-
Notifications
You must be signed in to change notification settings - Fork 537
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
There are summaries that summarize too many ops, indicating we are not summarizing often enough #10007
There are summaries that summarize too many ops, indicating we are not summarizing often enough #10007
Conversation
We need to put note into breaking.md and tell the world that
This is important, as most likely these settings will not work well for WB - their summaries are pretty heavy. Please also consider if we want to do #9990 first. I was looking around and discovered a lot of duplication there that sort made sense when we had it provided by service, but now it's one thing. I'd probably keep ISummaryConfiguration and move all summary related settings into ISummaryConfiguration. |
I'd rather open an issue to look into it. I see 1200 SummaryAckWaitTimeout events in telemetry over last 30 days. These are bugs. It's very hard to say without looking deeper into it if these bugs would be mitigated by lower timeout. |
I've opened #10008 with my thoughts on this. |
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.
Looks good, but we have to put a warning in breaking.md about this change and the fact that each experience should test if it works for them and see if they need to provide their own settings as opposed to relying on defaults.
Fix #9908
Initial attempt to adjust the Summary Configuration options now that we do not rely on the Service to provide them:
a) MaxOps (from 1000 - 100 ops)
b) MaxAckWaitTime (from 10 mins down to 2 mins).
@vladsud - I'm not sure if changing idleTime right now, together with MaxOps is the right thing to do - should we wait for more data ? Currently, idleTime is 15 secs.
@agarwal-navin FYI.