¶ Overview of Usage and Definitions
This section clarifies common terms and outlines Zelta's key features.
The simplest forms of the Zelta commands look like this:
Compare two datasets:
zelta match source/endpoint target/endpoint
Make a new or incremental backup:
zelta backup source/endpoint target/endpoint
Execute a series of backup jobs:
zelta policy
For point-of-use assistance, try zelta usage to show a quick list of subcommands and options, or zelta help to view the manual.
¶ Command Structure
Endpoint Format: [username@][hostname:]pool[/dataset][@snapshot]
- Local Dataset Example:
tank/data
- Remote Dataset Example 1:
server10.biz:tank/files
- Remote Dataset Example 2:
backup@server11.biz:pool/stuff
- Remote Snapshot Example:
twin@server12.biz:sink/log@today
Understanding the terminology used in Zelta's documentation and the broader context of ZFS and backup strategies will be helpful for geting the most out the Zelta Replication Suite. Though these terms can be used in different ways in different contexts, their meanings in this documentation is clarified below.
- Archives: Static replicas that don't need to be updated incrementally such as the backups of offline datasets or unused clone origins.
- Backups: Replicas that are expected to receive ongoing replication jobs.
- Dataset: A ZFS volume or filesystem (in Zelta, this won't refer to a snapshot unless specified).
- Endpoint: A dataset or snapshot object definition including its username, hostname, and pool, similar to the format used by SCP.
- Initiator: The host that initiates Zelta subcommands, which is useful for orchestrating replication between two remote hosts.
- Match: The most common snapshots (or source bookmark/target snapshot) between two replicas; if there is a match, an incremental replication is possible.
- Rel_Name: The relative child dataset name when comparing two replicas, similar to a relative directory path, .e.g., if replicating
A to B, a rel_name of x/y would refer to both A/x/y and B/x/y. A blank rel_name indicates the top of the pair of datasets. See zelta-match(8).
- Replica: A copy of a dataset tree used for a backup, archive, or failover standby.
- Rotate: Rename a dataset, and then clone it to the original name, which is a safe alternative to
zfs rollback.
- Savepoint: Zelta's internal name for a bookmark or snapshot.
- Source/Target: A dataset tree and its replica.
- Tree: A dataset and its children.
In addition, you should be aware of the following terms.
- General ZFS terms: Including bookmark, clone, filesystem, permissions, pool, properties, volume, replication, rollback, snapshot.
- General disaster recovery terms: Understanding redundancy versus backups, incremental vs full backups, RPO, RTO.
Zelta has a design that is oriented towards safe, recursive, and remote operation:
- All operations are recursive from the top of the indicated dataset.
- All operations can work on local or remote datasets.
- You never need to run Zelta as root. See "Setting Up Non-Root Replication" for more information.
For more examples, see Quick Start.