SIGTERM kills db-sync mid-pull; proposal: finish the current connection and exit between cycles

Open
#193 0 comments 0 reactions 0 assignees View on GitHub

@NicklasBlidmo is already working on this.

Since Oct 26, 2026.

  • #194 by @NicklasBlidmo — open

Assessment

Difficulty
3/5
Estimated time
1-2 days
Newbie friendliness
52/100
Issue type
Feature
Clarity
Mostly clear
Activity status
Active
Tech stack
docker, postgresql, python
Domain
backend, databases

Research direction

Start by reading dbsync_daemon.py and tracing dbsync_pull(), dbsync_push(), and run_daemon() to understand the cycle boundaries described in the issue. Implement the proposed SIGTERM shutdown behavior, then verify that a started pull or push can finish before the process exits and that SIGINT retains its default behavior.

Written by the indexing model from the issue text.

Description

dbsync_daemon.py installs no SIGTERM handler, so a docker stop,
docker compose up -d after a rebuild, or a service manager restart ends
the process wherever it happens to be. If that is inside pull() between
"GeoPackage updated" (the checkout is at the new version) and "base
schema updated" (the changeset has been applied to the database), the
checkout and the base schema are at the same version with different
content, and the next start refuses with

The db schemas already exist but 'base' schema is not synchronized with source GPKG. Running `dbsync_deamon.py` with `--force-init` should fix the issue.

--force-init then drops and rebuilds the schemas, which loses any
not-yet-pushed database edits and any views or grants on the schemas.
We hit this in production with four projects and a routine image
rebuild (one daemon was pulling 48 files at that moment).

Proposal, implemented in the PR that follows: a SIGTERM handler that only
sets a flag; dbsync_pull()/dbsync_push() check it before each
connection; run_daemon() checks it at the top of the cycle and between
pull and push; the sleep polls it once a second. A pull or push that has
started always completes; the process exits 0 at the next boundary.
SIGINT keeps its default so ctrl+c still kills immediately. A container
needs stop_grace_period long enough for one connection's pull
(we use 150 s).

Dominant language
Python
Stars
55
Forks
25
Avg merge
3d 22h
Merged PRs (30d)
1

Getting set up

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

More from MerginMaps/db-sync

All issues in MerginMaps/db-sync

Similar issues

More Python issues

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.