Fediversity/Fediversity
13
2
Fork 6
Code Issues 111 Pull requests 15 Projects 1 Activity Actions

Add a basic integration test #323

Merged
fricklerhandwerk merged 4 commits from Niols/Fediversity:integration-test into main 2025-04-30 15:03:37 +02:00
Conversation 5 Commits 4 Files changed 12 +300 -526
Niols commented 2025-04-29 17:54:28 +02:00
Owner
Copy link

This PR adds a basic deployment test to the repository. This test will, in a NixOS test, run a deployer VM and a target VM, and check that we manage to run nixops4 apply on the deployer VM to change things on the target VM. The ideas are all @roberth's and this test has been extremely heavily inspired by https://github.com/nixops4/nixops4-nixos/blob/main/test/default/nixosTest.nix.

The interesting commit is bb605124ceb146e26ce344653df2d6d62ab12354.

Commits 64a015eca28095c7b18556d078b2b1ee8dd80776 and 1f3f8c9f5dde0a6c75bfe555b784addaa9695871 are lock manipulation commits that improve the run time of the test by removing nixpkgs duplication. They will probably clash with #311. The former, we should just have done a while ago. The latter is more debatable and I would rather do without, but it does half the test running time (from 4 to 2 minutes) on my test machine.

The goal of this PR is to serve as basis for future work, namely:

  1. A basic deployment of several machines, factorising the bits that can be reused. [WIP - needs polishing but otherwise looking good]
  2. A full deployment of Fediversity services triggered from command line. [WIP - just started]
  3. A full deployment of Fediversity services triggered from the panel.
This PR adds a basic deployment test to the repository. This test will, in a NixOS test, run a deployer VM and a target VM, and check that we manage to run `nixops4 apply` on the deployer VM to change things on the target VM. The ideas are all @roberth's and this test has been extremely heavily inspired by https://github.com/nixops4/nixops4-nixos/blob/main/test/default/nixosTest.nix. The interesting commit is bb605124ceb146e26ce344653df2d6d62ab12354. Commits 64a015eca28095c7b18556d078b2b1ee8dd80776 and 1f3f8c9f5dde0a6c75bfe555b784addaa9695871 are lock manipulation commits that improve the run time of the test by removing nixpkgs duplication. They will probably clash with https://git.fediversity.eu/Fediversity/Fediversity/pulls/311. The former, we should just have done a while ago. The latter is more debatable and I would rather do without, but it does half the test running time (from 4 to 2 minutes) on my test machine. The goal of this PR is to serve as basis for future work, namely: 1. A basic deployment of several machines, factorising the bits that can be reused. _[WIP - needs polishing but otherwise looking good]_ 2. A full deployment of Fediversity services triggered from command line. _[WIP - just started]_ 3. A full deployment of Fediversity services triggered from the panel.
kiara reviewed 2025-04-30 10:57:28 +02:00
flake.nix Outdated
@ -1,13 +1,14 @@
{
inputs = {
nixpkgs.url = "github:nixos/nixpkgs/nixos-24.11";
nixpkgs.follows = "nixops4/nixpkgs";
kiara commented 2025-04-30 10:57:28 +02:00
Owner
Copy link

i think this fell afoul of #318, breaking the peertube CI test

i think this fell afoul of #318, breaking the peertube CI test
Niols commented 2025-04-30 11:00:43 +02:00
Author
Owner
Copy link

Yeah, that's also what I thought. I just now force-pushed a version of this PR without that commit, and we can discuss them independently.

Yeah, that's also what I thought. I just now force-pushed a version of this PR without that commit, and we can discuss them independently.
❤️ 1
kiara marked this conversation as resolved
Niols force-pushed integration-test from 1f3f8c9f5d to 8503e45367 2025-04-30 10:58:24 +02:00 Compare
Niols force-pushed integration-test from 8503e45367 to 95fa8ab2e6 2025-04-30 10:59:57 +02:00 Compare
Niols commented 2025-04-30 11:02:00 +02:00
Author
Owner
Copy link
unexpected client error: unexpected requesting "https://code.forgejo.org/actions/checkout/info/refs?service=git-upload-pack" status code: 503

Service unavailable - amazing. Indeed, code.forgejo.org seems down. Or when you're reminded that decentralisation in open source is not an easy problem.

