Redundant

Redundancy is a widely-used approach for making things robust (i.e. resistant to failure or attack); it involves having more than one of a thing (redundant copies). This can be an expensive approach to achieving that property (e.g. a bridge can be made robust against failure by providing a second complete, parallel, bridge), but it is fairly simple to do, and will work through almost all kinds of failure of the initial copy.

In data storage (both primary, and secondary storage), it is done by retaining additional information, stored with the basic data. If any of the data is corrupted, the redundant information can be used to recover from the loss. Examples there include:

• LINC tape implements straightforward redundancy by providing duplicates of all the tracks on the tape;
• RAM targeted for reliable services is protected by error-correcting codes;
• RAID can mirror disks, or add ECC;
• A server can have a second copy ready to take over if the first fails. This is called failover.

In data networking, examples include:

• retransmission, used in reliable byte stream protocols, and elsewhere;
• the Domain Name System makes heavy use of redundancy.