Page MenuHomePhabricator

gvisor pods are crashing (CrashLoopBackOff)
Closed, ResolvedPublic

Description

Events:
  Type     Reason          Age                   From     Message
  ----     ------          ----                  ----     -------
  Warning  BackOff         20m (x5423 over 21h)  kubelet  Back-off restarting failed container
  Normal   SandboxChanged  16m                   kubelet  Pod sandbox changed, it will be killed and re-created.
  Normal   Pulled          13m (x4 over 16m)     kubelet  Container image "gcr.io/k8s-minikube/gvisor-addon:3@sha256:23eb17d48a66fc2b09c31454fb54ecae520c3e9c9197ef17fcb398b4f31d505a" already present on machine
  Normal   Created         13m (x4 over 16m)     kubelet  Created container gvisor
  Normal   Started         13m (x4 over 16m)     kubelet  Started container gvisor
  Warning  BackOff         69s (x56 over 16m)    kubelet  Back-off restarting failed container

Logs:

2021/03/25 08:14:18 downloading binaries: downloading runsc: Get https://storage.googleapis.com/gvisor/releases/nightly/2020-02-14/runsc: dial tcp 142.250.185.240:443: i/o timeout

Event Timeline

s669m triaged this task as Normal priority.Thu, Mar 25, 9:21 AM
s669m created this task.

Discussion from Tech meeting (Hanno, Klaus, Marco)

  • Microk8s has support for Kata, best option is to contact ITCF and try Kata
s669m lowered the priority of this task from Normal to Low.Thu, Apr 1, 10:15 AM
s669m moved this task from In Progress to Resolved on the Kaapana (internal) board.
schererj claimed this task.