Mongo GridFS Dump

Dump a GridFS database to the file system.

This is used to support incremental backups, as long as files are write-once. Dump your GridFS database to the file system and use more proven backup mechanisms to secure the data in another location.

Files are dumped in two levels of subdirectories, based on their file ID. For example, using a destination of /backups/gridfs, a fs.files document with an _id of 4f8c973fcc9b365022000005 would be created at backups/gridfs/4f/8c/4f8c973fcc9b365022000005. As such, it's recommended you dump to file system that can handle a large directory tree. XFS seems to work well for our purposes.

NOTE: To speed up file count checks, this program expects your output directory to ONLY contain dumped files. If it contains other files, the counts will never match up.

Warning

This project was created to scratch an itch, and is not of the highest quality. You are welcomed to submit pull requests for something you feel could be improved.

Usage

Dumping GridFS files to the file system:

bundle install
./bin/mongo-gridfs-dump -s mongodb://user:pass@127.0.0.1/my_db -d /backups/gridfs -p fs

To get a full list of options:

./bin/mongo-gridfs-dump --help

If you perform an integrity check (-i) with a dump operation, and the check fails, the process will exit with status 1.

Restoring files dumped to the file system back into GridFS:

bundle install
./bin/mongo-gridfs-restore -s /backups/gridfs -d mongodb://user:pass@127.0.0.1/my_db -p fs

To get a full list of options:

./bin/mongo-gridfs-restore --help

To-Do

• Restore from file system to GridFS

Contributing

• Find something that could be improved
• Fix that something
• Open a pull request