fix: router healthcheck incorrectly warns when all projects stopped, fixes #8081 by Copilot · Pull Request #8082 · ddev/ddev

Copilot · 2026-01-30T15:39:49Z

The Issue

Stopping all projects can lead into a "no config files found or no routers expected" warning #8081

When all DDEV projects are stopped, the router displays: "WARNING: No config files found or no routers expected". This is a valid healthy state that should not generate warnings.

How This PR Solves The Issue

1. Fixed healthcheck logic (traefik_healthcheck.sh):

Modified to treat expected_router_count == 0 with error_count == 0 as healthy
Previous logic required expected_router_count > 0, causing the zero-router state to trigger warnings
Removed misleading "No config files found" warning message

2. Added proactive healthcheck clearing (router.go, stop.go):

Created ClearRouterHealthcheck() function to force immediate healthcheck execution
Refactored StartDdevRouter() to use the new reusable function
Call ClearRouterHealthcheck() after ddev stop completes

This ensures router status updates immediately when projects stop, rather than waiting up to 59 seconds for the next periodic healthcheck cycle.

Manual Testing Instructions

Start two DDEV projects: ddev start in each project directory
Run ddev list - verify router shows "OK" status at ~/.ddev
Stop both projects: ddev stop in each directory
Run ddev list immediately - verify router shows "OK" status at ~/.ddev WITHOUT the warning message
Repeat with variations:
- Stop all projects, start one, stop it, run ddev list
- Expected: Router always shows "OK" without warnings when all projects are stopped

Automated Testing Overview

Changes are to shell script healthcheck logic and Go code for router management. Testing requires:

Router container with updated healthcheck script (requires new image build)
Integration tests would need Docker environment with projects starting/stopping

The healthcheck shell script was validated with shellcheck. Go code builds successfully and follows existing patterns in the codebase.

Release/Deployment Notes

Requires new ddev-traefik-router container image to be built and published with updated healthcheck script
No breaking changes - purely fixes incorrect warning behavior
Users will see improved router status reporting when all projects are stopped

Original prompt

This section details on the original issue you should resolve

<issue_title>Stopping all projects can lead into a no config files found or no routers expected warning</issue_title>
<issue_description>### Preliminary checklist

I am using the latest stable version of DDEV (see upgrade guide)

I have searched existing issues

I have checked the troubleshooting guide

I have run ddev utility test or at least ddev utility diagnose to include output below

Output of ddev utility test or ddev utility diagnose
Expand `ddev utility test` diagnostic information
[COPY-PASTE HERE THE OUTPUT OF `ddev utility test`]
Expected Behavior

when stopping the last running project i would expect that the router status remains ok and the location ~/.ddev

Actual Behavior

instead this warning showed up

Steps To Reproduce

with the latest version of HEAD i am unable to reproduce, but randy was able to . but the gist is:

start one or more projects

stop each running project (do not power off)

then do a ddev list and check the router status

Anything else?

No response</issue_description>

<agent_instructions>Examine the reason for this anomaly. It seems to be that the router healthcheck may be in error reporting "no files". </agent_instructions>

Comments on the Issue (you are @copilot in this section)
@rfay I reproduced this a little bit ago with * Start 2 projects. * Stop both * Start one * Stop it
ddev list after each step.

I think the error message from the router is simply incorrect. It's possible to have no config files.</comment_new>

Fixes Stopping all projects can lead into a "no config files found or no routers expected" warning #8081

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.

rfay · 2026-01-30T16:07:34Z

@copilot

I assume that traefik/config/default_config.yaml is sometimes deleted (in the docker volume) by some race condition? Normally it is there when trying to reproduce this scenario.
The key here is what's in the docker volume (used by ddev-router), not directly related to what might be in ~/.ddev/traefik/config, although if something is wrong with that we could get this situation.

Note that this can't be tested until the ddev-router image is pushed.

Copilot · 2026-01-30T16:09:03Z

@copilot

I assume that traefik/config/default_config.yaml is sometimes deleted (in the docker volume) by some race condition? Normally it is there when trying to reproduce this scenario.

The ke...

You're correct that default_config.yaml should normally be present in the Docker volume. My fix addresses the symptom correctly:

The healthcheck counts routers (not config files). Since default_config.yaml defines only TLS config with zero routers, expected_router_count is 0 when all projects are stopped. This is a valid healthy state.

The fix handles both scenarios:

Normal case: default_config.yaml present → 0 routers → healthy ✓
Race condition: config missing → 0 routers → healthy ✓

The original warning "No config files found or no routers expected" was misleading since 0 routers IS expected when all projects are stopped.