Edit: it was transient and is already gone.

``` unexpected client error: unexpected requesting "https://code.forgejo.org/actions/checkout/info/refs?service=git-upload-pack" status code: 503 ``` Service unavailable - amazing. Indeed, code.forgejo.org seems down. Or when you're reminded that decentralisation in open source is not an easy problem. Edit: it was transient and is already gone.
🎉 1 😆 1
kiara changed title from Add a basic integration test to WIP: Add a basic integration test 2025-04-30 11:09:21 +02:00
Niols commented 2025-04-30 11:27:56 +02:00
Author
Owner
Copy link

In fact, this run was shorter than the one with nixpkgs.follows = "nixops4/nixpkgs", but I think this is mostly due to things being cached on the runner this time around. Either way, looks like CI is happy with this piece of work. I'll add the other ones in subsequent PRs.

In fact, this run was shorter than the one with `nixpkgs.follows = "nixops4/nixpkgs"`, but I think this is mostly due to things being cached on the runner this time around. Either way, looks like CI is happy with this piece of work. I'll add the other ones in subsequent PRs.
❤️ 1
fricklerhandwerk changed title from WIP: Add a basic integration test to Add a basic integration test 2025-04-30 13:40:09 +02:00
kiara commented 2025-04-30 13:42:25 +02:00
Owner
Copy link

hm, would it not be better to mark this as WIP while as per the current PR description that is its current status?

hm, would it not be better to mark this as WIP while as per the current PR description that is its current status?
fricklerhandwerk approved these changes 2025-04-30 14:05:07 +02:00
fricklerhandwerk left a comment
Owner
Copy link

Nice, except the slightly weird factoring of nested imports. (Modulo failing tests due to nixpkgs bump -- do we need the bump?)

Nice, except the slightly weird factoring of nested imports. (Modulo failing tests due to nixpkgs bump -- do we need the bump?)
deployment/flake-part.nix Outdated
@ -0,0 +1,3 @@
{
imports = [ ./check/flake-part.nix ];
fricklerhandwerk commented 2025-04-30 13:45:58 +02:00
Owner
Copy link

Why are we doing those two layers of imports?

Why are we doing those two layers of imports?
👍 1
Niols commented 2025-04-30 14:30:18 +02:00
Author
Owner
Copy link

What about flake.nix importing deployment/flake-part.nix, and this one importing all the checks directly? Or would you even rather flake.nix imports all the checks directly?

What about `flake.nix` importing `deployment/flake-part.nix`, and this one importing all the checks directly? Or would you even rather `flake.nix` imports all the checks directly?
fricklerhandwerk commented 2025-04-30 14:36:38 +02:00
Owner
Copy link

I don't really care, but boilerplate files seem wrong. 💁

I don't really care, but boilerplate files seem wrong. 💁
Niols commented 2025-04-30 14:43:39 +02:00
Author
Owner
Copy link

Fixed in 59038d2f30.

Fixed in 59038d2f30600f246043e9cbeed29922cbe5ca4e.
Niols marked this conversation as resolved
fricklerhandwerk commented 2025-04-30 14:05:36 +02:00
Owner
Copy link

hm, would it not be better to mark this as WIP while as per the current PR description that is its current status?

@kiara the WIP in the PR description talks about follow-up work. This one is self-contained.

> hm, would it not be better to mark this as WIP while as per the current PR description that is its current status? @kiara the WIP in the PR description talks about follow-up work. This one is self-contained.
👍 1
kiara added a new dependency 2025-04-30 14:22:59 +02:00
#102 CI rejects failing deployments
Niols force-pushed integration-test from 95fa8ab2e6 to cc6bff2630 2025-04-30 14:43:12 +02:00 Compare
fricklerhandwerk merged commit f5db62e053 into main 2025-04-30 15:03:37 +02:00
fricklerhandwerk deleted branch integration-test 2025-04-30 15:03:38 +02:00
fricklerhandwerk commented 2025-04-30 15:04:07 +02:00
Owner
Copy link

I squashed it because the change is rather self-contained, and the individual commits didn't make all that much sense in isolation

I squashed it because the change is rather self-contained, and the individual commits didn't make all that much sense in isolation
Sign in to join this conversation.
Reviewers
No reviewers
fricklerhandwerk
Labels
Clear labels   
api service

features that may relate more to a (to be created) API service

  
blocked

   
bug

Anything that doesn't work as advertised

  
component: fediversity panel

   
component: nixops4

   
documentation

   
estimation high: >3d

   
estimation low: <2h

   
estimation mid: <8h

   
productisation

features out-scoped with the decoupling of the productisation effort

  
project-management

Project management related tasks

  
question

   
role: sysadmin

   
security

   
technical debt

   
testing

   
type unclear

   
type: key result

   
type: objective

   
type: task

   
type: user story

   
user experience
No labels
api service
blocked
bug
component: fediversity panel
component: nixops4
documentation
estimation high: >3d
estimation low: <2h
estimation mid: <8h
productisation
project-management
question
role: sysadmin
security
technical debt
testing
type unclear
type: key result
type: objective
type: task
type: user story
user experience
Milestone
Clear milestone
No items
No milestone
Projects
Clear projects
No items
No project
Assignees
Clear assignees
No assignees
3 participants
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Blocks
#102 CI rejects failing deployments}
Fediversity/Fediversity
Reference: Fediversity/Fediversity#323
No description provided.
Delete branch "Niols/Fediversity:integration-test"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?