[SERVER-11734] Query performance degradation for 2.5-pre Created: 15/Nov/13  Updated: 11/Jul/16  Resolved: 17/Dec/13

Status: Closed
Project: Core Server
Component/s: Querying
Affects Version/s: 2.5.3
Fix Version/s: 2.5.5

Type: Task Priority: Major - P3
Reporter: Rui Zhang (Inactive) Assignee: Rui Zhang (Inactive)
Resolution: Done Votes: 0
Labels: 26qa
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Participants:

 Description   

2.5-pre nightly show 30% performance query degradation when comparing to 2.4.8 in some scenario

Test setup

  • the single collection has 1040162 doc
  • each doc is of size 1264, and have two field "_id" and "name" of type NumberLong with identical value. Both fields are indexed.
  • run query find({ "_id": randInt()).one() vs find( {"name": randInt() }

    ).one()

Version query with "_id" query with "name"
2.4.8 ops/s 14549 13510.6
2.4.8 CPU 29% 49%
2.5-pre (2013-11-13) ops/s 14319.4 10793.4
2.5-pre CPU 60% 85%

query explain show following difference

  • 2.5-pre show "nscannedObjectsAllPlans" : & "nscannedAllPlans" as 2, and 2.4.8 show them as 1

test is done with Ec2 m1.xlarge instance, and can be reproduced consistently



 Comments   
Comment by Rui Zhang (Inactive) [ 17/Dec/13 ]

based on test results, the initial issue is fixed by optimization. close this ticket.

Comment by Daniel Pasette (Inactive) [ 07/Dec/13 ]

please re-test with nightly.

Generated at Thu Feb 08 03:26:36 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.