Is there a configuration in esrally to replicate the track data X times or run index-append for a specific period of time?

Elastic Stack Elasticsearch
1

Hello,
I'm a newbie to elastic and esrally. I'm using esrally to benchmark our elastic cluster. Currently, I want to measure the index-append rate of our cluster using our sample data. With the default configuration, the esrally send the track data only one time which last for a few minutes, however, I want to run the test for a few hours to test the stability of our cluster. I cannot create a very large track data for such a test. I need an input parameter for esrally which specifies the numbers of replication of the track data or specifies the time of index-appending.

Thank you for your help!

2

You should be able to use the laps parameter to run the track multiple times in a row.

3

Thank you!
Using laps parameter there is a pause between each test. I know the pause time is quite small, but is there a way to have no pause between tests?

4

Then you probably need to modify the structure of the track. Am not sure exactly how to best do that if you have a single file you want to repeatedly replay.

It would be useful to have a parameter source that is able to generate infinite amounts of data based on sample files and I created an issue for this some time back. Please add comments and/or a +1 to this.

5

I have thumb up the issue. I hope that this feature will be added soon!

6

Yes, I think it is a feature that would make it a lot easier to work with Rally for a lot of people.

7

One alternative approach is described in this comment using jinja2 loops. One catch is that to do that it uses multiple indices, but in many cases (depending on the size of the corpus) this could be a realistic scenario.

8

Yes, it is very similar to what I need. Thank you!

9

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.