On Oct 15, 2008, at 9:08 AM, xzp5jj wrote:
I am running Zenoss 2.2 on CentOS 5.1. I don't know if this is a bug
or just normal operation for zenperfsnmp. I searched through the
forum and the Zenoss docs looking for the answer but couldn't find
anything. On some of my DNS servers, I have a pass statement in my
snmpd.conf that calls a perl script to parse the named.stats file
and populates the NET-SNMP MIB with traffic data that I then pull
and graph via a template. Occasionally I have an error where the
perl script hiccups and the OIDs I am pulling are not populated. The
script problem is my riddle to solve, but once zenperfsnmp gets a
"bad OID" error, the daemon will not pull data from those OIDs, even
though snmpd has subsequently executed the pass statement
successfully and the OIDs are now populated. The only way I have
found to "kick start" zenperfsnmp into polling those OID's again is
to restart zenperfsnmp. As I previously stated my questions are :
Is this a bug or normal operation for zenperfsnmp ?
If it is normal operation is there an option or magic switch that I
can set to tell zenperfsnmp to continue collecting on OID's after a
"bad OID" error is encountered ?
I haven't tested this, but a quick look at the code seems to indicate
that you could make one simple modification to $ZENHOME/Products/
ZenRRD/zenperfsnmp.py to make this happen. Search for the following
line in that file and remove it.
del proxy.oidMap[oid]
Ticket #3899: make zenperfsnmp keep trying bad OIDs
http://dev.zenoss.org/trac/ticket/3899
_______________________________________________
zenoss-users mailing list
[email protected]
http://lists.zenoss.org/mailman/listinfo/zenoss-users
