starred/headscale
1
0
Fork 0
mirror of https://github.com/juanfont/headscale.git synced 2026-01-11 20:00:28 +01:00
Code Issues 98 Packages Projects Releases 10 Wiki Activity

Verify all database connections are closed when process is killed #633

New Issue
Closed
opened 2025-12-29 02:21:24 +01:00 by adam · 6 comments
adam commented 2025-12-29 02:21:24 +01:00
Owner
Copy Link

Originally created by @TotoTheDragon on GitHub (Feb 13, 2024).

Why

SQLite sometimes has trouble with locks. This often occurs when there was a connection being used while the process is killed.

Description

Verify that on all ways of exiting an application the database connections are closed properly and implement where lacking

Originally created by @TotoTheDragon on GitHub (Feb 13, 2024). <!-- We typically have a clear roadmap for what we want to improve and reserve the right to close feature requests that does not fit in the roadmap, or fit with the scope of the project, or we actually want to implement ourselves. Headscale is a multinational community across the globe. Our language is English. All bug reports needs to be in English. --> ## Why <!-- Include the reason, why you would need the feature. E.g. what problem does it solve? Or which workflow is currently frustrating and will be improved by this? --> SQLite sometimes has trouble with locks. This often occurs when there was a connection being used while the process is killed. ## Description <!-- A clear and precise description of what new or changed feature you want. --> Verify that on all ways of exiting an application the database connections are closed properly and implement where lacking
adam added the enhancement label 2025-12-29 02:21:24 +01:00
adam closed this issue 2025-12-29 02:21:24 +01:00
adam commented 2025-12-29 02:21:25 +01:00
Author
Owner
Copy Link

@ohdearaugustin commented on GitHub (Feb 13, 2024):

This issue can be tested. Please add more information like which version on which system. Installed with docker or without. etc.

With which configuration can this issue be recreated.

@ohdearaugustin commented on GitHub (Feb 13, 2024): This issue can be tested. Please add more information like which version on which system. Installed with docker or without. etc. With which configuration can this issue be recreated.
adam commented 2025-12-29 02:21:25 +01:00
Author
Owner
Copy Link

@RainrZufall commented on GitHub (Feb 14, 2024):

This issue can be tested. Please add more information like which version on which system. Installed with docker or without. etc.

With which configuration can this issue be recreated.

I have the same problem.
I use Ubuntu, Docker and the current stable release image.

It always occurs when, for example, I have to restart the server because of a kernel update or Docker is being updated.
I then have to manually restart the container again using “docker restart headscale” so that it works again.

@RainrZufall commented on GitHub (Feb 14, 2024): > This issue can be tested. Please add more information like which version on which system. Installed with docker or without. etc. > > With which configuration can this issue be recreated. I have the same problem. I use Ubuntu, Docker and the current stable release image. It always occurs when, for example, I have to restart the server because of a kernel update or Docker is being updated. I then have to manually restart the container again using “docker restart headscale” so that it works again.
adam commented 2025-12-29 02:21:25 +01:00
Author
Owner
Copy Link

@kradalby commented on GitHub (Feb 15, 2024):

This is a fair concern, we should check the shutdown logic.

@kradalby commented on GitHub (Feb 15, 2024): This is a fair concern, we should check the shutdown logic.
adam commented 2025-12-29 02:21:25 +01:00
Author
Owner
Copy Link

@kradalby commented on GitHub (Feb 15, 2024):

So I've looked into this a bit, and SIGKILL (signal 9) cannot be trapped in Go, I think that is the main "missing" signal that we could receive that isnt handled, but there is no way to also do it. I'm not sure if there are any other situation within the code as of now.

@kradalby commented on GitHub (Feb 15, 2024): So I've looked into this a bit, and SIGKILL (signal 9) cannot be trapped in Go, I think that is the main "missing" signal that we could receive that isnt handled, but there is no way to also do it. I'm not sure if there are any other situation within the code as of now.
adam commented 2025-12-29 02:21:25 +01:00
Author
Owner
Copy Link

@TotoTheDragon commented on GitHub (Feb 20, 2024):

Seems like this issue only occurs when containers are killed and restart within the sqlite lock timeout. Cannot do anything to fix this, so I propose we close this @kradalby

@TotoTheDragon commented on GitHub (Feb 20, 2024): Seems like this issue only occurs when containers are killed and restart within the sqlite lock timeout. Cannot do anything to fix this, so I propose we close this @kradalby
adam commented 2025-12-29 02:21:26 +01:00
Author
Owner
Copy Link

@kradalby commented on GitHub (Feb 20, 2024):

Ok, let's fix the lock instead

@kradalby commented on GitHub (Feb 20, 2024): Ok, let's fix the lock instead
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
CLI
DERP
DNS
Nix
OIDC
SSH
bug
database
documentation
duplicate
enhancement
faq
good first issue
grants
help wanted
might-come
needs design doc
needs investigation
no-stale-bot
out of scope
performance
policy 📝
pull-request
Mirrored from GitHub Pull Request
 question
regression
routes
stale
tags
tailscale-feature-gap
well described ❤️
wontfix
No Label enhancement
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
adam
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: starred/headscale#633
Delete Branch "%!s()"

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?