11

Since GKE upgraded to 1.21.10-gke.2000 I see these errors in the konnectivity-agent logs every 5 minutes or so. I cannot detect any consequences to these problems, but I cannot get them to go away either. Does anyone know what the problem is and how I get it to go away?

pod_name: "konnectivity-agent-779eadf84d-98kjh"

"conn write failure" err="write tcp 10.48.0.2:40528->10.48.0.8:10250: use of closed network connection" connectionID=215
"connection read failure" err="read tcp 10.48.0.2:40528->10.48.0.8:10250: use of closed network connection"
Andy G
  • 321
  • 2
  • 9

2 Answers2

0

I have seen this error when the liveness probe is timing out. Try lengthening the timeoutSeconds on your livenessProbe and see if the problem goes away.

Refer this link for more information.

Ramesh kollisetty
  • 245
  • 1
  • 7
-1

I tried the below: Adding a firewall egress rule that allow you to connect to the port (use the port number getting displayed in the error message with the IP of the end point) , you could use the following command to add that rule. that should allow the konnectivity-agent to connect to the control plane.

gcloud compute firewall-rules create gke-node-to-konnectivity-service \
--allow=tcp:<port number> \
--direction=EGRESS \
--destination-ranges=<endpoint IP address > \
--target-tags=< node name> \
--priority=1000

Am using gke version 1.21.11. But did not work though

Sunil
  • 553
  • 1
  • 12
  • 30
  • Hi, how can you handle this in case of auto scaling where you have a lot of nodes? Also, what is the endpoint IP address you mentioned? Is it the IP of the master? – nvcnvn Jun 01 '22 at 15:38
  • Hi, the above is not working for me as expected – Sunil Jun 09 '22 at 04:34