-
Bug
-
Resolution: Unresolved
-
Normal
-
None
-
4.19.z
-
Quality / Stability / Reliability
-
False
-
-
None
-
None
-
None
-
None
-
None
-
Rejected
-
None
-
None
-
None
-
None
-
None
-
None
-
None
-
None
Description of problem:
When the scale tests create 72 RouteAdvertisements with qps=20 and burst=20 with 1:1 RA:CUDN, ovnkue-controller CPU uage is going above 800%. This is observed in all nodes (its a 18 worker node env)
I have attached the pprof, audit logs and must-gather which captured during the testing.
must-gather https://storage.scalelab.redhat.com/anilvenkata/bgp/rahighcpu/must-gather.local.581736730274221847.tar.gz
RA configuration https://storage.scalelab.redhat.com/anilvenkata/bgp/rahighcpu/log_routeadvertisement_all.yaml
CPU pprof https://storage.scalelab.redhat.com/anilvenkata/bgp/rahighcpu/pprof/ovnkube-cpu1.pprof
Memoy pprof https://storage.scalelab.redhat.com/anilvenkata/bgp/rahighcpu/pprof/ovnkube-heap1.pprof
This is a follow up of the bug https://issues.redhat.com/browse/OCPBUGS-54375
Dashboard showing 800% cpu usage by ovnkube-controller