Backup MongoDB to anywhere

Use a container to back up MongoDB in a practical way

docker pull atilaloise/mongodb-backup

Features:

• dump and restore
• dump For local volumes, windows shares or Bucket s3
• Simply enter the database, username and password
• It connects to any MOngodb available via network, it can be dockerizado or not
• Choose how often DUMPS occur
• Select the schedule that the dump should happen, or leave this default value as the time the container went up

To run a backup, run the image as a container with the correct parameters. Everything is controlled by environment variables passed pro container.

Por exemplo:

docker run -d --restart=always -e DB_DUMP_FREQ=60 -e DB_DUMP_BEGIN=2330 -e DB_DUMP_TARGET=/db -e MONGO_URI=mongodb://db:27017/database -v /local/file/path:/db atilaloise/mongodb-backup

The above example dumps the uri baseevery 60 minutes, starting at 11:30 pm local time

  MongoDB-backup:
    image: atilaloise/mongodb-backup
    environment:
      - MONGO_URI=mongodb://db:27017/database
      - DB_DUMP_TARGET=s3://dumps/MongoDBhomolog
      - DB_DUMP_BEGIN=1205
      - AWS_ACCESS_KEY_ID=teste
      - AWS_SECRET_ACCESS_KEY=aeuiaehuep4ssw0rd
      - AWS_ENDPOINT_URL=https://minio.lab.test
      - COMPRESSION=bzip2
    volumes:
      - /etc/localtime:/etc/localtime:ro
    networks:
      - app

• MONGO_URI: Connection string with mongoDB. MANDATORY.
• DB_DUMP_FREQ: Frequency of dump, in minutes. Default 1440 minutes = once per day.
• DB_DUMP_BEGIN: Time to do Dumps. Default Immediately while rotating the container. Define using the syntax:
  • Absolute: HHMM. for example at 23:30 use 2330 or at 04:15 use 0415
  • Relative: + MM. How many minutes after the container is started. Example + 0 (immediately), + 10 (in 10 minutes), or + 90 in an hour and a half
• DB_DUMP_CRON: Use crontab syntax to set scheduling (https://en.wikipedia.org/wiki/Cron), UNIQUE LINE.
• RUN_ONCE: Kills the container after running the backup. Useful if you use an external scheduler like rundeck. when using this variable, all other scheduling variables like DB_DUMP_FREQ, DB_DUMP_BEGIN and DB_DUMP_CRON will be ignored.
• DB_DUMP_DEBUG: Set to true, to show the full output of the scripts
• DB_DUMP_TARGET: Destination of the dumps. It needs to be a directory. If you want to send to multiple destinations, inform the targets by separating them with spaces. Supports 3 target formats:     * Local: If the value of DB_DUMP_TARGET starts with a/, the dump goes to a local destination, which must be a volume mounted inside the container.     * SMB: If the value of DB_DUMP_TARGET is an address insmb: // hostname / share / path /format then SMB will be used.     * S3: If the value of DB_DUMP_TARGET is an address in s3: // bucketname / path format then awscli will be used.
• AWS_ACCESS_KEY_ID: Access key of S3
• AWS_SECRET_ACCESS_KEY: Secret key of S3
• AWS_DEFAULT_REGION: Region. Do not set anything when using minio as destination
• AWS_ENDPOINT_URL: address of S3 Server. ex: https://minio.lab.test
• SMB_USER: SMB user. Use when DB_DUMP_TARGET is an addresssmb: //. You can leave this variable blank and pass it all in the smb: // url.
• SMB_PASS: PasswordSMB. Use when DB_DUMP_TARGET is an addresssmb: //. You can leave this variable blank and pass it all in the smb: // url.
• COMPRESSION: Compress backup. They are supported: gzip (default) and bzip2
• DB_DUMP_KEEP_PERMISSIONS: holds dump file permissions during copying using cp -a. If the target filesystem does not work properly with this, you can disable it by setting DB_DUMP_KEEP_PERMISSIONS = false. Default = true.

