I only have a single retention configured in my storage-schemas.conf
[default_rentions]
pattern = .*
#retentions = 60s:7d,120s:31d,600s:365d,30m:9y
retentions = 15s:7d,5m:30d,15m:10y
All other collectd data is saved as expected, however any custom metric gathered via the Exec plugin is only saved for just a week.
collectd.conf config:
LoadPlugin exec
<Plugin exec>
Exec "centos:centos" "/etc/collectd/site-benchmarks.pl"
</Plugin>
I've already tried forcefully manually resizing the existing whisper files using the retention times that I want it to be, but that didn't seem to fix the issue.
find ./ -type f -name '*.wsp' -exec whisper-resize.py --nobackup {} 15s:7d 5m:30d 15m:10y \;
I've also already tried removing the corresponding *.wsp files so that graphite builds them scratch, and that also didn't help. All new custom exec collected data is still only being saved for a week.
Anyone have any ideas on why only custom collectd exec plugin's data is saved with a retention of a week?
Update: I've even verified the updated retention configs have been applied to the whisper files. Example test wsp file whisper metadata output:
[centos@ip-172-16-16-124 apache-response-time]$ whisper-dump.py gauge-test.wsp
Meta data:
aggregation method: average
max retention: 315360000
xFilesFactor: 0.5
Archive 0 info:
offset: 52
seconds per point: 15
points: 40320
retention: 604800
size: 483840
Archive 1 info:
offset: 483892
seconds per point: 300
points: 8640
retention: 2592000
size: 103680
Archive 2 info:
offset: 587572
seconds per point: 900
points: 350400
retention: 315360000
size: 4204800
How long does the executed script take to run?
It might be a possibility that points are committed slower than once every 15 seconds which results in missing points. If there are not enough points carbon will refuse to aggregate the data. If that is the case you could lower the xFilesFactor which describes how many datapoints there must be present to aggregate.