The following information outlines the bad block management features and functionality supported by Intel® VROC, including the Intel® VROC sub-products: Intel® VROC (VMD NVMe* RAID), Intel® VROC (SATA RAID) and Intel® VROC (Non-VMD NVMe* RAID). To learn about specific features supported by each Intel® VROC sub-product, refer to the following resources:
Intel® VROC Features | Bad block management is one of the features of the Intel® VROC family of products. To learn about other features of Intel® VROC, refer to the Intel® Virtual RAID on CPU (Intel® VROC) Technical Product Specification for Windows*. |
The Intel® VROC family of products provides 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 marking 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.
This feature is supported in the Pre-OS as well as the Intel® VROC driver for Windows*.
When Intel® VROC is performing a read operation and encounters a bad block condition on one of the member drives, it will mark the same sector on the other RAID member drive(s) as a bad block sector and reflect it in the Intel® VROC metadata.
When Intel® VROC is performing a RAID volume rebuild and encounters a full bad block table, it will stop rebuilding the RAID volume at that point and will leave the RAID volume in a degraded state.
When Intel® VROC is performing a read operation and encounters a bad block condition, that read command response will indicate that a failure has occurred.
Intel® VROC (SATA RAID) supports Bad Block Management while the system is performing Intel® VROC (SATA RAID) Read Patrol activities. The Intel® VROC (SATA RAID) Read Patrol feature is disabled by default and must be manually enabled via the Intel® VROC Graphical User Interface (GUI) application. During the Intel® VROC (SATA RAID) Read Patrol process, if a bad block is encountered, an attempt will be made to correct the bad block condition.
When the Intel® VROC encounters multiple bad blocks in the corresponding RAID volume member (corresponding Logical Block Addressing (LBA)’s), the RAID volume state will not be changed if encountered during one of the following processes:
Intel® VROC will log bad block conditions found in the Windows* Event Log.
Intel® VROC will attempt to recover from a bad block condition whenever possible. This includes when Bad Block Management is performed during Read Patrol, RAID volume initialization and Verify and Repair processes.
The Intel® VROC GUI application will report any bad block activity logged through the Intel® VROC GUI interface.