-
Notifications
You must be signed in to change notification settings - Fork 962
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
perf: imporve end to end queue processing latency #3772
Conversation
Welcome @fengruotj! |
hi @kevin-wangzefeng, this is the perf: imporve end to end queue processing latency |
9160b72
to
39e69bb
Compare
Hi, please sign off your commit with |
/ok-to-test |
Signed-off-by: tanjie.master <tanjiemaster@gmail.com>
39e69bb
to
9b94a73
Compare
Hi, I've already signed off my commit wiht |
/assign @Monokaix @kevin-wangzefeng |
/lgtm |
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.
/lgtm
@hwdef I've considered this as well. The main issue is that the state transition of the queue involves queuecommand, and commandWorker is primarily used to handle the queuecommand CRD, which is then passed to the worker for processing by workqueue. The main performance bottleneck lies in both the worker and the commandWorker. I believe concurrency should be applied to both. |
@fengruotj |
/approve |
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: Monokaix The full list of commands accepted by this bot can be found here. The pull request process is described here
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
The current implementation of queue controller handles queue state operations and queuecommand operations for sequentially which may come into a performance bottleneck in performance-critical clusters. For example the end user may increase queuecommand & queue creation speed, so this pr add an option to make queue & queuecommand creation parallelly to handle such case.
Here, queue-controller workqueue pending process object metrics.
After perf: imporve end to end queue processing latency, the queue-controller workqueue pending process object finally reduced.