-
Notifications
You must be signed in to change notification settings - Fork 12
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
Swap performance testing implementation to use Hyperfoil. #34
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.
Some changes are requested with initial review.
deployment/inventory/.settings/org.eclipse.core.resources.prefs
Outdated
Show resolved
Hide resolved
The initial review relate to an outdated commit.
In README.md, TOC has not been updated |
what is It contains
:O |
The purpose of |
some comments left. Waiting for feedback |
Feedback provided to comments. |
I'm not familiar with the codebase, so I will not comment about the code changes, but I'd like to give my opinion as a user of the tool. We've been using this tool to benchmark many versions of 3scale. Jmeter provides all we need, which includes things like: ability to configure the number of connections, the throughput, and nice reports. Why are we changing to Hyperfoil? What problems does it solve? Also, I'm a bit concerned about changing the tool because from the moment we change, we won't be able to compare with previous results. Has anyone tried to run the same benchmarks we've run for the latest version of 3scale with hyperfoil to see how results compare? |
Swapping out the driver to use Hyperfoil has been advocated so that load testing can be achieved easily in a microservices architecture. The driver also supports the ability to distribute the load generation. Amongst one or more Agents. |
JMeter also supports Distributed Testing https://jmeter.apache.org/usermanual/jmeter_distributed_testing_step_by_step.html |
The following params
are duplicated in several files called
and
Please, make sure more attributes are not duplicated and only used files are being committed. |
I was thinking the same. |
I just tested this branch, and still get the same error:
|
Can you raise an issue on the hyperfoil_test project here please ? |
It would be better if you do to explain configuration and environment. I am not familiar with hyperfoil parameters and configuration. |
It isn't anything to do with Hyperfoil as far as I can see. Don't you think this is an Ansible error reported for the shell module execution ?
then execute the run.yml playbook again to see if this solves the problem ? |
If you suggest to modify some third party source of code, please open issue to that repo suggesting to them, not to me. If it fails to me, it could fail to anyone else using it. |
I don't mind creating the issue against the hyperfoil_test project for you to help you with this issue. |
Even worse message,
Does it work for you? |
On a RHEL/Fedora platform it has always worked. |
Last attempt. This PR is not the place to test your roles.
|
From this test it looks like the issue with the Ansible "conflicting action statements: shell, warn" has been solved. I'll raise an issue with the relevant hyperfoil_test project. |
Is it ready to be tested again? I see there are new changes. New review is needed. |
Yes it is ready to be tested. |
Proposed refactor: #54 waiting for you input @whitingjr |
Another proposed refactor: #59 |
…. Added role to install dependencies. Added platform specific package name variable loading for Java and unzip. Aligned interface with shared connections. Removed duplicate variables.
This PR will replace the JMeter injector with Hyperfoil.
Included in this PR are an integration with Ansible playbooks. Updates to the documentation for users to get started with the replacement injector.