Recent AppArmor issue on Linux kernel 7.0.0-43
Nov 17, 2026
I was doing a routine maintenance and reboot a node as usual, but this time it's been 10 minutes and some workloads are not up. Turns out all cloudnative-pg containers are unhealthy, it's used up all connections, and I can't use psql to clear the connections.
Turns out it's a known AppArmor issue on kernel 7.0.0-43. The fix is to disable AppArmor until a proper fix is introduced. For cloudnative-pg, adding metadata per this issue can fix unhealthy cloudnative-pg cluster issues, but you'll have to force remove the database pods so the metadata is applied.
It also affects authentik as well, there is no official fix, but you can patch an authentik pod to disable AppArmor:
kubectl patch deployment authentik-server --type=strategic \
-p '{"spec":{"template":{"spec":{"containers":[{"name":"server","securityContext":{"appArmorProfile":{"type":"Unconfined"}}}]}}}}'And KubeVirt is also affected too, but you can apply the same patch via:
kubectl patch vm $vmName --type=merge -p '{"spec":{"template":{"metadata":{"annotations":{"container.apparmor.security.beta.kubernetes.io/compute":"unconfined"}}}}}'