This is an anecdotal observation that our instance gets very slow after extensive use of the interactive 3D volume viewer. The project in which the volume viewer was activated seems affected the most. The way to fix this is to re-booting the instance and usually, cryosparcm stop does not cut it so I have to send a stop signal to the supervisor process (using the kill command without -9).
Did anyone else experience this? What log files would should I look into to troubleshoot the issue?
Not sure if this is the same issue or not, but we have had a similar issue for as long as we’ve been using CryoSPARC (v4.6 currently) on our cluster. Every couple weeks we have to restart everything, because things get extremely slow or nonresponsive, primarily on the CryoSPARC Live interface (can’t see exposures, can’t see 2Ds or any graphics when things get really bad) and also the other place we see it is the volume viewer. We’ve never tracked if it tracks with use of the volume viewer and it seems to be across all projects / users for us.
We’ve never been able to track down where the issue comes from but it is annoying since we have to go into maintenance mode and/or coordinate with all our users to not start jobs until everything can be reset. Often just resetting the web interface part fixes it although sometimes we have to restart the instance. Of note, it is not an issue with jobs actually running. Just an issue with visualizing them. Of note, we’ve tried accessing the interface in numerous ways, including local portforwarding in combination with a URL, and that has not had any impact on the issue.
Hi Daniel,
It may be unrelated, but we had been experiencing a similar-sounding issue with command_vis that was apparently fixed with the v4.7 update.
I suppose if you’re still experiencing this having applied the update then it may indeed be unrelated–or the fix was inefficacious.
Cheers,
Yang
We recently upgraded to 4.7. I hope this fixed the problem. I will report back after we have used it for some time.