Possible memory on endpoint detail pages

[bakito]

Describe the bug

When I open an endpoint detail page endpoints/xxx in the browser, the memory consumption on the service increases on every refresh. After some refreshes the container crashes with OutOfMemory.

It seems that the memory increases on each request on the backend. Are there some cached data, that is not released?

What do you see?

The memory consumption increases on each refresh made on an endpoint detail page which leads to an OutOfMemory eventually.

What do you expect to see?

The memory consumption should not increase on each refresh.

List the steps that must be taken to reproduce this issue

1. Configure an endpoint
2. Open the detail page of this endpoint
3. Hit refresh several times

Version

v5.7.0

Additional information

We are running the docker image in kubernetes. Starting the Container hat 64Mb memory, we extended the memory up to 256Mb but the OOM occurs still after some refreshes.

The used memory during runtime without much browser access is around 50Mb

[TwiN]

I gave it a shot, and while I do see the memory increasing a bit more than I'd like to see, it eventually goes back down.

I do acknowledge that I am curious about why the memory increases so much, but there doesn't seem to be a memory leak.

[muhlba91]

same here. if i basically continue refreshing the status page all the time, i hit the same error. i hit OOM because kubernetes deployment kills the pod if it consumes more than the limit set. it's pretty consistent though. you see an increase with the number of refreshes, especially when checks are happening at the same time. memory consumption decreases over time (maybe some timeout or garbage collection?) but very slowly, in my opinion.

[221bshashank]

I am not sure if my case is same, but pod gets crashed after some time due to memory threshold reach maybe. This was not an issue earlier.