fix: Unload flow when removing script instead of removing

[Bravo555]

Proposed changes

Unload flow when one of its scripts is removed instead of removing. Removing sometimes made flow not start back up when script was restored. Instead, unloading keeps it in an unloaded list that is checked when relevant script is reloaded.

This is an alternative fix to #4031 using the same test suite, but addresses flow not being reloaded without assuming stale, out-of-order FileDeleted messages, which I could not reproduce. Instead my hypothesis is that the flow wasn't being reloaded because reload_script looks at unloaded flows list to restart flows, but the flow wasn't in that list because it was removed instead of just unloaded.

Types of changes

• Bugfix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Improvement (general improvements like code refactoring that doesn't explicitly fix a bug or add any new functionality)
• Documentation Update (if none of the other choices apply)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)

Paste Link to the issue

Checklist

• I have read the CONTRIBUTING doc
• I have signed the CLA (in all commits with git commit -s. You can activate automatic signing by running just prepare-dev once)
• I ran just format as mentioned in CODING_GUIDELINES
• I used just check as mentioned in CODING_GUIDELINES
• I have added tests that prove my fix is effective or that my feature works
• I have added necessary documentation (if appropriate)

Further comments

[codecov]

Codecov Report

❌ Patch coverage is 0% with 19 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
crates/extensions/tedge_flows/src/actor.rs	0.00%	14 Missing ⚠️
crates/extensions/tedge_flows/src/registry.rs	0.00%	5 Missing ⚠️

📢 Thoughts on this report? Let us know!

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[github-actions]

Robot Results

✅ Passed	❌ Failed	⏭️ Skipped	Total	Pass %	⏱️ Duration
852	0	3	852	100	2h43m29.694894s

[reubenmiller]

I was testing with the PoC sm-plugin for a flow, and upgrading an existing flow results in the following log entries:

Mar 11 20:17:08 7e289213ec5f tedge-mapper[808]: 2026-03-11T20:17:08.746513095Z  WARN flows: Removing a script used by a flow protobuf-xform: /etc/tedge/mappers/local/flows/protobuf-xform/lib/main.js
Mar 11 20:17:08 7e289213ec5f tedge-mapper[808]: 2026-03-11T20:17:08.757372524Z  INFO flows: Reloading flow script /etc/tedge/mappers/local/flows/protobuf-xform/lib/main.js

The WARN level for the flow removal is overkill (INFO would be more appropriate), however interesting enough, #4031, only yields a single "Reloading flow" log entry on updates (which is preferable from a users perspective given an update for a user should trigger a reload and the "remove" operation is unnecessary (again from the users point of view).

[Bravo555]

The WARN level for the flow removal is overkill

Indeed we were removing the flow first and the script after, in order to not trigger that warning, but it would be better to adjust it to not appear if removing a script for a flow that's unloaded.

[Bravo555]

Removed the extra warning and added one more test case for moving a script.

[reubenmiller]

Removed the extra warning and added one more test case for moving a script.

I just tested the update version and there is a lot of noise when updating a single flow (via the poc sm-plugin). The following entries show with 1 software update (which replaces the flow.toml and lib/main.js)...6 log entries in total.

Mar 12 15:20:34 7a272ddfd7d7 tedge-mapper[2303]: 2026-03-12T15:20:34.835061855Z  INFO flows: Loading flow /etc/tedge/mappers/local/flows/protobuf-xform/flow.toml
Mar 12 15:20:34 7a272ddfd7d7 tedge-mapper[2303]: 2026-03-12T15:20:34.848035591Z  INFO flows: Removing flow /etc/tedge/mappers/local/flows/protobuf-xform/flow.toml
Mar 12 15:20:34 7a272ddfd7d7 tedge-mapper[2303]: 2026-03-12T15:20:34.848198424Z  INFO flows: Loading flow /etc/tedge/mappers/local/flows/protobuf-xform/flow.toml
Mar 12 15:20:34 7a272ddfd7d7 tedge-mapper[2303]: 2026-03-12T15:20:34.874713772Z  INFO flows: Reloading flow script /etc/tedge/mappers/local/flows/protobuf-xform/lib/main.js
Mar 12 15:20:34 7a272ddfd7d7 tedge-mapper[2303]: 2026-03-12T15:20:34.874764272Z  INFO flows: Unloading flow /etc/tedge/mappers/local/flows/protobuf-xform/flow.toml
Mar 12 15:20:34 7a272ddfd7d7 tedge-mapper[2303]: 2026-03-12T15:20:34.874907939Z  INFO flows: Loading flow /etc/tedge/mappers/local/flows/protobuf-xform/flow.toml

In comparison to the other PR #4031, which produced the following entries (but yes also not perfect).

Mar 12 15:29:40 b92dc435bbf5 tedge-mapper[820]: 2026-03-12T15:29:40.978003119Z  INFO flows: Loading flow /etc/tedge/mappers/local/flows/protobuf-xform/flow.toml
Mar 12 15:29:40 b92dc435bbf5 tedge-mapper[820]: 2026-03-12T15:29:40.992939339Z  INFO flows: Loading flow /etc/tedge/mappers/local/flows/protobuf-xform/flow.toml
Mar 12 15:29:41 b92dc435bbf5 tedge-mapper[820]: 2026-03-12T15:29:41.01615384Z  INFO flows: Reloading flow script /etc/tedge/mappers/local/flows/protobuf-xform/lib/main.js
Mar 12 15:29:41 b92dc435bbf5 tedge-mapper[820]: 2026-03-12T15:29:41.043625023Z  INFO flows: Reloading flow script /etc/tedge/mappers/local/flows/protobuf-xform/lib/main.js

The main point is that the first set of log entries are confusing for the user and show that we're being very inefficient when processing flows as we're removing, loading, unloading, loading...so this just makes the system harder to reason with, and I would assume it would also be consuming additional resources doing all of these things.

[Bravo555]

The main point is that the first set of log entries are confusing for the user and show that we're being very inefficient when processing flows as we're removing, loading, unloading, loading...so this just makes the system harder to reason with, and I would assume it would also be consuming additional resources doing all of these things.

Indeed, in #4031 logs are more concise, because some events are explicitly ignored, with the motivation that I think is not entirely correct (can't observe FileDeleted events), but the fix itself ends up being correct (when replacing flow assets we ignore temporary deletions when we think the files are going to be recreated soon, so there's less noise).

Then I think we can go forward with #4031, but with the addition of Removing script unloads flow and restoring script loads flow again test, added in 153278c, to make sure regular removals of script files cause the flow to be unloaded/removed as expected.

[Bravo555]

Selected alternative #4031 to be merged.