You can schedule backups in 3 ways

• RUN_ONCE: Rotate once and exit.
• DB_DUMP_FREQ and DB_DUMP_BEGIN: Set the time interval for the backup to occur (in minutes) and the backup start time;
• DB_DUMP_CRON: Run according to a crontab schedule.

If a backup scheduled with cron exceeds the time to start next, then this second backup will be ignored. For example, if you set up an hourly backup as follows:

0 * * * *

If the backup that started at 1:00 p.m. ends at 2:05 p.m., then the next backup will only happen at 3:00 p.m. This behavior is dictated by the Cron algorithm

If you set the 3 scheduling options, they will have the following precedence over each other.

1. RUN_ONCE runs once and exits, everything else is ignored.
2. DB_DUMP_CRON: runs according to the cron schedule set and ignores DB_DUMP_FREQ and DB_DUMP_BEGIN.
3. DB_DUMP_FREQ and DB_DUMP_BEGIN: If the other options were not set

By default, the process does not run as root. The default user of this image is the appuser with UID / GID 1005. You may not be able to access some volumes whose permissions only give administrators access.

In these cases, you have two options:

• Rotate the container as root, docker run --user 0 ... or, in docker - compose.yml, set user0
• Give write permission to the UID or GID 1005.

To do the dump, backup-mongodb needs to connect to the database via the network. You ** are required ** to inform the pgsql server - which can be the pgsql container name inside the stack or standalone docker, or any hostname accessible from the backups container - via the MONGO_URI variable, using the hostname or the server ip.

docker run -d --restart=always -e MONGO_URI=mongodb://db:27017/database -e DB_DUMP_FREQ=60 -e DB_DUMP_BEGIN=2330 -e DB_DUMP_TARGET=/db  -v /local/file/path:/db atilaloise/mongodb-backup

This is where you want the backup to be saved. The file will always be compressed and will follow the format of the following name:

db_backup_YYYYMMDDHHmm.<compression>

As:

• YYYY = Year "4 digits"
• MM = Month range 01-12
• DD = date range 01-31
• HH = hour range 00-23
• mm = minute range 00-59
• compression = appropriate extension for the chosen compression type gz (gzip; bz2 (bzip2)

By default, the backup directory is / backup inside the container. Ensure that you have mounted some external volume, after all it is no use to backup the die with the container.

If you use a smb: // host / share / path url, the backup will be sent to an SMB server. if you need to pass credentials, use smb: // user: pass @ host / share / path.

Note that for smb, if the user belongs to a domain, you should use ';' as a separator between user; domain to pass correctly in the url.

So, smb://mydom;myuser:pass@host/share/path

If the url is of type s3: // bucket / path, the backup will be saved in a S3 bucket.

Credentials must be passed in AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY and AWS_DEFAULT_REGION variables when using amazon

For interoperable systems such as MInio, digitalOcean, and other "S3 like" servers, enter the http address through the variable AWS_ENDPOINT_URL for example https: // minio.lab.test and set the DB_DUMP_TARGET to s3://bucketname/path.

To restore a backup you can also use this same image.

Just use the following environment variables:

• MONGO_URI: Connection String . MANDATORY.
• DB_RESTORE_TARGET: Backup path that will be restored, either local or SMB or S3
• DB_DUMP_DEBUG: Enables the debug of the output of the scripts and shows in the container's logs.
• To use S3, you must set AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY and AWS_DEFAULT_REGION

Examples:

1. Restore from a local file: docker run -e MONGO_URI=mongodb://db:27017/database -e DB_RESTORE_TARGET=/backup/db_backup_201509271627.gz -v /local/path:/backup atilaloise/mongodb-backup
2. Restore from SMB share: docker run -e MONGO_URI=mongodb://db:27017/database -e DB_RESTORE_TARGET=smb://smbserver/share1/backup/db_backup_201509271627.gz atilaloise/mongodb-backup
3. Restore from S3: docker run -e MONGO_URI=mongodb://db:27017/database -e AWS_ACCESS_KEY_ID=awskeyid -e AWS_SECRET_ACCESS_KEY=secret -e AWS_DEFAULT_REGION=eu-central-1 -e DB_RESTORE_TARGET=s3://bucket/path/db_backup_201509271627.gz atilaloise/mongodb-backup