Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Performance degradation with 7.13 against 7.12 #5910

Closed
MartinDemberger opened this issue Jul 16, 2021 · 1 comment
Closed

Performance degradation with 7.13 against 7.12 #5910

MartinDemberger opened this issue Jul 16, 2021 · 1 comment

Comments

@MartinDemberger
Copy link

MartinDemberger commented Jul 16, 2021

NEST/Elasticsearch.Net version:
NEST 7.12.1

Elasticsearch version:
7.13.0

.NET runtime version:
net5.0

Operating system version:
Windows 10

Description of the problem including expected versus actual behavior:
When I run a query with an aggregation the execution ist much slower against a elasticsearch server 7.13 than against 7.12.

With 7.13 the query is executed in an average von 4.8 ms
With 7.12 the query is executed in an average von 0.1 ms

I'm not sure if this is a problem of the NEST client or the elastic server. It this is the wrong place can you please move the report to correct repository.

Steps to reproduce:
https://gist.github.com/MartinDemberger/a58465ba5ddc2042ae1756040b2318df

  1. Start a local elastic server with version 7.13
  2. Execute the test in the appended project 10x and note the output
  3. Start a local elastic server with version 7.13
  4. Execute the test in the appended project 10x and note the output

Expected behavior
The runtime should be almost equal

Provide ConnectionSettings (if relevant):

Provide DebugInformation (if relevant):

@stevejgordon
Copy link
Contributor

stevejgordon commented Jul 16, 2021

Hi @MartinDemberger. Thanks for raising this and the comprehensive repro code. I made some modifications to analyse results over multiple searches, and did manage to reproduce this. After speaking to our search engineers they believe this is the result of a known issue which is fixed in this PR which will be released in 7.14.0. I ran a further test against a pre-release version of 7.14.0 and indeed, the performance regression is removed. In fact, the results are looking faster overall.

7.12.0:
Aggs: 35,305.92ns
Collector: 17,084.70ns
Query: 160,051.20ns
Search loop time: 3,087ms

7.12.1:
Aggs: 38,216.61ns
Collector: 19,089.20ns
Query: 168,745.60ns
Search loop time: 3,518ms

7.13.0:
Aggs: 792,880.90ns
Collector: 805,869.18ns
Query: 53,972.00ns
Search loop time: 5,504ms

7.13.3:
Aggs: 765,656.90ns
Collector: 790,476.64ns
Query: 54,005.90ns
Search loop time: 4,773ms

7.14.0:
Aggs: 18,465.88ns
Collector: 17,370.50ns
Query: 96,179.00ns
Search loop time: 2,481ms

I don't think it's worth transferring this given that a fix is already merged.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants