centreon-plugins
centreon-plugins copied to clipboard
Performance statistics Dell Powerstore Check
Hi,
Is it possible to add performance statistics to the Dell Powerstore check? Kind of the same as on the Dell Powervault checks. The most important part for us are latency statistics.
Hi,
What kind of performance data? I assume you would expect it in hardware components.
If the API provides some data why not, but we will need curl and sample results to develop it.
For the Dell Powervault there is a controller statistics component and a volume statistics component. I would like to see the same for the Powerstore check, so not only performance data, but also checks on these data if everything is OK. So for example when there is too much latency we get an alert.
The API can provide all these data, when kind of input can I provide? Can I provide this private?
@Hbitvof could you give me the units for the following values ? (latency an bandwitdth)
"avg_write_latency": 381.44846,
"avg_read_bandwidth": 55678444,
@Hbitvof could you give me the units for the following values ? (latency an bandwitdth)
"avg_write_latency": 381.44846, "avg_read_bandwidth": 55678444,
Following the documentation that's: latency in microseconds bytes per second
Thanks. It will be available thursday
Could you test with that version ? https://github.com/centreon/centreon-plugins/archive/refs/heads/MON-14466-powerstore-performance.zip
Could you test with that version ? https://github.com/centreon/centreon-plugins/archive/refs/heads/MON-14466-powerstore-performance.zip
Yes I can. What part is changed so I know how what commands I need to run?
@Hbitvof with sources above, try the new --plugin=storage::dell::powerstore::restapi::plugin --mode=clusters that should answer to your request.
@Hbitvof with sources above, try the new --plugin=storage::dell::powerstore::restapi::plugin --mode=clusters that should answer to your request.
I get the following output when running the new part:
could you provide the --debug result ?
could you provide the --debug result ?
It's weird. Other modes are working ? The curl command is working with endpoint: /api/rest/cluster
?
It's weird. Other modes are working ? The curl command is working with endpoint:
/api/rest/cluster
?
Apologies, I did something wrong. I tried it again with the right commandline and I get:
UNKNOWN: 405 Method Not Allowed
Could you provide the --debug ?
Could you provide the --debug ?
I think I know what the problem is. Method should be POST instead of GET for /api/rest/metrics/generate
Could you redownload it and test it ? (i have done a patch)
Could you redownload it and test it ? (i have done a patch)
Can you provide a download link again? I just downloaded it again from https://github.com/centreon/centreon-plugins/archive/refs/heads/MON-14466-powerstore-performance.zip but got the same error.
It's the good link:
https://github.com/centreon/centreon-plugins/archive/refs/heads/MON-14466-powerstore-performance.zip
Could you provide the --debug ? (it should be a POST).
I get another error now:
It's better :) You can redownload it and retest it ? (i have a done a new patch)
The check is working now!
I only see some strange output in the performance data:
How can I set warning/critical thresholds for this check?
I have removed the microseconds units. Thresholds are in the help:
=item B<--warning-*> B<--critical-*>
Thresholds.
Can be: 'clusters-detected',
'read-iops-5m', 'read-iops-30m', 'read-iops-1h', 'read-iops-24h',
'write-iops-5m', 'write-iops-30m', 'write-iops-1h', 'write-iops-24h',
'read-latency-5m', 'read-latency-30m', 'read-latency-1h', 'read-latency-24h',
'write-latency-5m', 'write-latency-30m', 'write-latency-1h', 'write-latency-24h',
'read-bandwidth-5m', 'read-bandwidth-30m', 'read-bandwidth-1h', 'read-bandwidth-24h',
'write-bandwidth-5m', 'write-bandwidth-30m', 'write-bandwidth-1h', 'write-bandwidth-24h'.
I have removed the microseconds units. Thresholds are in the help:
=item B<--warning-*> B<--critical-*> Thresholds. Can be: 'clusters-detected', 'read-iops-5m', 'read-iops-30m', 'read-iops-1h', 'read-iops-24h', 'write-iops-5m', 'write-iops-30m', 'write-iops-1h', 'write-iops-24h', 'read-latency-5m', 'read-latency-30m', 'read-latency-1h', 'read-latency-24h', 'write-latency-5m', 'write-latency-30m', 'write-latency-1h', 'write-latency-24h', 'read-bandwidth-5m', 'read-bandwidth-30m', 'read-bandwidth-1h', 'read-bandwidth-24h', 'write-bandwidth-5m', 'write-bandwidth-30m', 'write-bandwidth-1h', 'write-bandwidth-24h'.
Thanks for your quick actions!
Is it possible to convert the microsecond units to miliseconds? That's more common in storage performance counters.
Yes i will update it
It's merged!
It's merged!
I think you forgot to convert the output (/1000) :).
It should be done: https://github.com/centreon/centreon-plugins/pull/3838/files#diff-29f6c085963d5cf9ad2e222b9f080ad492476c4896ff21e13175e29cc8e9b351R227
It should be done: https://github.com/centreon/centreon-plugins/pull/3838/files#diff-29f6c085963d5cf9ad2e222b9f080ad492476c4896ff21e13175e29cc8e9b351R227
Well, it doesn't seem to work (we don't have this latency on our storage :)):