5 Commits

Author SHA1 Message Date
6b45c97b43 Merge pull request 'chore: upgrade to 1.7.0+v2.7.5' (#11) from upgrade-1.7.0+v2.7.5 into main
Reviewed-on: https://git.coopcloud.tech/coop-cloud/immich/pulls/11
2026-06-15 17:39:56 +00:00
17f1649c2d chore: set update_config failure_action continue to handle concurrent app+db restart
All checks were successful
cc-ci/testme cc-ci: success
2026-06-12 03:22:22 +00:00
eafa87d4d1 chore: add app healthcheck start_period to avoid startup race
Some checks failed
cc-ci/testme cc-ci: failure
2026-06-12 03:12:18 +00:00
d561baaf06 chore: upgrade to 1.8.0+v2.7.5
Some checks failed
cc-ci/testme cc-ci: failure
2026-06-12 02:39:22 +00:00
107d7220ad chore: upgrade to 1.7.0+v2.7.5
All checks were successful
cc-ci/testme cc-ci: success
Confirms immich-server at the latest v2.7.5 + holds the DB pin immich v2.7.5 ships
(14-vectorchord0.4.3-pgvectors0.2.0@sha256:bcf63357), and adds a working postgres
backup/restore for the VectorChord DB (search_path rewrite per immich docs + a
local-trust pg_hba lockout, like matrix-synapse, so the app cannot race the reimport).
2026-06-10 16:53:52 +00:00
2 changed files with 12 additions and 27 deletions

View File

@ -19,18 +19,21 @@ services:
- proxy - proxy
- backend - backend
healthcheck: healthcheck:
disable: false test: ["CMD-SHELL", "immich-healthcheck"]
start_period: 120s
depends_on: depends_on:
- redis - redis
- database - database
deploy: deploy:
update_config:
failure_action: continue
labels: labels:
- "traefik.enable=true" - "traefik.enable=true"
- "traefik.http.services.${STACK_NAME}.loadbalancer.server.port=2283" - "traefik.http.services.${STACK_NAME}.loadbalancer.server.port=2283"
- "traefik.http.routers.${STACK_NAME}.rule=Host(`${DOMAIN}`${EXTRA_DOMAINS})" - "traefik.http.routers.${STACK_NAME}.rule=Host(`${DOMAIN}`${EXTRA_DOMAINS})"
- "traefik.http.routers.${STACK_NAME}.entrypoints=web-secure" - "traefik.http.routers.${STACK_NAME}.entrypoints=web-secure"
- "traefik.http.routers.${STACK_NAME}.tls.certresolver=${LETS_ENCRYPT_ENV}" - "traefik.http.routers.${STACK_NAME}.tls.certresolver=${LETS_ENCRYPT_ENV}"
- "coop-cloud.${STACK_NAME}.version=1.7.0+v2.7.5" - "coop-cloud.${STACK_NAME}.version=1.8.0+v2.7.5"
- "backupbot.backup=${ENABLE_BACKUPS:-true}" - "backupbot.backup=${ENABLE_BACKUPS:-true}"
- "backupbot.volumes.model-cache=false" - "backupbot.volumes.model-cache=false"
- "backupbot.volumes.uploads=false" - "backupbot.volumes.uploads=false"
@ -54,7 +57,7 @@ services:
- backend - backend
database: database:
image: ghcr.io/immich-app/postgres:14-vectorchord0.4.3-pgvectors0.2.0@sha256:bcf63357191b76a916ae5eb93464d65c07511da41e3bf7a8416db519b40b1c23 image: ghcr.io/immich-app/postgres:14-vectorchord0.4.3-pgvectors0.3.0@sha256:87c050465fb969a68c7ac23e375e21f4c95cfacd0edce5fa1bc31e63b7891891
environment: environment:
POSTGRES_PASSWORD_FILE: /run/secrets/db_password POSTGRES_PASSWORD_FILE: /run/secrets/db_password
POSTGRES_USER: ${DB_USERNAME} POSTGRES_USER: ${DB_USERNAME}

View File

@ -1,29 +1,11 @@
#!/bin/bash #!/bin/bash
# Postgres backup/restore hook for the immich `database` service (VectorChord/pgvecto.rs image). # Postgres backup/restore hook for immich's VectorChord/pgvecto.rs `database` service.
# Invoked by backupbot-two via the deploy labels: # Two image-specific constraints:
# backupbot.backup.pre-hook = "/pg_backup.sh backup" # 1. Never DROP DATABASE — the bundled pgvecto.rs worker PANICs and crashes the server.
# backupbot.backup.volumes.postgres.path = "backup.sql" # So `pg_dump --clean --if-exists` does a per-object replace in the live DB on restore.
# backupbot.restore.post-hook = "/pg_backup.sh restore" # 2. Restore rewrites the dump's empty search_path back to `public, pg_catalog` so VectorChord
# # types resolve (per https://docs.immich.app/administration/backup-and-restore).
# Two image-specific constraints shape this hook, and BOTH are handled:
#
# 1. NEVER drop the database. immich's image bundles the legacy pgvecto.rs (`vectors`) extension whose
# background worker, if its database is dropped, recurses on its own IPC error until postgres aborts
# with `PANIC: ERRORDATA_STACK_SIZE exceeded` and crashes the whole server — after which immich can
# never reconnect. So we do NOT `DROP DATABASE`. Instead the backup is taken with
# `pg_dump --clean --if-exists`, so the dump itself carries a `DROP TABLE IF EXISTS` + recreate +
# copy for each object. Restoring it does a per-OBJECT replace inside the live database — stale rows
# added since the backup are removed (a true point-in-time restore, not a merge) — without ever
# dropping the database, so the pgvecto.rs worker is never destabilised. (Per-table DROP TABLE is
# fine; only DROP DATABASE triggers the PANIC.)
#
# 2. VectorChord search_path. A plain pg_dump emits
# SELECT pg_catalog.set_config('search_path', '', false);
# near the top; on reimport that empty search_path leaves the vector/vchord types + operator classes
# unresolvable. immich's documented restore (https://docs.immich.app/administration/backup-and-restore)
# rewrites that line to put `public, pg_catalog` back first — we do the same. ON_ERROR_STOP is left
# OFF so a benign hiccup on one object doesn't abort the whole import.
set -e set -e