Closed niki-on-github closed 2 years ago
I second this - somehow some of my pods started doing this since yesterday. Haven't been able to figure out the root cause just yet.
Edit: I'm gonna have to add in that after testing your PR, my containers still crash, but it's very inconsistent.
For example: I have 5 services running in my vpn
namespace. Two of which keep on crashing - the other 3 are fine.
The crashing ones stay in CrashLoopBackoff
untill I manually delete the pods and then they'll be fine for approx 8-ish hours.
My logs are as of yet inconclusive to what exactly causes this, but my sidecar container is throwing the exact same error as @niki-on-github describes above.
Description of the change:
To use the reconnect functionality I had to make a small adjustment otherwise the Sidecar stays in
CrashLoopBackOff
and restarts constantly. See log below:And as a result, a connection to the gateway can no longer be established.
Description: Because the pod already has the route setting the default route does not longer exist and the script is not able to determine the variable
K8S_GW_IP
. But we don't need this in this case anyway because the corresponding routing setting is already set.Possible drawbacks:
None