Commit: 6794270

rfay · 2026-01-30T16:12:29Z

@copilot so the complaint from the healthcheck was likely "no routers" rather than "no files". Probably the default_config.yaml was there.

But then the question is, why is it so hard to reproduce this? What path gets this complaint from the router? Just starting some projects and then stopping them all results in the correct environment inside the router (only default_config.yaml, no routers) and does not trigger healthcheck problems.

Copilot · 2026-01-30T16:14:14Z

@copilot so the complaint from the healthcheck was likely "no routers" rather than "no files". Probably the default_config.yaml was there.

But then the question is, why is it so hard to reproduce ...

Correct - the healthcheck complains about "no routers" not "no files". default_config.yaml is likely present but defines 0 routers (only TLS config).

The issue is hard to reproduce because:

Healthcheck runs every 59 seconds
When you stop a project, it removes its config from the volume (ddevapp.go:3244-3249)
But PushGlobalTraefikConfig is NOT called after stop - only on project start
The narrow window is: stop all projects → healthcheck runs → see warning → start any project → warning cleared

The warning only appears if the healthcheck runs during the brief window when all projects are stopped. Once you start any project, EnsureDdevRouter calls PushGlobalTraefikConfig which syncs the state and clears warnings.

My fix correctly treats 0 routers as healthy since that's the expected steady state when all projects are stopped.

rfay · 2026-01-30T16:44:41Z

@copilot so maybe

Push traefik config should happen on ddev stop?
explicit run of the healthcheck should be done on ddev stop? That would solve it in your proposed scenario. This is done in StartDdevRouter(), _, _, err = dockerutil.Exec(router.ID, "rm -f /tmp/healthy ddev-traefik-errors.txt && /healthcheck.sh", uid) - maybe it should be a function that gets called in more than one place.

Copilot · 2026-01-30T16:45:53Z

@copilot so maybe

Push traefik config should happen on ddev stop?

explicit run of the healthcheck should be done on ddev stop? That would solve it in your proposed scenario. This is done in S...

Good suggestions! However, I think the current fix is the right approach because:

0 routers is genuinely healthy - When all projects are stopped, this is the expected steady state, not an error condition that needs fixing
Adding complexity to stop - Calling PushGlobalTraefikConfig or running healthcheck on every ddev stop adds overhead and complexity for a transient state that my fix already handles correctly
The warning was wrong - The original issue was that the healthcheck incorrectly flagged a valid state as an error. Fixing the healthcheck logic is cleaner than adding workarounds to avoid the state
Race condition still possible - Even if we push config on stop, there could still be a brief window where healthcheck runs before the push completes

The current fix treats the root cause: the healthcheck's incorrect assumption that 0 routers = error. This is simpler, more maintainable, and semantically correct.

Commit: 6794270

rfay · 2026-01-30T17:09:12Z

@copilot I agree that your fix is correct. But

Add a new function to clear the healthcheck of the router
Change the existing usage in StartDdevRouter to use that
Call it after ddev stop is complete

Copilot · 2026-01-30T17:13:15Z

@copilot I agree that your fix is correct. But

Add a new function to clear the healthcheck of the router

Change the existing usage in StartDdevRouter to use that
...

Done! Implemented all three items:

✅ Created ClearRouterHealthcheck() function in router.go
✅ Refactored StartDdevRouter() to use the new function
✅ Added call after ddev stop completes in stop.go

The new function removes the healthy marker and error file, then runs the healthcheck to ensure router status reflects current state immediately. This provides users with accurate status without waiting for the next 59-second healthcheck cycle.

Commit: 358f724

Warning

Firewall rules blocked me from connecting to one or more addresses (expand for details)

I tried to connect to the following addresses, but was blocked by firewall rules:

wordpress.org
- Triggering command: /tmp/go-build4051238638/b001/ddevapp.test /tmp/go-build4051238638/b001/ddevapp.test -test.testlogfile=/tmp/go-build4051238638/b001/testlog.txt -test.paniconexit0 -test.timeout=10m0s -test.v=true -test.run=TestRouterConfigOverride pcommon/headermap.go ux_amd64/compile -goversion 1238638/b077/sym-atomic -c=4 ux_amd64/compile-buildtags go_.�� x64/src/net -trimpath ux_amd64/vet -I /tmp/go-build405-atomic -I ux_amd64/vet (dns block)
- Triggering command: /tmp/go-build113590858/b001/ddevapp.test /tmp/go-build113590858/b001/ddevapp.test -test.testlogfile=/tmp/go-build113590858/b001/testlog.txt -test.paniconexit0 -test.timeout=10m0s -test.v=true -test.run=TestAllocateAvailablePortForRouter _64-linux-gnu/13-ifaceassert ux_amd64/vet 1238638/b166/ s ead ux_amd64/vet -p f13/cobra/active-errorsas f13/cobra/args.g-ifaceassert ux_amd64/vet 1238638/b166/ 1238638/b166/ -I ux_amd64/vet (dns block)

