or something to jabber_dispatcher.py to prevent this to happen?
--Ingeborg
Ingeborg Østrem Hellemo -- ingeborg.hellemo@uit.no Dep. of Information Technology --- Univ. of Tromsø
On Mon, Jan 25, 2010 at 12:45:31PM +0100, Ingeborg Hellemo wrote:
Is it possible to add a timeout or something to jabber_dispatcher.py to prevent this to happen?
Created https://bugs.launchpad.net/nav/+bug/515400 for this issue and subscribed you to the bug. Should have a workaround or fix within the day :)
On Tue, Jan 26, 2010 at 04:00:37PM +0100, Stokkenes Vidar wrote:
Are there any modifications planned to maybe supress queued up SMS messages after a crash?
Created https://bugs.launchpad.net/nav/+bug/515403 to address this issue.
On Tue, 26 Jan 2010 16:00:37 +0100 Stokkenes Vidar Vidar.Stokkenes@hn-ikt.no wrote:
Not entirely related, but I might as well mention it..
I hope the developers can add some kind of logics to supress queued SMS messages in the alert queue after restarting smsd after a crash. We've had situations where smsd has pushed out 100-200 SMS messages after a restart, often more than 20 messages to one mobile phone.
Are there any modifications planned to maybe supress queued up SMS messages after a crash?
All queued SMS messages can be suppressed manually prior to a restart using the -c option to smsd.py See "smsd.py -h" for more help on command line options to the daemon.
-
Ingeborg Hellemo -
Morten Brekkevold -
Stokkenes Vidar -
Thomas K. Adamcik -
Thomas K. Adamcik