[SERVER-43939] Add service time metric to server status Created: 10/Oct/19 Updated: 22/Jul/20 Resolved: 22/Jul/20 |
|
| Status: | Closed |
| Project: | Core Server |
| Component/s: | None |
| Affects Version/s: | None |
| Fix Version/s: | None |
| Type: | New Feature | Priority: | Major - P3 |
| Reporter: | Kevin Arhelger | Assignee: | Michael Gargiulo |
| Resolution: | Incomplete | Votes: | 4 |
| Labels: | None | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Issue Links: |
|
||||
| Sprint: | Execution Team 2019-12-30, Execution Team 2019-12-30, Execution Team 2020-07-13, Execution Team 2020-07-27 | ||||
| Participants: | |||||
| Description |
|
The average latency metrics in serverStatus are very useful, except when queuing is occurring on the server. When queuing begins, the time waiting for locks and tickets is counted average latency. A new metric that measures just the time an operation was running (or the time spent waiting for locks and tickets allowing for a derived metric) would help better explain why queuing occurred (the service time increased). |
| Comments |
| Comment by Eric Milkie [ 22/Jul/20 ] |
|
The Admission Control project will address service time metrics. |
| Comment by Eric Milkie [ 07/Jan/20 ] |
|
I'm going to take this out of Quick Wins – with the locking changes we're making in the master branch it's unclear how much of a benefit this will be. |