docker and duplicity multivol_snapshot

[finevine]

Hi,

I'm facing a problem to understand multivol_snapshot of dupliciy.

Trying to restore manually a db from tar.gz files, I cannot untar them because they are on multivol_snapshot.

I have 2 docker-compose using volumerize: one for preprod (rc) and one for production:

rcbackup:
  # https://github.com/blacklabelops/volumerize/tree/master/backends/AmazonS3
  # https://github.com/blacklabelops/volumerize/issues/97
    image: blacklabelops/volumerize:1.6
    container_name: RcBackup
    env_file:
      - ../../.env
    restart: always
    depends_on:
      - rcdb
      - rcweb
    volumes:
      - /var/run/docker.sock:/var/run/docker.sock  
      - /etc/timezone:/etc/timezone:ro          
      - volumerizerccache:/volumerize-cache
      - '../../db_backup:/source/db_backup:ro'
      - '../../media:/source/media:ro'
    environment:         
      - VOLUMERIZE_SOURCE=/source
      - VOLUMERIZE_CACHE=/volumerize-cache
      - VOLUMERIZE_TARGET=s3://s3.eu-west-3.amazonaws.com/${VOLUMERIZE_USER}
      - AWS_ACCESS_KEY_ID=${VOLUMERIZE_USER_ACCESS_KEY}
      - AWS_SECRET_ACCESS_KEY=${VOLUMERIZE_USER_SECRET_KEY}
      - TZ="Europe/Paris"
      - VOLUMERIZE_JOBBER_TIME=0 35 3 * * *
      # - VOLUMERIZE_JOBBER_TIME=0 1/15 * * * *
      - VOLUMERIZE_FULL_IF_OLDER_THAN=7D
      - JOB_NAME2=RemoveOldBackups
      - JOB_COMMAND2=/etc/volumerize/remove-older-than 7D --force
      - JOB_TIME2=0 40 3 * * *
    networks:
      - intern
    labels:
      - traefik.enable=false
volumes:
  rcpgdb:
  volumerizerccache:
  rcredis-data:

  prodbackup:
  # https://github.com/blacklabelops/volumerize/tree/master/backends/AmazonS3
  # https://github.com/blacklabelops/volumerize/issues/97
    image: blacklabelops/volumerize:1.6
    container_name: ProdBackup
    env_file:
      - ../../.env
    restart: always
    depends_on:
      - proddb
      - prodweb
    volumes:
      - /var/run/docker.sock:/var/run/docker.sock  
      - /etc/timezone:/etc/timezone:ro          
      - volumerize-cache:/volumerize-cache
      - '../../db_backup:/source/db_backup:ro'
      - '../../media:/source/media:ro'
    environment:         
      - VOLUMERIZE_SOURCE=/source
      - VOLUMERIZE_CACHE=/volumerize-cache
      - VOLUMERIZE_TARGET=s3://s3.eu-west-3.amazonaws.com/${VOLUMERIZE_USER}
      - AWS_ACCESS_KEY_ID=${VOLUMERIZE_USER_ACCESS_KEY}
      - AWS_SECRET_ACCESS_KEY=${VOLUMERIZE_USER_SECRET_KEY}
      - TZ="Europe/Paris"
      - VOLUMERIZE_JOBBER_TIME=0 35 3 * * *
      - VOLUMERIZE_FULL_IF_OLDER_THAN=7D
      - JOB_NAME2=RemoveOldBackups
      - JOB_COMMAND2=/etc/volumerize/remove-older-than 365D --force
      - JOB_TIME2=0 40 3 * * *
    networks:
      - intern
    labels:
      - traefik.enable=false
volumes:
  prodpgdb:
  volumerize-cache:
  prodredis-data:

First, my volumes were named the same but since I've read that they must not be shared, I've changed the one for preprod to volumerizerccache

But now on the preprod, when I run docker exec RcBackup backup I'm also facing a multivol_snapshot. I cannot understand why.

Any help? 🙏 🙏 Thank you!

[blacklabelops]

I do not understand your question. What is a 'multivol_snapshot'? What does it mean 'to face a multishot_snapshot'?

Why do you rename the volumes, because you run them on the same machine? If you run them on the same machine, why do you rename the volumes but point them to the same folder?

[finevine]

It means that when I download a duplicity-full.20220706T0000Z.vol1.difftar I have 2 folders:

• multivol_snapshot
• snapshot

Files in snapshot can be red in my Macos finder but not thoses on multivol_snapshot. What I have now understand is that I should not try to restore files manually by downloading archive. I should only use volumerize command to do so.

Yes I run the containers in the same machine. I guess I should not point them to the same folder. But I've understood that source is a temp folder. So this is not a big deal is it?

And thank you for your answer!