Task #9150
closedAs a Pulp user, I want to validate that my setup is working
0%
Description
Ticket moved to GitHub: "pulp/pulpcore/2032":https://github.com/pulp/pulpcore/issues/2032
This has been reported in the Pulp Community Survey as a concern that some "Checks" documentation would help alleviate.
"how do I validate my setup is working." ... I've written my own checkout test cases to address this because it wasn't obvious how to reuse the Pulp CI checks.
Can we provide documentation as to how a user could reuse the Pulp CI checks? Alternatively, can we provide some guidance around how users can know when all is good?
Updated by daviddavis over 3 years ago
We have some basic app checks:
https://github.com/pulp/pulpcore/blob/master/pulpcore/app/checks.py
They use the django system check framework:
https://docs.djangoproject.com/en/3.2/topics/checks/
Maybe we could document how to run these checks and also add some more checks.
Updated by daviddavis over 3 years ago
mcorr, can you be more explicit about what sorts of checks might benefit users?
Also, today during our pulpcore meeting we discussed relying on the status api (accessible at /pulp/api/v3/status/ and via the pulp status
CLI command), which should give some picture about the health of the system (see example below). So maybe it's just worth documenting this?
{
"versions": [
{
"component": "core",
"version": "3.15.0.dev"
},
{
"component": "file",
"version": "1.9.0.dev"
},
{
"component": "ansible",
"version": "0.10.0.dev"
}
],
"online_workers": [
{
"pulp_href": "/pulp/api/v3/workers/2fc7dfd6-3d4d-4a3d-8b2a-a9f641fbc4d7/",
"pulp_created": "2021-08-03T12:39:44.395610Z",
"name": "211023@pulp3-source-fedora.crake.example.com",
"last_heartbeat": "2021-08-03T13:56:49.351986Z"
}
],
"online_content_apps": [],
"database_connection": {
"connected": true
},
"redis_connection": {
"connected": true
},
"storage": {
"total": 42006183936,
"used": 3561533440,
"free": 36280442880
}
}
Updated by mcorr over 3 years ago
Hi David. No idea why I didn't get notified of this!
The idea I had in my head was something like the Docker hello world scenario, where in their troubleshooting docs, you usually run that docker hello world command to see can you communicate with Docker and pull images, before you go hunting for a bigger problem.
Your suggestion looks good also!
Updated by ipanova@redhat.com about 3 years ago
- Quarter changed from Q3-2021 to Q4-2021
Updated by pulpbot almost 3 years ago
- Description updated (diff)
- Status changed from NEW to CLOSED - DUPLICATE