partitioning logic runs too often
Bug #1554738 reported by
gordon chung
This bug affects 1 person
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Gnocchi |
Fix Released
|
Low
|
Trevor McCasland |
Bug Description
gnocchi uses the reporting worker to partition tasks across each of the processing workers. by default reporter runs every 10s and processor worker every 5s. unfortunately, in any environment, small or large, the processor takes much longer than 10s to process it's data. because of that, the repartition messages that the reporter sends up queues up so the processor ends up running through a list of stale repartition messages before it can grab a new set of metrics.
we should either set reporter to report less frequently OR get it broadcast less frequently. i'm inclined to put a min period of 1min that the reporter can run at.
Changed in gnocchi: | |
status: | New → Triaged |
importance: | Undecided → Low |
Changed in gnocchi: | |
assignee: | nobody → Trevor McCasland (twm2016) |
Changed in gnocchi: | |
milestone: | none → 2.1.0 |
status: | Fix Committed → Fix Released |
To post a comment you must log in.
Fix proposed to branch: master /review. openstack. org/290542
Review: https:/