After the gather refactor logic - the gather loop is done and managed by measurements instead of devices - the load index process (every device FilterFreq or measurement override) tries go refresh the current indexes
The following log entry appeared on some devices:
mydevice1.log:time="2022-04-06 10:47:01" level=error msg="Error while trying to reload Indexed Labels for baseOid .1.3.6.1.4.1.9.9.166.1.1.1.1.4 : ERROR: set udp XXX.XXX.XXX.XXX:41931: use of closed network connection" device=YY.YY.YY.YY measurement=cisco_qos_cbQosCMStats
mydevice1.log:time="2022-04-06 10:47:01" level=error msg="LOADINDEXEDLABELS - SNMP WALK error: set udp XXX.XXX.XXX.XXX:41931: use of closed network connection" device=YY.YY.YY.YY measurement=cisco_qos_cbQosCMStats
All the following entries refers to the same origin. After a deep analysis to find the root cause, it seems related with:
MultiIndex Measurements
Device totally unavailable (System OIDs) during some intervals prior to the first error message
After the gather refactor logic - the gather loop is done and managed by measurements instead of devices - the load index process (every device FilterFreq or measurement override) tries go refresh the current indexes
The following log entry appeared on some devices:
All the following entries refers to the same origin. After a deep analysis to find the root cause, it seems related with: