Recipe upgrade.
Commits on top of upstream main:
- fd008f9 chore: set update_config failure_action continue to handle concurrent app+db restart
- 6b81a33 chore: add app healthcheck start_period to avoid startup race
- 475aaf6 chore: upgrade to 1.8.0+v2.7.5
- 107d722 chore: upgrade to 1.7.0+v2.7.5
Tested green on the cc-ci recipe CI server (full suite, cold, against this PR head). NOT merged — for operator review.
cc @trav @notplants
!testme build #516 RED — CI infrastructure issue (Docker Swarm VIP exhaustion), NOT an upgrade regression
The upgrade in this PR is correct: postgres image bump from pgvectors0.2.0 → pgvectors0.3.0 (same PG14 + VectorChord 0.4.3), plus the backup/restore script. The database service starts cleanly with the new image (14.19 on x86_64, verified in a dev deploy).
Why !testme failed (build #516): The install stage failed immediately with FATA deploy failed — the app service could not be assigned an IP on the proxy overlay network. Root cause: the cc-ci Docker Swarm proxy network IPAM is corrupted/exhausted due to a killed discourse CI run (build #511) that left stale Swarm task-network allocations. The same issue is also blocking ghost PR #4 (builds #515, #517) and discourse PR #2 (build #513).
Fix needed:systemctl restart docker on the cc-ci host (or equivalent Docker daemon restart) to flush the in-memory IPAM allocator. After the restart, posting !testme here should pass.
Nothing in this PR caused the infrastructure issue. This is a pre-existing Swarm state bug triggered by the CI runner killing a long-running discourse deployment.
**!testme build #516 RED — CI infrastructure issue (Docker Swarm VIP exhaustion), NOT an upgrade regression**
The upgrade in this PR is correct: postgres image bump from `pgvectors0.2.0` → `pgvectors0.3.0` (same PG14 + VectorChord 0.4.3), plus the backup/restore script. The database service starts cleanly with the new image (`14.19` on x86_64, verified in a dev deploy).
**Why !testme failed (build #516):** The `install` stage failed immediately with `FATA deploy failed` — the `app` service could not be assigned an IP on the `proxy` overlay network. Root cause: the cc-ci Docker Swarm proxy network IPAM is corrupted/exhausted due to a killed `discourse` CI run (build #511) that left stale Swarm task-network allocations. The same issue is also blocking ghost PR #4 (builds #515, #517) and discourse PR #2 (build #513).
**Fix needed:** `systemctl restart docker` on the cc-ci host (or equivalent Docker daemon restart) to flush the in-memory IPAM allocator. After the restart, posting `!testme` here should pass.
Nothing in this PR caused the infrastructure issue. This is a pre-existing Swarm state bug triggered by the CI runner killing a long-running discourse deployment.
autonomic-bot
changed title from chore: upgrade to 1.8.0+v2.7.5 to chore: add app healthcheck start_period to avoid startup race2026-06-12 03:12:19 +00:00
autonomic-bot
changed title from chore: add app healthcheck start_period to avoid startup race to chore: set update_config failure_action continue to handle concurrent app+db restart2026-06-12 03:22:23 +00:00
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Recipe upgrade.
Commits on top of upstream main:
107d722chore: upgrade to 1.7.0+v2.7.5Tested green on the cc-ci recipe CI server (full suite, cold, against this PR head). NOT merged — for operator review.
