When running cAdvisor on a Podman-based system, cAdvisor periodically queries the Podman API to collect container metadata. During this process, cAdvisor appears to treat Pod IDs as Container IDs, resulting in Podman returning:
level=info msg="Request Failed(Not Found): no container with name or ID \"3f52b41ef23a0c29544eaccafa0e482f3c811430de35eb9c8d39d1ec8bb59623\" found: no such container"
These messages are logged every 60 seconds by the Podman systemd service and create significant log noise in /var/log/messages or journalctl.
We are creating pods using podman kube play command.
There is currently no way to filter out infra containers or Pod IDs from cAdvisor's runtime API calls.
# podman ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
d64e93249f5e appliance/autosupport:7.0.0 /sbin/init 8 hours ago Up 8 hours (healthy) autosupport
2ecc072b1795 appliance/etcd:3.6.4 /usr/local/bin/et... 7 hours ago Up 7 hours etcd
1cb4a03e87c9 appliance/svc/auth:7.0.0 7 hours ago Up 7 hours (healthy) authservice
053658f3ab48 appliance/registry:3.0.0 /etc/distribution... 7 hours ago Up 7 hours registry
05568e479d4c /svc/mgmt:7.0.0 7 hours ago Up 7 hours (healthy) mgmtserver
6cc225b5a627 svc/gateway:7.0.0 /usr/sbin/nginx -... 7 hours ago Up 7 hours (healthy) 10.81.202.68:443->443/tcp, 10.81.202.68:1443->1443/tcp webui
4c3a323d8800 lsvc/settings:7.0.0 7 hours ago Up 7 hours (healthy) settings
b29d328bf38a localhost/podman-pause:4.9.4-rhel-1762969518 7 hours ago Up 7 hours 3f52b41ef23a-infra
e2b60393a755 localhost/podman-pause:4.9.4-rhel-1762969518 7 hours ago Up 7 hours 4f4aa37568d2-infra
3389b212793a /svc/peer:7.0.0 7 hours ago Up 7 hours (healthy) peer-pod-peer
77af637b2ba7 /appliance/netinsights:6.2.0 7 hours ago Up 7 hours (healthy) remotemgmt-nic
a1d5213e9ffc flex/svc/gateway:7.0.0 /usr/sbin/nginx -... 7 hours ago Up 7 hours (healthy) hostapi
5df8dcc5a80f localhost/podman-pause:4.9.4-rhel-1762969518 7 hours ago Up 7 hours 65e843376519-infra
d04ff306a2e5 /appliance/rabbitmq:3.12.4-12 7 hours ago Up 7 hours (healthy) pubsub-rabbitmq
ee57e9f04417 /svc/pubsub:7.0.0 7 hours ago Up 7 hours (healthy) pubsub-server
8bd3c2f1b879 /appliance/prometheus:3.5.0 --config.file=/et... 7 hours ago Up 7 hours prometheus
# podman pod list
POD ID NAME STATUS CREATED INFRA ID # OF CONTAINERS
65e843376519 pubsub Running 7 hours ago 5df8dcc5a80f 3
4f4aa37568d2 remotemgmt Running 7 hours ago e2b60393a755 2
3f52b41ef23a peer-pod Running 7 hours ago b29d328bf38a 2
Pods and there corresponding infra containers
POD Infra Container
65e843376519 65e843376519-infra
4f4aa37568d2. 4f4aa37568d2-infra
3f52b41ef23a 3f52b41ef23a-infra
Environment
cAdvisor version v0.49.1
Podman version 4.9.4-rhel
OS: RHEL
Container runtime API: Podman Docker-compat API (/run/podman/podman.sock)
cAdvisor service ExecStart:
cadvisor \
-docker=unix:// \
-housekeeping_interval=30s \
-raw_cgroup_prefix_whitelist="/machine.slice/libpod" \
-store_container_labels=false \
-disable_metrics=advtcp,app,cpuLoad,cpu_topology,cpuset,disk,diskIO,hugetlb,memory_numa,oom_event,percpu,perf_event,process,referenced_memory,resctrl,sched,tcp,udp \
-enable_metrics=cpu,memory,network
When running cAdvisor on a Podman-based system, cAdvisor periodically queries the Podman API to collect container metadata. During this process, cAdvisor appears to treat Pod IDs as Container IDs, resulting in Podman returning:
level=info msg="Request Failed(Not Found): no container with name or ID \"3f52b41ef23a0c29544eaccafa0e482f3c811430de35eb9c8d39d1ec8bb59623\" found: no such container"These messages are logged every 60 seconds by the Podman systemd service and create significant log noise in /var/log/messages or journalctl.
We are creating pods using
podman kube playcommand.There is currently no way to filter out infra containers or Pod IDs from cAdvisor's runtime API calls.
Pods and there corresponding infra containers
Environment
cAdvisor version v0.49.1
Podman version 4.9.4-rhel
OS: RHEL
Container runtime API: Podman Docker-compat API (/run/podman/podman.sock)
cAdvisor service ExecStart: