[SERVER-35721] Improve the statistics that BenchRunner::finish returns Created: 21/Jun/18  Updated: 06/Dec/22  Resolved: 05/Nov/21

Status: Closed
Project: Core Server
Component/s: Performance
Affects Version/s: None
Fix Version/s: None

Type: Task Priority: Major - P3
Reporter: Dianna Hohensee (Inactive) Assignee: Backlog - Server Tooling and Methods (STM) (Inactive)
Resolution: Won't Fix Votes: 0
Labels: tig-benchrun
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Issue Links:
Related
related to SERVER-61295 Complete TODO listed in SERVER-35721 Closed
Assigned Teams:
Server Tooling & Methods
Participants:

 Description   

This bit of code is responsible for returning different types of operations per second, but doesn't label them as per second. E.g., instead of returning

{..., "inserts/sec": 50, ...}

, it returns

{..., "insert": 50, ...}

. This is misleading. errCount output, meanwhile, is total number of ops failed, not per second, and not by op. It is therefore hard to compare succeeded and failed ops.

Requesting the following particular additions:

  • Change misleading "<op_name>" field names to "<op_name>/sec", which is correct.
  • Add total successful ops for each op type.
  • Add errCount result per op type.
  • Add errCount per op type per second, for ease of comparison to op type per second – I'm extrapolating the usefulness of this one, it's not something I ran into wanting.

This may require workload updates in the workloads code repository as well, since the results of benchRun are returned and parsed by string. E.g. any workloads reporting particular ops/sec use res["insert"] to get what will be changed to res["inserts/sec"].


Generated at Thu Feb 08 04:40:43 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.