-
Notifications
You must be signed in to change notification settings - Fork 77
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
Rename parameters for Distributed Workload Generation - Issue 258 #407
Rename parameters for Distributed Workload Generation - Issue 258 #407
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Changes look good overall. Was curious on the screen shot, do you know why the last operation only had 77% complete and there was no report published?
Hi @IanHoang , Here is the screenshot of the log from the coordinator node. It shows a warning as it cannot reach the worker ips by GET. |
@dosa-chammandi That makes sense, can you clear the data in the worker ips by deleting the corpus in |
Hi @IanHoang , |
…ensearch-project#407) Signed-off-by: vivek palakkat <[email protected]>
…ensearch-project#407) Signed-off-by: vivek palakkat <[email protected]> Signed-off-by: AkshathRaghav <[email protected]>
… 258 (opensearch-project#407)" This reverts commit 3adacc0.
Description
The terminology used in Distributed Workload Generation is updated.
Once all the hosts have had the daemon successfully start on them, we run the following command on the host we designated as --coordinator-ip.
Run test command on coordinator
opensearch-benchmark execute_test --pipeline=benchmark-only --workload= --load-worker-coordinator-hosts=<IP2,IP3> --target-hosts=
The parameter "--load-worker-coordinator-hosts" is renamed to "worker-ips" to be consistent with --node-ip and --coordinator-ip.
Issues Resolved
[#258]
Testing
To use Distributed Workload Generation, users have to first start the opensearch-benchmark daemon on all hosts they plan to distribute load from. One of the hosts --referred to as the load coordinator host -- coordinates requests while the others -- referred to as worker-ips -- will drive the requests / load. For example, if we have three hosts -- denoted by IP1, IP2, and IP3 --, we would do run the corresponding command on each host:
Run on IP1 Host
opensearch-benchmarkd start --node-ip=<IP1> --coordinator-ip=<IP1>
Run on IP2 Host
opensearch-benchmarkd start --node-ip=<IP2> --coordinator-ip=<IP1>
Run on IP3 Host
opensearch-benchmarkd start --node-ip=<IP3> --coordinator-ip=<IP1>
Once all the hosts have had the daemon successfully start on them, run the following command on the host we designated as --coordinator-ip in the commands above (in this case, IP1):
Run test command on IP1 Host
opensearch-benchmark execute-test --pipeline=benchmark-only --workload=<workload> --worker-ips=<IP2,IP3> --target-hosts=<endpoint>
This starts the test. IP1 is the coordinator and delegates a portion of ingestion and search requests for IP2 and IP3 to send to the target cluster.
It is expected to run the tests successfully. The following screenshot is an output from a test.
By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license.
For more information on following Developer Certificate of Origin and signing off your commits, please check here.