#2121: Updating PostgreSQL 14 → 18 for Swarmica

Отредактирована: сегодня

This guide is applicable for Swarmica versions 6.1.2 and higher.

Instructions

Create a backup of the database:

cd /root/swarmica

# Full logical dump of all databases and roles inside the container
docker exec -it --env-file .env swarmica-postgres sh -c 'pg_dumpall -U $POSTGRES_USER > /backups/full_backup_pg14.sql'

# Copy the dump to the host
docker cp swarmica-postgres:/backups/full_backup_pg14.sql /root/swarmica/

Check the backup:

ls -lh /root/swarmica/full_backup_pg14.sql
grep -c "CREATE DATABASE" /root/swarmica/full_backup_pg14.sql

If the second command returned ≥ 1 — the dump contains databases, everything is fine.

Stop Swarmica services:

cd /root/swarmica
docker compose down

Delete the old data volume:

PostgreSQL 18 will not start on a volume initialized by version 14, the container will crash with an incompatibility error. Therefore, the volume must be deleted and created again.

docker volume rm swarmica_swarmica_postgres_data

Create a new empty volume:

docker volume create swarmica_swarmica_postgres_data

Edit docker-compose.yml. Open the file in a text editor and change image and volume in the postgres service as shown below:

<...>
postgres:
  image: postgres:18-bookworm                          # ← CHANGED: was postgres:14.2-bullseye
  container_name: swarmica-postgres
  shm_size: 1gb
  command: postgres -N 200 -c work_mem=8MB -c temp_buffers=16MB -c shared_buffers=1GB -c track_activity_query_size=16384
  healthcheck:
    test: ["CMD-SHELL", "pg_isready -U ${POSTGRES_USER} -d ${POSTGRES_DB}"]
    interval: 10s
    timeout: 5s
    retries: 5
    start_period: 10s
  restart: always
  volumes:
    - swarmica_postgres_data:/var/lib/postgresql:z     # ← CHANGED: was /var/lib/postgresql/data
<...>

Start only PostgreSQL 18:

docker compose up -d postgres

Wait for the healthy status:

docker compose ps
docker compose logs postgres | tail -20

The logs should contain database system is ready to accept connections. This means that initdb created a new PostgreSQL 18 cluster in the fresh volume.

Copy the database backup from step 1 into the container:

docker cp /root/swarmica/full_backup_pg14.sql swarmica-postgres:/tmp/full_backup_pg14.sql
docker exec -it swarmica-postgres ls -lh /tmp/full_backup_pg14.sql

Restore data from the backup:

docker exec -it swarmica-postgres sh -c 'psql -U $POSTGRES_USER -d postgres < /tmp/full_backup_pg14.sql'

Start the remaining Swarmica services:

cd /root/swarmica
docker compose up -d

Perform final procedures:

docker exec --env-file .env -it swarmica-django-1 python manage.py post_start

Delete the temporary dump from the container:

docker exec -it swarmica-postgres rm /tmp/full_backup_pg14.sql

Keep the dump on the host (/root/swarmica/full_backup_pg14.sql) for several days as insurance. Delete it only after complete confidence in stability.

#2121: Updating PostgreSQL 14 → 18 for Swarmica

Отредактирована: сегодня

This guide is applicable for Swarmica versions 6.1.2 and higher.

Instructions

Create a backup of the database:

cd /root/swarmica

# Full logical dump of all databases and roles inside the container
docker exec -it --env-file .env swarmica-postgres sh -c 'pg_dumpall -U $POSTGRES_USER > /backups/full_backup_pg14.sql'

# Copy the dump to the host
docker cp swarmica-postgres:/backups/full_backup_pg14.sql /root/swarmica/

Check the backup:

ls -lh /root/swarmica/full_backup_pg14.sql
grep -c "CREATE DATABASE" /root/swarmica/full_backup_pg14.sql

If the second command returned ≥ 1 — the dump contains databases, everything is fine.

Stop Swarmica services:

cd /root/swarmica
docker compose down

Delete the old data volume:

PostgreSQL 18 will not start on a volume initialized by version 14, the container will crash with an incompatibility error. Therefore, the volume must be deleted and created again.

docker volume rm swarmica_swarmica_postgres_data

Create a new empty volume:

docker volume create swarmica_swarmica_postgres_data

Edit docker-compose.yml. Open the file in a text editor and change image and volume in the postgres service as shown below:

<...>
postgres:
  image: postgres:18-bookworm                          # ← CHANGED: was postgres:14.2-bullseye
  container_name: swarmica-postgres
  shm_size: 1gb
  command: postgres -N 200 -c work_mem=8MB -c temp_buffers=16MB -c shared_buffers=1GB -c track_activity_query_size=16384
  healthcheck:
    test: ["CMD-SHELL", "pg_isready -U ${POSTGRES_USER} -d ${POSTGRES_DB}"]
    interval: 10s
    timeout: 5s
    retries: 5
    start_period: 10s
  restart: always
  volumes:
    - swarmica_postgres_data:/var/lib/postgresql:z     # ← CHANGED: was /var/lib/postgresql/data
<...>

Start only PostgreSQL 18:

docker compose up -d postgres

Wait for the healthy status:

docker compose ps
docker compose logs postgres | tail -20

The logs should contain database system is ready to accept connections. This means that initdb created a new PostgreSQL 18 cluster in the fresh volume.

Copy the database backup from step 1 into the container:

docker cp /root/swarmica/full_backup_pg14.sql swarmica-postgres:/tmp/full_backup_pg14.sql
docker exec -it swarmica-postgres ls -lh /tmp/full_backup_pg14.sql

Restore data from the backup:

docker exec -it swarmica-postgres sh -c 'psql -U $POSTGRES_USER -d postgres < /tmp/full_backup_pg14.sql'

Start the remaining Swarmica services:

cd /root/swarmica
docker compose up -d

Perform final procedures:

docker exec --env-file .env -it swarmica-django-1 python manage.py post_start

Delete the temporary dump from the container:

docker exec -it swarmica-postgres rm /tmp/full_backup_pg14.sql

Keep the dump on the host (/root/swarmica/full_backup_pg14.sql) for several days as insurance. Delete it only after complete confidence in stability.