HP NetRAID-4M Configuration and Upgrade Guide (Release 5)

Chapter 9 Issues and Problem Resolution

controllers and associated containers into a single file. For example, if there are two

controllers, then two commands must be issued, forcing the configuration of each

controller to be saved into a separate file.

Workaround: Use multiple commands to make multiple files for saving the multi-

controller configuration. All of the separate files must be saved on the same diskette.

25 Lack of a warning when multiple containers are sharing the same

spare

Problem: When using the CCU to create another container, drives previously

assigned as failover drives appear on the list of available drives for new containers.

You can create a container with those drives even though they are supposed to be

failover drives.

Background: The NetRAID-4M works on the concept of “free space” rather than

dedicated spare disks. The FAST utility allows you to create a container with a

previously assigned failover drive; however, it does post a warning to the user. The

warning message is: “One or more of the disks selected for use is assigned as a spare

disk to one or more containers. Using this space may result in loss of rebuild

capability for one or more containers. Are you sure you wish to create this

container?” One must think beyond the traditional definition of RAID and hot

spares. With conventional array solutions, spare disks are assigned on a one-to-one

basis for each redundant array. The NetRAID-4M’s controller management

software's ability to view and manage storage as a pool of free space allows you to

assign spare disks on a one-to-many basis. Because each spare disk is not dedicated

to a single fault-tolerant container the same spare space can be shared by multiple

containers. This powerful method of flexible spare assignment can significantly

reduce the number of dedicated spare disks needed to protect against disk failures.

If the NetRAID-4M detects an unrecoverable error during I/O on a redundant

container, it checks for an available spare disk that has sufficient free space to handle

the rebuild, and starts the rebuild process.

Workaround: Keep track of which drives are assigned to containers and which are

assigned as failovers. A failover space not currently used for storage is available to

be assigned as failover space for another container or even as a container itself.

26 Order in which failover drives are listed