[SERVER-61244] Investigate performance of $group + $top compared to covered $sort + $group + $first plan against timeseries collection Created: 04/Nov/21  Updated: 16/Nov/21  Resolved: 16/Nov/21

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

Type: Task Priority: Major - P3
Reporter: Mihai Andrei Assignee: Ruslan Abdulkhalikov (Inactive)
Resolution: Done Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Sprint: QE 2021-11-15, QE 2021-11-29
Participants:

 Description   

This should also compare the $group + $top pipeline against the $lookup workaround that exists.



 Comments   
Comment by Ruslan Abdulkhalikov (Inactive) [ 10/Nov/21 ]

pawel.terlecki 

  1. yes, single laptop (and seems like not very fast lol)
  2. I think it is warm data since I ran count() first. I can re-test it on cold data to have a better picture
Comment by Pawel Terlecki [ 09/Nov/21 ]

The new optimization is crushing it. Nice.

1. Is it a single machine test?

2. Were the measurements taken on cold or warm data caches? It seems to me that all the data should fit in memory in all these cases. If the caches were indeed warm, how could count take a second or $group/topn 14 seconds?

Generated at Thu Feb 08 05:51:56 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.