1. Administrator's Guide
  2. Administer
  3. Backup and Recovery
  4. Repairing a Failed Zone by Replacing Hardware

Repairing a Failed Zone by Replacing Hardware

If all of the machines belonging to a zone fail, and quorum is maintained, you can replace them by using new, different Storage Nodes deployed to the same zone.

If a zone fails but quorum is lost, you can perform a failover instead. To do this, see Performing a Failover.

For example, suppose a store consists of three zones; zn1, deployed to the machines on the first floor of a physical data center, zn2, deployed to the machines on the second floor, and zn3, deployed to the third floor. Additionally, suppose that a fire destroyed all of the machines on the second floor, resulting in the failure of all of the associated Storage Nodes. In this case, you need to replace the machines in the zn2 zone; which can be accomplished by doing the following:

  1. Replace each individual Storage Node in the failed zone with new, different Storage Nodes belonging to same zone (zn2), although located in a new physical location. To do this, follow the instructions in Replacing a Failed Storage Node. Make sure to remove each old Storage Node after performing the replacement.

  2. After replacing and then removing each of the targeted SNs, the zone to which those SNs belonged should now contain the new SNs.