[SERVER-65321] Create microbenchmark that tracks latency distribution for queueing operations. Created: 07/Apr/22 Updated: 06/Dec/22 Resolved: 15/Jun/22 |
|
| Status: | Closed |
| Project: | Core Server |
| Component/s: | None |
| Affects Version/s: | None |
| Fix Version/s: | None |
| Type: | Task | Priority: | Major - P3 |
| Reporter: | Daniel Gomez Ferro | Assignee: | Backlog - Storage Execution Team |
| Resolution: | Won't Do | Votes: | 0 |
| Labels: | None | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Assigned Teams: |
Storage Execution
|
| Participants: |
| Description |
|
The library we use for microbenchmarks (Google benchmark) doesn't report latency distributions, just latency averages. We are very interested in outliers and tail latencies for this project, we need a microbenchmark that generates useful latency distributions. If rolling our own solution or incorporating a third party library requires too much effort, a standalone, off-tree experiment would be acceptable to help us choose the right implementation for the queue. |