ui: timeseries dashboard pages are really slow on large clusters #30287
Labels
A-webui-general
Issues on the DB Console that span multiple areas or don't have another clear category.
A-webui-scaling
C-performance
Perf of queries or internals. Solution not expected to change functional behavior.
S-3-ux-surprise
Issue leaves users wondering whether CRDB is behaving properly. Likely to hurt reputation/adoption.
I was running a 120 node cluster last week as part of TPC-C 50k testing, and the UI was really, really bogging down my browser whenever I had pages with graphs open. I believe it was a client side issue (although I'm sure that loading all the timeseries in the database isn't easy, either) given the way Chrome occasionally became unresponsive.
It didn't help that sometimes the time span selector would automatically switch from 10 minutes to 6 hours when I would reload or switch dashboards without me telling it to. I don't know what was going on there.
I realize this isn't a particularly useful issue since I didn't record any data about the slowness or track it down to anything specific, but I'm planning to spin up another such cluster again for a day this week. Is there anything you'd find useful for me to do/record then? Would you like to poke at it when it's up and running?
@couchand
The text was updated successfully, but these errors were encountered: