Cpu meters stop after stopping/resuming an instance
Bug #1154596 reported by
Yolanda Robla
This bug affects 1 person
| Affects | Status | Importance | Assigned to | Milestone | |
|---|---|---|---|---|---|
| Ceilometer |
Invalid
|
Undecided
|
Eoghan Glynn | ||
Bug Description
I'll show the case:
I have an instance with resource id xxx
I do a nova stop and a nova start for a while. After that, when i do a call with ceilometerclient, i'm stopping having meterings from cpu. I receive one/several blank meterings, and then it stops collecting cpu:
| xxx | cpu | cumulative | 1.0177841e+14 | ns | 2013-03-13T13:04:19 |
| xxx | cpu | cumulative | 1.0188838e+14 | ns | 2013-03-13T13:14:19 |
| xxx | cpu | cumulative | | ns | 2013-03-13T13:24:19 |
Other machines have been reporting cpu usage after that date. Also i can see disk usage for that machines after that date, but the cpu meterings are missing.
| Changed in ceilometer: | |
| assignee: | nobody → Eoghan Glynn (eglynn) |
Revision history for this message
| Eoghan Glynn (eglynn) wrote | #1 |
| Changed in ceilometer: | |
| status: | New → Invalid |
Revision history for this message
| Eoghan Glynn (eglynn) wrote | #2 |
To post a comment you must log in.
This is a case of the later samples not being noticed in the ceilometerclient output, as these are not sorted in timestamp order.
The later samples can be seen post-restart with something like:
ceilometer sample-list -c cpu -r $INSTANCE_ID | awk '/2013/ {print $(NF-1)}' | sort