We are experiencing intermittent freezing issues with Lens since the release of the new design. The application starts working fine but becomes progressively less responsive after a short period of use, eventually rendering the UI unusable. This happens even when Lens is idle, and it becomes increasingly difficult to close the application. Restarting temporarily resolves the issue, but the problem soon recurs.
We tested this on multiple machines running Windows 10 and 11.
Versions tested:
Stable: 2024.11.261604-latest
Beta: 2024.12.182210-beta (tested to see if the issue was resolved).
Behaviour:
Lens slows down noticeably within about 10 minutes of use.
Memory usage increases rapidly during this time, correlating with the performance degradation.
Ultimately, the UI stops rendering, requiring a restart.
Hi @pkeuter, thanks for reporting. We got similar reports recently and are investigating the root cause of the issue.
Out of curiosity, how many clusters do you have connected when this happens? and when you mentioned that Lens is “idle”, do you have tabs open at all ? or it even happens without checking any tabs?
Any chance you could share a recording? if you don’t want to share it in the forums, you can also send it here support@k8slens.dev, for us to take a look at it. It would be really helpful
Thank you for your message. I typically have one connection open at a time, and at most three simultaneously. However, this doesn’t seem to impact performance.
Currently, we aren’t experiencing performance issues when all tabs are closed. By “idle,” I meant that we’re not interacting with the application. That said, even with just one tab open and no content displayed, the freezing still occurs.
Initially, it works fine for about a minute, but then it starts to hang intermittently, first for a second, then for two seconds after a couple of minutes. Over time (around 20 minutes, though not precisely), it becomes completely unresponsive.
I am going to look into making a recording for you to review, but you will see that the entire interface just becomes unresponsive.
Thank you very much @pkeuter for providing this information, it definitely helps narrowing down a few things. If you can provide the recording it would be great. Meanwhile, we are solving a couple of bugs that we think could be part of the cause for this. We’ll share with you the version holding the fixes as soon as we can produce it (for validation of the fixes)
I’m also experiencing this, and this only happens when I connect to a particular cluster. other clusters work fine.
if I disconnect from this particular cluster in between freezes, the issue stops. If I reconnect to that cluster, the issue starts again. If I keep browsing that cluster, after few freezes, the entire UI hangs and I have to restart Lens.
During a freeze, a Lens process is using a lot of CPU:
in terms of pods and nodes it is actually much smaller than other clusters.
but:
it has the highest number of events, but it is not much higher than the others. its number of events is on average 1.5 times higher than the next cluster.
it has the highest number of CRDs (211), other clusters only have 144.
Thank you so much for taking the time to share your findings with us — we really appreciate it!
We’ve actually already been working on a fix for this issue. While it’s not directly related to the number of CRDs, we strongly believe it’s connected and addresses the underlying cause you’ve encountered.
