On Sep 29, 2008, at 9:38 AM, hshah wrote:
Graphs for all my devices have gaps in them, this usually seems to
happen around the time when zenperfsnmp logs shows following logs
for one or 2 servers. The problem is that the graphs for all the
servers/routers end up with gaps at regular intervals incl the
problem devices(their graphs are worst).. I have two questiions
- What could be wrong with the problem server for zenoss to throws
the following errors
INFO zen.zenperfsnmp: total=17 good=16 bad=0 time=39.096727
2008-09-29 23:03:28 WARNING zen.zenperfsnmp: There are still 1
devices to query
2008-09-29 23:03:28 WARNING zen.zenperfsnmp: Problem devices: [server]
2008-09-29 23:03:28 WARNING zen.zenperfsnmp: Waiting one more cycle
period
2008-09-29 23:08:28 WARNING zen.zenperfsnmp: There are still 1
devices to query
2008-09-29 23:08:28 WARNING zen.zenperfsnmp: Problem devices:
[server']
2008-09-29 23:08:28 WARNING zen.zenperfsnmp: Devices status is not
clearing. Restarting.
- Is there any way to make the other graphs works while still
monitoring these "problem devices"?
My install is the latest 2..2.3 install and i am using the
ethernet64 templates, the server is a dual core 2.5ghz new server.
I have also tried deleting the pickle files from the the problematic
servers to no avail.
The logic is a bit lacking right now. However, there is a ticket open
to address it already. In the meantime you really need to figure out
why these problem devices are taking so long to respond to SNMP, and
failing that you will need to set them to zSnmpMonitorIgnore or
they'll affect your entire collection cycle.
Ticket #3567: zenperfsnmp "Waiting one more cycle period" needs to be
smarter
http://dev.zenoss.org/trac/ticket/3567
_______________________________________________
zenoss-users mailing list
[email protected]
http://lists.zenoss.org/mailman/listinfo/zenoss-users
