Description of problem: This bug is opened to backport fix : https://github.com/cri-o/cri-o/pull/5508 Version-Release number of selected component (if applicable): OCP 4.11 Actual results: Percpu memory usage is high: $ cat proc/meminfo | awk '{print $2 " " $1}'| sort -rn | awk '{print $1 " = " int($1/1024) "MB - "int($1/1024/1024)"GB " $2}' | grep -E 'MemTotal|MemFree|Buffers|Cached|Percpu' 32897520 = 32126MB - 31GB MemTotal: 19215360 = 18765MB - 18GB Percpu: 4547680 = 4441MB - 4GB Cached: 449000 = 438MB - 0GB MemFree: 3108 = 3MB - 0GB Buffers: 0 = 0MB - 0GB SwapCached: Expected results: Not high usage of percpu memory. Additional info: child of bug https://bugzilla.redhat.com/show_bug.cgi?id=2004037#c121
% oc get clusterversion NAME VERSION AVAILABLE PROGRESSING SINCE STATUS version 4.11.0-0.nightly-2022-08-31-193418 True False 5h31m Cluster version is 4.11.0-0.nightly-2022-08-31-193418 sh-4.4# while :; do podman run --name=test1 --replace centos /bin/echo 'running'; done ae97e06331bb7ec9ed7c5cd9e018003f0a45b7cea189a634d6bced7e52051b31 running …. sh-4.4# grep Per /proc/meminfo Percpu: 6000 kB sh-4.4# cat /proc/cgroups | column -t #subsys_name hierarchy num_cgroups enabled cpuset 9 68 1 cpu 2 286 1 cpuacct 2 286 1 blkio 8 286 1 memory 5 368 1 devices 3 282 1 freezer 6 68 1 net_cls 4 68 1 perf_event 12 68 1 net_prio 4 68 1 hugetlb 7 68 1 pids 11 286 1 rdma 10 68 1 sh-4.4# date Thu Sep 1 12:45:46 UTC 2022 sh-4.4# grep Per /proc/meminfo Percpu: 6048 kB sh-4.4# cat /proc/cgroups | column -t #subsys_name hierarchy num_cgroups enabled cpuset 9 65 1 cpu 2 282 1 cpuacct 2 282 1 blkio 8 282 1 memory 5 370 1 devices 3 278 1 freezer 6 65 1 net_cls 4 65 1 perf_event 12 65 1 net_prio 4 65 1 hugetlb 7 65 1 pids 11 282 1 rdma 10 65 1
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 (Moderate: OpenShift Container Platform 4.11.3 packages 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-2022:6287