dardriveBasic concepts¶
File types¶
Dardrive stores backups and catalogs as .dar files, in the archive_store and catalog_store respectively. The first slice for each individual catalog will be saved with extended attributes on which dardrive represents the metadata needed to import an archive store on another system.
Jobs¶
Backup Jobs are definitions within the main config file, which define the storage and backup paths, and a series of properties such as compression, encryption, rotation, etc.
Backup types¶
Dardrive may produce 4 different kinds of backups for a given job:
Full filesystem backup, are dar backups with root path specified with the root option, all the files are stored in one or more slices of the same dar base. They are produced either forcing a full backup or when the last full is olther than the maximum age in days allowed for a job (specified by the diffdays option).
Incremental filesystem backups are dar archives which saves only files that have changed since reference full backup. Dardrive uses as a reference a catalog file corresponding with the last full backup for a given job.
Database dumps are simple mysqldump files of a given database or a collection of databases in SQL format.
Gzipped database dumps are gzipped versions of database dumps.
Additionally, all kinds of backups can be encrypted, file system ones use dar’s internal encryption, database dumps use openssl symmetric encryption.
When dardrive runs a filesystem backup several actions take place:
1. It chooses whether to run an incremental or full backup.2. It creates an entry on its internal database.3. Runs the dar backup and builds parity files if configured, also if the par_local option is.5. Save metadata as extended attributes of the first slice of the archive.7. Appends the logs from previous operations to the database.8. Adds the catalog from the backup to a dar_manager database.9. Run the rotation algorithm, which may modify the database, delete old archives, and update extended attributes of previous archives.