Guidelines for RAID Recovery

When faced with a RAID failure necessitating recovery, the initial step is to secure details about the original array configuration. This involves documenting both the hardware and logical RAID configurations.

For the hardware configuration, it is essential to label member disks, controller ports, and cables with a marker. This labeling facilitates easy reference to specific disks during the recovery process.

The logical RAID configuration, crucial for recovery, encompasses parameters such as RAID type, number of member disks, disk order, block size (if applicable), and the start offset for data on member disks (if provided by the RAID controller).

This configuration information can usually be found in the RAID controller BIOS or RAID monitoring and configuration software. Once all these parameters are identified, the RAID recovery is considered complete. The subsequent information pertains to the preparation for RAID recovery, which involves reconstructing missing RAID configuration details like block size, parity position, disk and data order, among others.

It's imperative to document any intermediate configurations if changes are made during the recovery process, such as swapping disks or cables or experimenting with different block sizes. This documentation ensures clarity if one of the intermediate configurations proves successful.

In preparation for member disk recovery, it is recommended to disconnect disks from the controller. Attempting RAID recovery with incorrect parameters may result in conflicting data ordering schemes. RAID recovery software cannot analyze multiple schemes simultaneously. Additionally, avoid attempting to rebuild an array using the same or a different controller without accurate RAID parameters, as it can lead to data loss or irreparable damage.

Creating disk image files from separate disks, rather than the entire array, is advised. This is crucial for certain RAID types, such as RAID 10 (mirrors) and RAID 5 (parity), where some components are inaccessible without disassembling the array.

The environment for RAID recovery poses challenges related to array sizes and data capacity. As array sizes increase, so do the required number of ports and recovery time. Adequate free disk space is essential, ranging from the size of the original disk images to the destriped image of the array and the recovered files. For example, recovering a RAID 5 array of 12 TB may necessitate three times that amount of disk space. Careful consideration of a configuration with sufficient resources, both in terms of capacity and speed, is crucial for successful RAID recovery.