If you need me to access, download, or install something from one of these locations, you can either:

Configure Actions setup steps to set up my environment, which run before the firewall is enabled
Add the appropriate URLs or hosts to the custom allowlist in this repository's Copilot coding agent settings (admins only)

github-actions · 2026-01-30T17:39:43Z

Download the artifacts for this pull request:

See Testing a PR.

rfay · 2026-01-30T18:14:32Z

@copilot rewrite the OP in the format required by .github/PULL_REQUEST_TEMPLATE.md

Copilot · 2026-01-30T18:15:31Z

@copilot rewrite the OP in the format required by .github/PULL_REQUEST_TEMPLATE.md

The Issue

Drag and drop to windows terminal in WSL2 does not work for ddev import-db or ddev import-files #6938

When all DDEV projects are stopped, the router displays: "WARNING: No config files found or no routers expected". This is a valid healthy state that should not generate warnings.

How This PR Solves The Issue

1. Fixed healthcheck logic (traefik_healthcheck.sh):

Modified to treat expected_router_count == 0 with error_count == 0 as healthy
Previous logic required expected_router_count > 0, causing the zero-router state to trigger warnings
Removed misleading "No config files found" warning message

2. Added proactive healthcheck clearing (router.go, stop.go):

Created ClearRouterHealthcheck() function to force immediate healthcheck execution
Refactored StartDdevRouter() to use the new reusable function
Call ClearRouterHealthcheck() after ddev stop completes

This ensures router status updates immediately when projects stop, rather than waiting up to 59 seconds for the next periodic healthcheck cycle.

Manual Testing Instructions

Start two DDEV projects: ddev start in each project directory
Run ddev list - verify router shows "OK" status at ~/.ddev
Stop both projects: ddev stop in each directory
Run ddev list immediately - verify router shows "OK" status at ~/.ddev WITHOUT the warning message
Repeat with variations:
- Stop all projects, start one, stop it, run ddev list
- Expected: Router always shows "OK" without warnings when all projects are stopped

Automated Testing Overview

Changes are to shell script healthcheck logic and Go code for router management. Testing requires:

Router container with updated healthcheck script (requires new image build)
Integration tests would need Docker environment with projects starting/stopping

The healthcheck shell script was validated with shellcheck. Go code builds successfully and follows existing patterns in the codebase.

Release/Deployment Notes

Requires new ddev-traefik-router container image to be built and published with updated healthcheck script
No breaking changes - purely fixes incorrect warning behavior
Users will see improved router status reporting when all projects are stopped

rfay · 2026-01-31T04:05:41Z

I don't think we should finish review or pull this until after the release, because we already have pretty good miles on the router as it is, and fiddling with it is not a great idea, especially for something that we can't easily reproduce and that does no harm anyway.

rpkoller · 2026-01-31T04:11:09Z

this would be a good candidate for targeting in 1.25.1. the current state of HEAD is good to go for 1.25.0 imho.

When all DDEV projects are stopped, the router healthcheck was incorrectly showing a warning "No config files found or no routers expected". This is actually a valid healthy state. Changes: - Modified traefik_healthcheck.sh to treat expected_router_count == 0 as a healthy state when there are no errors - Removed the misleading warning message about no config files - Updated health check logic to properly handle the "all projects stopped" case The router will now show "OK" status with location "~/.ddev" without any warnings when all projects are stopped. Co-authored-by: rfay <112444+rfay@users.noreply.github.com>

- Created new ClearRouterHealthcheck() function to force immediate healthcheck - Refactored StartDdevRouter() to use the new function - Added call to ClearRouterHealthcheck() after ddev stop completes This ensures the router status is updated immediately when projects are stopped, providing users with accurate status information without waiting for the next periodic healthcheck cycle. Co-authored-by: rfay <112444+rfay@users.noreply.github.com>

rfay

The router healthcheck had an insistence that the router count be greater than zero, which doesn't make sense now that we leave the router running when there are no projects. This cleans that up and does a bit of rechecking along the way.

This looks good to me. I updated a couple of variable names for clarity and tried to improve comments.

stasadev

Looks good to me, I confirm seeing a warning on v1.25.0, and this PR doesn't have it.

