Intel S2600CP Software User Guide for Windows* - Page 19

RAID Features 2.5.19 Bad Block Management Intel® RSTe will provide support for Bad Block Management. In the course of rebuilding a degraded RAID volume, where one of the member disks has failed or been removed, and is being replaced by a 'spare' drive, the redundant contents of the other drive(s) are read and then used to reconstruct data to be written to the spare drive. In case a read failure occurs sometime during this rebuild process, the data to be written to the spare will not be available and therefore lost. In this scenario, rather than mark the entire RAID volume as failed, we can mark only those sectors on the spare that are known to have indeterminate data, in a log of such bad sectors. This bad block management log can be used to reflect error status whenever any attempts are made to access those sectors of the spare. 2.5.20 Dirty Stripe Journaling Intel® RSTe will provide support for Dirty Stripe Journaling (DSJ). DSJ is used to help speed up RAID 5 write power loss recovery by storing the write stripes that were in progress at the time of the failure. The DSJ allows rapid recovery without having to rebuild the entire volume. The DSJ is only utilized when disk write cache is DISABLED. 2.5.21 Partial Parity Logging (PPL) Intel® RSTe will provide support for Partial Parity Logging (PPL). PPL is used to record the results of XORing old data with old parity. PPL is currently saved as part of the RAID member information and is only utilized when writing RAID 5 parity. It helps protect against data loss when a power failure or a system crash occurs by allowing data to be rebuilt by utilizing the PPL information. 2.5.22 OS Installation Intel® RSTe will provide the OS appropriate RSTe driver files required for installation during the OS setup onto a drive or RAID volume attached to either the AHCI or SCU controllers. 2.5.23 Selectable Boot Volume Intel® RST 3.0 will support the ability to select any volume as the OS boot volume. The OS installer will be able to install the operating system onto RAID volume. There will be no need for RAID management (for example, volume creation/deletion) support from within OS installer. 2.5.24 Auto Rebuild Intel® RSTe will provide support for the ability to automatically rebuild a failed or degraded RAID volume. This feature will begin when a member disk of the array has failed and a suitable replacement disk with sufficient capacity is available. As soon as the failure occurs the rebuild process will begin automatically, using the marked Hot Spare disk, without user intervention. If a marked Hot Spare disk is not present, the automatic rebuild process will begin under the following conditions:  Another free disk is plugged into the same directly attached physical location as the failed drive  The newly inserted disk size is at least as large as the amount of space used per disk in the current array  The newly inserted disk must be the same type (SAS/SATA) as the disk being replaced or the rebuild will not start.  If the newly inserted disk contains Intel® RSTe (or Intel® RST) metadata with current status of member being offline or contains no Intel® RSTe (or Intel® RST) metadata.  The newly inserted disk has not reported a SMART event. 10 Intel® RSTe User's Guide