7 days monitoring with Kubernetes plugin times out



I’m using Netdata Cloud Beta architecture with agents v1.32.0 on parent-child architecture to monitor a Kubernetes cluster.

I have some trouble trying to show timespans superior to 4-5 days on the merged Kubernetes graphs only. (it doesn’t happen on a single node view or with the nodes graphs). The query times out, sometimes end up showing something but with 2 nodes erroring (I have 3 nodes on the k8s cluster).
Also after the query, the whole graphs are erroring, even if I set the timespan back to 15 or 5min, and end up showing again after maybe 2 minutes.

The parent and the childs are not showing any error in the logs, I couldn’t find anything to enable “debug/verbose logs” on containers. I would like some help to debug this, if possible :slight_smile:.


Last 15min K8S plugin

Last 4 days K8S plugin

After some secs - showing ERROR but still showing the graphs

Last 7 days K8S plugin

After some secs

After a minut or something

Last 7 days standard cpu graph

After some secs


I’m running Netdata with parent-child architecture on 3 kubernetes nodes. Each node has 4 cores and 16GB or RAM, whenever I try to show 7 days on the kubernetes plugin I can see the resources usage growing (as expected), but I do not expect it to crash with this usage. the netdata parent has no resource limit.
1769m means 1769 milliCPU which represent 1.7 core usage.

NAME   CPU(cores)   CPU%   MEMORY(bytes)   MEMORY%
node    2404m        61%    9087Mi          62%
NAME                              CPU(cores)   MEMORY(bytes)
netdata-child-467pp               63m          143Mi
netdata-child-88vq8               52m          162Mi
netdata-child-q6w9q               71m          157Mi
netdata-parent-77d87f5979-7857g   1769m        2381Mi

Tried with Firefox 94.0 and Chromium 96.0.4664.45

Also I’m on your discord channel with the same username if you want to reach me there.


Case handled by papazach directly on Discord.
Will keep this thread updated with the info I have.