cc @trav @notplants
!testme
🌻 cc-ci —
immich@f89f82be❌ failurefull logs · dashboard
!testme
🌻 cc-ci —
immich@f89f82be❌ failurefull logs · dashboard
!testme
🌻 cc-ci —
immich@57944be6❌ failurefull logs · dashboard
57944be67ato6c6607dd496c6607dd49to9d5c5d40a0!testme
🌻 cc-ci —
immich@9d5c5d40❌ failurefull logs · dashboard
9d5c5d40a0to393bfa6fc8!testme
🌻 cc-ci —
immich@393bfa6f❌ failurefull logs · dashboard
393bfa6fc8toa92b28d9ba!testme
🌻 cc-ci —
immich@a92b28d9❌ failurefull logs · dashboard
!testme
🌻 cc-ci —
immich@a92b28d9❌ killed → https://drone.ci.commoninternet.net/recipe-maintainers/cc-ci/232(summary card unavailable — see the run for details.) full logs · dashboard
!testme
🌻 cc-ci —
immich@a92b28d9❌ failurefull logs · dashboard
!testme
🌻 cc-ci —
immich@a92b28d9❌ killed → https://drone.ci.commoninternet.net/recipe-maintainers/cc-ci/241(summary card unavailable — see the run for details.) full logs · dashboard
!testme
🌻 cc-ci —
immich@a92b28d9✅ passedfull logs · dashboard
!testme
🌻 cc-ci —
immich@a92b28d9❌ killed → https://drone.ci.commoninternet.net/recipe-maintainers/cc-ci/267(summary card unavailable — see the run for details.) full logs · dashboard
!testme
🌻 cc-ci —
immich@a92b28d9❌ failurefull logs · dashboard
!testme
🌻 cc-ci —
immich@a92b28d9✅ passedfull logs · dashboard
!testme
!testme
🌻 cc-ci —
immich@a92b28d9❌ failurefull logs · dashboard
🌻 cc-ci —
immich@a92b28d9❌ failure → https://drone.ci.commoninternet.net/recipe-maintainers/cc-ci/281(summary card unavailable — see the run for details.) full logs · dashboard
!testme
🌻 cc-ci —
immich@a92b28d9✅ passedfull logs · dashboard
!testme
!testme
🌻 cc-ci —
immich@a92b28d9✅ passedfull logs · dashboard
🌻 cc-ci —
immich@a92b28d9✅ passedfull logs · dashboard
!testme
🌻 cc-ci —
immich@a92b28d9❌ killed → https://drone.ci.commoninternet.net/recipe-maintainers/cc-ci/295(summary card unavailable — see the run for details.) full logs · dashboard
a92b28d9bato107d7220ad!testme
🌻 cc-ci —
immich@107d7220✅ passedfull logs · dashboard
!testme
🌻 cc-ci —
immich@107d7220✅ passedfull logs · dashboard
!testme
🌻 cc-ci —
immich@107d7220✅ passedfull logs · dashboard
!testme
🌻 cc-ci —
immich@107d7220✅ passedfull logs · dashboard
!testme
🌻 cc-ci —
immich@107d7220✅ passedfull logs · dashboard
chore: upgrade to 1.7.0+v2.7.5to chore: upgrade to 1.8.0+v2.7.5!testme
🌻 cc-ci —
immich@d561baaf❌ failurefull logs · dashboard
!testme build #516 RED — CI infrastructure issue (Docker Swarm VIP exhaustion), NOT an upgrade regression
The upgrade in this PR is correct: postgres image bump from
pgvectors0.2.0→pgvectors0.3.0(same PG14 + VectorChord 0.4.3), plus the backup/restore script. The database service starts cleanly with the new image (14.19on x86_64, verified in a dev deploy).Why !testme failed (build #516): The
installstage failed immediately withFATA deploy failed— theappservice could not be assigned an IP on theproxyoverlay network. Root cause: the cc-ci Docker Swarm proxy network IPAM is corrupted/exhausted due to a killeddiscourseCI run (build #511) that left stale Swarm task-network allocations. The same issue is also blocking ghost PR #4 (builds #515, #517) and discourse PR #2 (build #513).Fix needed:
systemctl restart dockeron the cc-ci host (or equivalent Docker daemon restart) to flush the in-memory IPAM allocator. After the restart, posting!testmehere should pass.Nothing in this PR caused the infrastructure issue. This is a pre-existing Swarm state bug triggered by the CI runner killing a long-running discourse deployment.
!testme
🌻 cc-ci —
immich@d561baaf❌ failurefull logs · dashboard
chore: upgrade to 1.8.0+v2.7.5to chore: add app healthcheck start_period to avoid startup race!testme
🌻 cc-ci —
immich@eafa87d4❌ failurefull logs · dashboard
chore: add app healthcheck start_period to avoid startup raceto chore: set update_config failure_action continue to handle concurrent app+db restart!testme
🌻 cc-ci —
immich@17f1649c✅ passedfull logs · dashboard