[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Need to kill pod



Hi,

We have a strange thing with heapster, after upgrading to 1.1.3:


4:27:07 PMNormalKilling Killing container with docker id 97b2aa607729: Need to kill pod.
4:23:47 PMNormalCreated Created container with docker id 97b2aa607729
4:23:47 PMNormalStarted Started container with docker id 97b2aa607729
4:23:46 PMNormalPulled Successfully pulled image "docker.io/openshift/origin-metrics-heapster:latest"
4:23:44 PMNormalScheduled Successfully assigned heapster-iwjir to node-1
4:23:44 PMNormalPulling pulling image "docker.io/openshift/origin-metrics-heapster:latest"
Heapster is scrapping metrics correctly, and every 3-5 minutes, the pod restarts for now reason. Apparently, kubernetes "needs" to kill it, but no idea why.
The logs of the pod / container don't show any clue:

[...]
I0227 21:26:35.000146       1 manager.go:79] Scraping metrics start: 2016-02-27 21:26:00 +0000 UTC, end: 2016-02-27 21:26:30 +0000 UTC
I0227 21:26:35.799862       1 manager.go:152] ScrapeMetrics: time: 799.529596ms size: 483
I0227 21:27:05.000138       1 manager.go:79] Scraping metrics start: 2016-02-27 21:26:30 +0000 UTC, end: 2016-02-27 21:27:00 +0000 UTC
I0227 21:27:05.786697       1 manager.go:152] ScrapeMetrics: time: 786.458797ms size: 483
<= exit here

the 2 others pods (hawkular and cassandra) don't restart.
We can see some graphs in the metrics tab, with some holes when heapster is restarting:

Inline image 1
Any idea what could go wrong?

Thanks

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]