Sorry, I’m sending the message back to you;
I was also on Gladys 3 and I could barely figure out why something wasn’t working with putty, but now with Gladys 4, I’m a bit overwhelmed. I mentioned it briefly in a comment on a video you made, but I saw that gaetanb76 had asked « How to concretely test? »
I think you need to get the image here Images 
to test development services
But after…?
It was in June, (but remained unanswered)
And I would also like to test the images but how to do it?
Anyway, thanks for all the work done by the community and especially by you
Thanks for everything
These are Docker images, so if you don’t have the basics (this is not a criticism), I would advise you not to start without having some knowledge of the subject.
If you tell us which image you want to test, we can give you the command to execute.
I split the topic because it didn’t have much to do with the original issue in the other place 
Good evening VonOx,
no problem, indeed I don’t have the basics of Docker.
For the image, I had an RFlink on Gladys 3 to control my 433Mhz devices and I wanted to reuse it, so I wanted to test the Arduino image
thanks for your answers 
Pierre Gilles:
I try to put the answers on the right topics, but well, it doesn’t always work,
thanks for the message management
The Arduino image doesn’t seem to be under active development, and the PR doesn’t have much to do with the Rflink, so it won’t help you
However, there is an Rflink PR that is quite advanced (or almost finished), I invite you to check out this post:
After that, as @VonOx said, we are talking about a service under development, so if you want to be able to launch it, you need to have some skills.
Hello PierreGilles,
This is what I would like to do: test PRs. I have installed the entire Windows Dev environment, but I don’t know how to proceed further (I don’t know if this is what we use to test PRs?). I was supposed to test the PR on the Sonoff eWeling, but I don’t know the steps to get there. I think that with a small description like « Develop a Gladys service, » I could deepen what I don’t know, and I would have a model to follow (like in the good old days of V3 
)
Thanks again for taking the time for us.
If it’s for testing purposes only (not for development), it’s better to launch a Docker image « in real conditions » rather than doing this in a development environment (which doesn’t really reflect reality and isn’t necessarily easy to set up).
To launch a Docker image, the process is the same as described on the Gladys installation page via Docker (Installation avec Docker | Gladys Assistant), i.e., launching a Docker container on the machine where you want to test Gladys (as close as possible to what you have in production, a Raspberry Pi for example).
The only difference is that you need to change the name of the image in the docker run (to launch an image generated by another developer), and to set volumes that correspond to your test environment.
Let’s break down the docker run command from the site:
docker run -d \
--log-opt max-size=10m \
--restart=always \
--privileged \
--network=host \
--name gladys \
-e NODE_ENV=production \
-e SERVER_PORT=80 \
-e TZ=Europe/Paris \
-e SQLITE_FILE_PATH=/var/lib/gladysassistant/gladys-production.db \
-v /var/run/docker.sock:/var/run/docker.sock \
-v /var/lib/gladysassistant:/var/lib/gladysassistant \
-v /dev:/dev \
-v /run/udev:/run/udev:ro \
gladysassistant/gladys:v4
Here you notice that at the end, we specify that we want to launch a container from the image gladysassistant/gladys:v4, this is the « official » image that we distribute
In the case of a test scenario, for example in the case of the RFLink PR, @ProtZ shared a Docker image on his personal Docker Hub account, and the image is named ngeissel/gladys:rflinkdev.
If you want to launch an RFLink test image on your Pi, I can suggest the following command:
docker run -d \
--log-opt max-size=10m \
--restart=always \
--privileged \
--network=host \
--name gladys-test-rflink \
-e NODE_ENV=production \
-e SERVER_PORT=8001 \
-e TZ=Europe/Paris \
-e SQLITE_FILE_PATH=/var/lib/gladysassistant/gladys-production.db \
-v /var/run/docker.sock:/var/run/docker.sock \
-v /var/lib/gladysassistant_test_rflink:/var/lib/gladysassistant \
-v /dev:/dev \
-v /run/udev:/run/udev:ro \
ngeissel/gladys:rflinkdev
You will notice several changes:
- I changed the name of the container to « gladys-test-rflink » to avoid a conflict if you already have a Gladys running.
