From: Echn Athon (athon_ra@web.de)
Date: Fri Sep 03 2004 - 09:12:11 EDT
Hi all,
we have build a product that heavily relies on generic data services under Sun Cluster 3.0/3.1.
Our gds are configured with a Thorough_Probe_Intervall of 60 seconds,
but at some instances we discover our Probe_command being invoked up to 5 times per minute.
What we found out so far:
1) Configuring a port-list with additional spare ports for future use, is no good idea.
It seems as if the system does additional probing for every port no process is listening on.
2) The first entry in port-list seems to have a special meaning: if no process listens on this port,
the cluster software is changing the status of this ressource-group to degraded.
3) Even if all ports in port-list are listened on, the real probing intervall may be well below the
configured 60 seconds.
This behaviour is independent of the return code of the Probe_command (0 or 100).
Is anybody out there in the know of
- what may be the problem or
- what different causes may lead to additional probing or
- what manual may cover these issues in detail.
Any hints are appreciated. Thanks in advance
martin
________________________________________________________________
Verschicken Sie romantische, coole und witzige Bilder per SMS!
Jetzt neu bei WEB.DE FreeMail: http://freemail.web.de/?mc=021193
_______________________________________________
sunmanagers mailing list
sunmanagers@sunmanagers.org
http://www.sunmanagers.org/mailman/listinfo/sunmanagers
This archive was generated by hypermail 2.1.7 : Wed Apr 09 2008 - 23:29:23 EDT