Rack and chassis failures primarily include the following:
- Rack failure: This includes network disconnection or power failure of the rack itself, as well as the malfunction of one or more servers within the rack.
- Chassis failure: This includes disconnection or power loss of the chassis itself, as well as faults in one or more hosts within the chassis.
- Host failure: This includes scenarios where the host itself shuts down, restarts, or experiences network issues. It also covers situations where failure of one or more data disks on the host renders it unavailable.
Note:
The types of tolerable faults described below only apply to the security of the data itself. Data availability must also consider whether the fault causes more than half of the master nodes in the cluster to fail.