- I changed the port on which Gladys listens to 8001 to avoid a conflict if you already have a Gladys running.
- I changed the Gladys data volume (folder in which Gladys data will be mounted on the disk), here to set « /var/lib/gladysassistant_test_rflink » (Thus, if you want to clean your pi after running this file, you just need to delete this folder).
- I changed the image to « ngeissel/gladys:rflinkdev ».
Once your container is running, you will be able to access Gladys by typing « IP_RASPBERRY_PI:8001 ».
If you want to stop this container:
docker stop gladys-test-rflink
If you want to delete this container:
docker rm gladys-test-rflink
If you want to fetch a new version of the image
(if the developer has made progress on his integration):
docker pull ngeissel/gladys:rflinkdev
Note that a Docker container is immutable: if you have pulled a new image, you need to stop the container, delete the container, and then docker run a new container
Conclusion
Here is an example in the case of RFLink, but you can change the name of the image to whatever you want.
It might be Chinese if you don’t know Docker, in my opinion it’s worth reading the Docker basics if you don’t know much about it
The Sonoff Ewelink is already merged and available in Gladys 4 for a while now!
Hello Pierre Gilles
Thank you very much for this detailed explanation, I will now be able to try testing, now that I have a guide.
I’m already demystifying Docker by following YouTube tutorials.
Yes, I know, and now it works very well and without latency for me, but testing was a few months ago on a PR from Ptit-Nico, after the changes he made.
In any case, thank you, to you and the community, for all the work done and the support you provide us with,
No problem Don’t hesitate if you have any issues during your tests!
@pierre-gilles I’m reactivating this discussion to ask essentially the same question as Psoy: how to test a PR in development, in the case where there is no complete image integrating this PR?
@jparbel Two possibilities:
-
Ask the developer to set up an image, or set up an image yourself (It’s just clicking in GitHub, no commands to execute…). This is the simplest solution, it takes 30 seconds to click a button on GitHub to start a build, and then about 1 hour for an initial build, then 5 minutes for subsequent builds thanks to Docker cache
-
Set up a Gladys development environment, but this is intended for an advanced audience who already knows how to develop on a Node.js / React stack (Tutorial on the site under MacOS/Linux => Mettre en place un environnement de développement sous Mac/Linux | Gladys Assistant or Windows => Mettre en place un environnement de développement sous Windows | Gladys Assistant). I do not recommend starting down this path if you have never developed on this stack (after all, it can be learned, but it’s not done in 5 minutes, it’s a profession :D)
Following up on the last message, do you know how to properly configure the creation of a Docker test image?
Here’s what I did to set up the builds:
1 - Created a Docker Hub account + a repo to receive the test images bertrandda/gladys
2 - On my GitHub fork, I added 3 secrets DOCKERHUB_USER « bertrandda », DOCKERHUB_PASSWORD the access token generated from my Docker account, and DOCKERHUB_REPO « gladys »
3 - Launched the image creation from « Actions » → « Build gladys dev images »
But it keeps failing at the end, do you know why? https://github.com/bertrandda/Gladys/actions/runs/1998893565
We see « authorization failed » in the log
Is the user/password incorrect?
Edit: you need to put bertrandda/gladys for the repo
For the username/password it should be fine, the « Login to DockerHub » part is well Succeeded.
I just restarted with bertrandda/gladys in the repo, I’ll let you know if it works.
Thanks @VonOx it works now
Hello,
When testing, should we create a local account with our email or can we restore a Gladys Plus backup?
If we choose to create a local account, can we use the same email address or should we use a different one?
Create a local account, I advise against restoring a Gladys Plus backup, because that will connect this test build to your Gladys Plus production account (and therefore disconnect your real instance)
It doesn’t matter, put toto@toto.fr, it doesn’t change anything The email isn’t used, it’s just a purely local identifier for the login