@Pedro76 wrote:
Today, when I go to "Workloads" this message keeps appearing "Missing data from clusters blabla (deadline exceeded)".
I got this exact same issue. Below is just some details if this helps anyone else.
On project 1.
We were running a cluster with 2 nodepools, both on version 1.22.13-gke.1000.
I upgraded the cluster master to 1.23.12-gke.100. No issues.
I created a new node pool, and then moved over pods from one of the existing pools. Then deleted the empty old pool.
I started seeing this error, but nothing else seemed off. It also seemed to be intermittent.
I then created a new node pool once more to replace the remaining existing old one. The moment that nodepool was up I started seeing the error the whole time, but also no further effects to the cluster. Could still do day to day.
Once I removed the old nodepool however, not only did I keep getting the error, but about 4 out of 5 calls failed. Tried kubectl and GKE UI, kept getting the error and some of the pod listings etc also didn't respond. Not sure if the 2 are related.
I did also see the error on another cluster that I have not touched. This was on project 2 so completely unrelated to project 1. But seemed much less frequent on that side. Also all other api calls seemed to work on that cluster.
It seems to have gotten less frequent as the day went on. Api calls also slowly started responding as expected. This morning I am yet to see the error once.
While I had issues with the cluster itself, all pods running on the cluster was unaffected as far as I could tell. They all served traffic and responded in a timely manner.
I have upgraded many times, first time seeing this happen.
Additional info:
Project 1 is on europe-west2-b for both control plane and node zones
Project 2 is on europe-west1-b for the control plane and europe-west1-b, europe-west1-d, europe-west1-c for node zones.
Hope this info can help someone.
Update:
I do still get it now and again, but still fairly infrequent
LinkedIn
Twitter