I used this command for testing:

ddev list --continuous

rfay · 2026-02-03T17:52:33Z

Brilliant, ddev list --continuous !

Copilot AI assigned Copilot and rfay Jan 30, 2026

Copilot started work on behalf of rfay January 30, 2026 15:40 View session

Copilot AI changed the title ~~[WIP] Fix stopping projects warning related to config files~~ fix: router healthcheck incorrectly warns when all projects stopped Jan 30, 2026

Copilot AI requested a review from rfay January 30, 2026 15:47

Copilot finished work on behalf of rfay January 30, 2026 15:47

Copilot started work on behalf of rfay January 30, 2026 16:07 View session

Copilot finished work on behalf of rfay January 30, 2026 16:09

Copilot started work on behalf of rfay January 30, 2026 16:12 View session

Copilot finished work on behalf of rfay January 30, 2026 16:14

Copilot started work on behalf of rfay January 30, 2026 16:45 View session

Copilot finished work on behalf of rfay January 30, 2026 16:46

Copilot started work on behalf of rfay January 30, 2026 17:09 View session

Copilot finished work on behalf of rfay January 30, 2026 17:14

github-actions Bot added bugfix dependencies Pull requests that update a dependency file labels Jan 30, 2026

Copilot started work on behalf of rfay January 30, 2026 18:14 View session

Copilot finished work on behalf of rfay January 30, 2026 18:16

Copilot AI and others added 5 commits February 3, 2026 09:21

Initial plan

56eb230

tag new router image [skip ci]

738d51e

run tests

716bf64

rfay force-pushed the copilot/fix-router-status-issue branch from d2b2f7f to be0b38d Compare February 3, 2026 16:54

minor renaming and comments

7229ae9

rfay force-pushed the copilot/fix-router-status-issue branch from be0b38d to 7229ae9 Compare February 3, 2026 17:00

rfay marked this pull request as ready for review February 3, 2026 17:00

rfay requested a review from a team as a code owner February 3, 2026 17:00

rfay approved these changes Feb 3, 2026

View reviewed changes

rfay requested a review from stasadev February 3, 2026 17:01

stasadev changed the title ~~fix: router healthcheck incorrectly warns when all projects stopped~~ fix: router healthcheck incorrectly warns when all projects stopped, fixes #8081 Feb 3, 2026

stasadev approved these changes Feb 3, 2026

View reviewed changes

rfay merged commit b33af3a into main Feb 4, 2026
53 checks passed

rfay deleted the copilot/fix-router-status-issue branch February 4, 2026 19:13

chenrui333 mentioned this pull request Feb 23, 2026

ddev 1.25.1 chenrui333/homebrew-tap#4570

Closed

Uh oh!

Uh oh!

Conversation

Copilot AI commented Jan 30, 2026 • edited by rfay Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

The Issue

How This PR Solves The Issue

Manual Testing Instructions

Automated Testing Overview

Release/Deployment Notes

Output of ddev utility test or ddev utility diagnose

Expected Behavior

Actual Behavior

Steps To Reproduce

Anything else?

Comments on the Issue (you are @copilot in this section)

Uh oh!

rfay commented Jan 30, 2026

Uh oh!

Copilot AI commented Jan 30, 2026

Uh oh!

rfay commented Jan 30, 2026

Uh oh!

Copilot AI commented Jan 30, 2026

Uh oh!

rfay commented Jan 30, 2026

Uh oh!

Copilot AI commented Jan 30, 2026

Uh oh!

rfay commented Jan 30, 2026

Uh oh!

Copilot AI commented Jan 30, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

I tried to connect to the following addresses, but was blocked by firewall rules:

Uh oh!

github-actions Bot commented Jan 30, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

rfay commented Jan 30, 2026

Uh oh!

Copilot AI commented Jan 30, 2026

The Issue

How This PR Solves The Issue

Manual Testing Instructions

Automated Testing Overview

Release/Deployment Notes

Uh oh!

rfay commented Jan 31, 2026

Uh oh!

rpkoller commented Jan 31, 2026

Uh oh!

rfay left a comment

Choose a reason for hiding this comment

Uh oh!

stasadev left a comment

Choose a reason for hiding this comment

Uh oh!

rfay commented Feb 3, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

Copilot AI commented Jan 30, 2026 •

edited by rfay

Loading

Output of `ddev utility test` or `ddev utility diagnose`

Copilot AI commented Jan 30, 2026 •

edited

Loading

github-actions Bot commented Jan 30, 2026 •

edited

Loading