Bug 2002868

Since in previous versions we had OVS metrics, we should consider this a regression.

Note that this is not specific to ovnkube, but applies to SDN too.

Part of the ovn-kubernetes solution is https://github.com/ovn-org/ovn-kubernetes/pull/2723 . Still need to work out with upstream whether we enable the metrics as part of ovnkube-node (which we already expose with kbue-rbac-proxy) or whether it's done as part of a another container with the standalone upstream metrics executable.

It could be a different daemonset that both SDN and OVN can use, but that's a lot more work/book-keeping (different image too) and the code isn't huge, so can be duplicated for SDN as well.

This was enabled for ovn-kubernetes by https://github.com/openshift/cluster-network-operator/pull/1393 and the core functionality was present in ovnkube since the beginning of 4.11.

To be clear, this has been enabled & available in 4.11 ovnkube/CNO since late April 2022. Just missed the tie between PR and bug.

Tested and verified in 4.11.0-rc.4

sh-4.4# curl 127.0.0.1:29105/metrics | grep -E 'process_virtual_memory_bytes|process_cpu_seconds_total|process_virtual_memory_max_bytes|process_resident_memory_bytes'
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100 3# HELP ovs_db_process_cpu_seconds_total Total user and system CPU time spent in seconds.
13# TYPE ovs_db_process_cpu_seconds_total counter
80ovs_db_process_cpu_seconds_total 1.63
    0 31380    0     0   74# HELP ovs_db_process_resident_memory_bytes Resident memory size in bytes.
7k# TYPE ovs_db_process_resident_memory_bytes gauge
 ovs_db_process_resident_memory_bytes 3.2751616e+07
     0 --:--:-- --:--:-- --:--:--  766k
# HELP ovs_db_process_virtual_memory_bytes Virtual memory size in bytes.
# TYPE ovs_db_process_virtual_memory_bytes gauge
ovs_db_process_virtual_memory_bytes 9.6464896e+07
# HELP ovs_db_process_virtual_memory_max_bytes Maximum amount of virtual memory available in bytes.
# TYPE ovs_db_process_virtual_memory_max_bytes gauge
ovs_db_process_virtual_memory_max_bytes 1.8446744073709552e+19
# HELP ovs_vswitchd_process_cpu_seconds_total Total user and system CPU time spent in seconds.
# TYPE ovs_vswitchd_process_cpu_seconds_total counter
ovs_vswitchd_process_cpu_seconds_total 25.24
# HELP ovs_vswitchd_process_resident_memory_bytes Resident memory size in bytes.
# TYPE ovs_vswitchd_process_resident_memory_bytes gauge
ovs_vswitchd_process_resident_memory_bytes 1.95637248e+08
# HELP ovs_vswitchd_process_virtual_memory_bytes Virtual memory size in bytes.
# TYPE ovs_vswitchd_process_virtual_memory_bytes gauge
ovs_vswitchd_process_virtual_memory_bytes 7.06281472e+08
# HELP ovs_vswitchd_process_virtual_memory_max_bytes Maximum amount of virtual memory available in bytes.
# TYPE ovs_vswitchd_process_virtual_memory_max_bytes gauge
ovs_vswitchd_process_virtual_memory_max_bytes 1.8446744073709552e+19
sh-4.4# exit
exit
[weliang@weliang ~]$ oc get clusterversion
NAME      VERSION       AVAILABLE   PROGRESSING   SINCE   STATUS
version   4.11.0-rc.4   True        False         24m     Cluster version is 4.11.0-rc.4
[weliang@weliang ~]$

Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory (Important: OpenShift Container Platform 4.11.0 bug fix and security update), and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

https://access.redhat.com/errata/RHSA-2022:5069