Refactor Hyperfoil document #234
Conversation
docs/tutorial/hyperfoil.adoc
Outdated
|
|
||
| Run with and replace `targethost` value with your needs | ||
| ``` | ||
| targethost=$USER@localhost |
There was a problem hiding this comment.
why does the guide introduce the extra step of creating the environment variable? Is this necessary?
| The previous tutorial created a qDup yaml that downloaded Quarkus getting started and ran a single curl against the `quarkus dev` process. We are going to start where it left off: | ||
|
|
||
| ```yaml | ||
| == Running the experiment |
There was a problem hiding this comment.
Please add a sentence after the heading as a code section immediately after the heading can leave a user to scroll past the code section to find out why the code section exists then scrolling back up to the code section to interact with it.
docs/tutorial/hyperfoil.adoc
Outdated
| ``` | ||
|
|
||
| == Splitting the roles | ||
| The previous tutorial created an `example` role to run all of ours scripts. This works but a proper performance test will normally isolate the load generation (currently `curl`) from the process we are testing (currently `quarkus dev`). | ||
| qDup isolates scripts by assigning them to roles that are running on different `hosts`. To isolate `test-endoing` from `getting-started` we need to introduce a second role. Let us call it `hyperfoil`. The `roles` section will now look like the following: | ||
| The previous tutorial created a role to run all of ours scripts. This works but a proper performance test will normally isolate the load generator (`hyperfoil`) from the process we are testing (`quarkus dev`). |
There was a problem hiding this comment.
Removed (hyperfoil) and (quarkus dev)
docs/tutorial/hyperfoil.adoc
Outdated
|
|
||
| The first command is to `wget` the hyperfoil release but we are going to start with `cd /tmp` like we did in `getting-started` so that we don't add files to the current home directory. | ||
| ==== setup-hyperfoil | ||
| It will download, install Hyperfoil locally and configure the `single-request.hf.yaml` file to use `http://localhost:8080` instead of `http://hyperfoil.io` |
There was a problem hiding this comment.
The first sentence after a new heading should probabably use a proper noun as the topic of the first sentence instead of a pronoun. Starting with it will force a reader to go back to the previous section to identify the topic of the new section.
For this situation, the rendered version is easy to read. I have addressed all the comments. If you agree with the proposed solution, don't forget to squash it before merging.
|
docs/tutorial/hyperfoil.adoc
Outdated
|
|
||
| ```yaml | ||
| == Running the experiment | ||
| The script will clone the `quarkus-quickstarts` repository, install `quarkus-cli` if needed, and, start the application. In parallel, it will download and install link:https://hyperfoil.io[Hyperfoil]. Once the application is ready, it will start a simple benchmark. In the end, it will upload the result and stop the `quarkus-quickstarts` application. |
There was a problem hiding this comment.
it will upload the result
I think you mean it will download the result to qDup
There was a problem hiding this comment.
I think you mean it will download the result to qDup
No. It will update the result to Horrem.
Maybe should I update the sentence adding "to Horreum" ?
There was a problem hiding this comment.
I don't see anything in the hyperfoil.adoc that would upload it to Horreum. Am I missing a curl in the guide?
https://github.com/diegolovison/qDup/blob/hyp_doc/docs/tutorial/hyperfoil.adoc
I personally do not think the example should include Horreum upload because the example is for Hyperfoil not Horreum + Hyperfoil and keeping the user focused is valuable.
No description provided.