After upgrading to 4.2.3 on Centos 6 x86_64 the zenmodeler daemon seems to be misbehaving. After a couple of hours the daemon will begin to comsume large CPU and memory resources on the server. At the same time Zenoss will report that the heartbeat to the zenoss daemon has been lost. The log file for zenmodeler does not show any weirdness when this is happening. The only thing that I see when the zenmodeler daemon begins to act up is the following entry in the zenstatus.log file:
2013-05-04 16:12:51,153 INFO zen.zenstatus: 44 devices processed (0 datapoints)
2013-05-04 16:12:51,156 INFO zen.collector.scheduler: Tasks: 145 Successful_Runs: 9721 Failed_Runs: 0 Missed_Runs: 0 Queued_Tasks: 0 Running_Tasks: 0
Below is the zenmodeler config:
force True
duallog False
hubhost localhost
uid zenoss
collage 0
zenhubpinginterval 30
enableRegex assword:
nowmi False
checkstatus False
promptTimeout 10
watchdog False
eventflushseconds 5.0
allowduplicateclears False
ntlmv2auth False
commandTimeout 10
monitor localhost
hubusername admin
user zenoss
maxqueuelen 5000
## Commenting out by config parser on 2013-05-03T21:55:40.151884: unknown option
## zodb-portscantimeout 5
concurrentSessions 10
hubpassword zenoss
loginTimeout 10
existenceTest test -f %s
enable False
cycletime 720
duplicateclearinterval 0
initialHubTimeout 30
logseverity 20
maxlogsize 10240
passwordRegex assword:
loginRegex ogin:.$
now False
parallel 1
loginTries 1
maxbackuplogs 3
writetries 2
communities False
debug False
termlen False
keyPath ~/.ssh/id_dsa
eventflushchunksize 50
hubport 8789
Any ideas?