-
Notifications
You must be signed in to change notification settings - Fork 3.9k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[WIP] ui: show IO metrics in node list #27660
[WIP] ui: show IO metrics in node list #27660
Conversation
Use new library `gopsutil` to gather IO stats from the host kernel and save them as time series. Caveat: these counters are for the entire host over its lifetime. As such, the first value cockroach records may be very high, and if we plot these charts as derivatives, the derivative from zero to the first recorded data point will be much higher than subsequent derivatives, making a plot where it's hard to see changes after that. A future commit will gather per-process IO stats, but gopsutil only provides that on Linux. Release note (admin ui change): gather disk and network IO statistics and save them as time series.
This commit updates the `vendor` submodule to point at github.com/cockroachdb/vendored master, including this new library for hardware stats. Release note: None
Release note: None
Release note: None
Release note: None
Release note: None
Release note: None
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
It's not clear from the screenshot whether those are (a) cumulative for the life of the node or (b) the units are just missing a /s
(e.g. MiB/s
instead of just MiB
). It's also not clear whether they're (c) instantaneous measurements or (d) average/median/nth percentile over some time period.
And please ignore this last comment since I'm sure it's already been thought through, but to me these seem less valuable than the other info on this already densely packed overview page.
Thanks for the feedback Alex. They'll definitely need a lot more labelling before this merges. (b) and (c) are correct — they're derivatives of IO counters (i.e. MiB/s); i.e. the most recent values you would see in the charts in the new Hardware dashboard (#27626). Also, it's definitely still up for debate what metrics (if any) are worth squeezing into this table. Just sort of put everything here to start so we can winnow it down. Will discuss with @piyush-singh, @couchand and @josueeee tomorrow. |
Just going to rely on the timeseries screens to look at these IO metrics. |
server
andbuild
commits are part of #27137, which will go in first. (the firstui
commit is between them in the GitHub UI, but not in the git rebase interface…)("Net IO" and "Disk IO" columns are new)