Flushing host cache data before rebuilding degraded redundant virtual disk

1. A storage management method, comprising:

receiving, by a host, a degrade signal indicating a DEGRADED state of a virtual disk associated with the host;

responsive to receiving the degrade signal, performing rebuild-flush operations comprising:

writing, to the virtual disk and to a hot spare drive (HSP) associated with the virtual disk, valid-modified data, stored in a host storage cache, associated with the virtual disk; and

writing, to the HSP, valid-unmodified data, stored in the host storage cache, associated with the virtual disk; and

rebuilding the virtual disk, said rebuilding including, for each virtual disk logical block address (LBA):

responsive to determining that the virtual disk LBA corresponds to valid cache data in the host storage cache, skipping to a next virtual disk LBA; and

responsive to determining that the virtual disk LBA does not correspond to valid cache data in the host storage cache:

regenerating user data associated with the LBA; and

storing rebuilt user data resulting from said regenerating to the HSP.

2. The method of claim 1 , further comprising:

maintaining hash table data indicating virtual disk LBAs cached in the host storage cache.

3. The method of claim 2 , wherein maintaining the hash table data includes:

maintaining a valid-unmodified hash table indicating virtual disk LBAs having corresponding valid-unmodified entries in the host storage cache; and

maintaining a valid-modified hash table indicating virtual disk LBAs having corresponding valid-modified entries in the host storage cache.

4. The method of claim 3 , wherein performing the rebuild-flush operations comprises:

locking storage cache lines associated with the virtual disk to prevent eviction; and

generating a rebuild-flush I/O request corresponding to the virtual disk LBAs, said generating including:

setting, in an I/O structure corresponding to the rebuild-flush I/O request, a rebuild-flush flag indicating whether a corresponding virtual disk LBA is a valid-unmodified LBA; and

submitting the rebuild-flush I/O request.

5. The method of claim 4 , further comprising:

processing, by a storage driver, the rebuild-flush I/O request in accordance with the flush-rebuild flag, said processing including:

writing valid-unmodified virtual disk LBAs to the HSP; and

writing valid-modified virtual disk LBAs to the virtual disk and to the HSP.

6. The method of claim 2 , further comprising:

handling, while said rebuilding is in progress, an I/O transaction targeting an I/O transaction LBA, said handling comprising:

determining whether an LBA associated with the I/O transaction maps to a completed rebuild region of the virtual disk, a current rebuild region of the virtual disk, or a pending rebuild region of the virtual disk;

responsive to determining that the I/O transaction LBA maps to a completed rebuild region, executing the I/O transaction; and

responsive to determining that the I/O transaction LBA corresponds to a current rebuild region, queuing the I/O transaction to execute when rebuilding of the current rebuild region completes.

7. The method of claim 6 , wherein handling the I/O transaction further comprises:

responsive to determining that the I/O transaction LBA corresponds to a pending rebuild region, determining whether the I/O transaction LBA matches an LBA in the hash table data;

responsive to determining that the I/O transaction LBA matches a hash table LBA, processing the I/O transaction by accessing:

hard disk drives associated with the virtual disk; and

the HSP; and

responsive to determining that the I/O transaction LBA does not match a hash table LBA, regenerating a stripe associated with the I/O transaction before processing the I/O transaction.

8. An information handling system, comprising:

a central processing unit (CPU);

a storage cache device for use in providing a host storage cache for a virtual disk;

a memory including processor-executable storage instructions, that, when executed by the CPU, cause the CPU to perform operations comprising:

receiving a degrade signal indicating a DEGRADED state of the virtual disk associated with the host storage cache;

responsive to receiving the degrade signal, performing rebuild-flush operations comprising:

writing, to the virtual disk and to a hot spare drive (HSP) associated with the virtual disk, valid-modified data stored in a host storage cache and associated with the virtual disk; and

writing valid-unmodified data associated with the virtual disk to the HSP; and

rebuilding the virtual disk, said rebuilding including, for each virtual disk logical block address (LBA):

responsive to determining that the virtual disk LBA does not correspond to valid cache data in the host storage cache:

regenerating user data associated with the LBA; and

storing rebuilt user data resulting from said regenerating to the HSP.

9. The information handling system of claim 8 , wherein the host storage cache comprises a solid state drive storage cache device and storage cache metadata and wherein the virtual disk is a RAID virtual disk implemented on a group of hard disk drives.

10. The information handling system of claim 9 , wherein the operations include:

maintaining a valid-unmodified hash table indicating virtual disk LBAs having corresponding valid-unmodified entries in the host storage cache; and

maintaining a valid-modified hash table indicating virtual disk LBAs having corresponding valid-modified entries in the host storage cache.

11. The information handling system of claim 10 , wherein performing the rebuild-flush operations comprises:

locking storage cache lines associated with the virtual disk to prevent eviction; and

generating a rebuild-flush I/O request corresponding to the virtual disk LBAs, said generating including:

setting, in an I/O structure corresponding to the rebuild-flush I/O request, a rebuild-flush flag indicating whether a corresponding virtual disk LBA is a valid-unmodified LBA; and

submitting the rebuild-flush I/O request.

12. The information handling system of claim 11 , wherein the operations include:

processing, by a storage driver, the rebuild-flush operation in accordance with the flush-rebuild flag, said processing including:

writing valid-unmodified virtual disk LBAs to the HSP; and

writing valid-modified virtual disk LBAs to the virtual disk and the HSP.

13. The information handling system of claim 12 , wherein the operations include:

handling an I/O transaction while said rebuilding is in progress, said handling comprising:

determining whether an LBA associated with the I/O transaction maps to a completed rebuild region of the virtual disk, a current rebuild region of the virtual disk, or a pending rebuild region of the virtual disk;

responsive to determining that the I/O transaction LBA maps to a completed rebuild region, executing the I/O transaction; and

responsive to determining that the I/O transaction LBA corresponds to a current rebuild region, queuing the I/O transaction to execute when rebuilding of the current rebuild region completes.

14. The information handling system of claim 13 , wherein handling the I/O transaction further comprises:

responsive to determining that the I/O transaction LBA corresponds to a pending rebuild region, determining whether the I/O transaction LBA matches an LBA in either of the hash tables;

responsive to determining that the I/O transaction LBA matches a hash table, LBA processing the I/O transaction by accessing:

hard disk drives associated with the virtual disk; and

the HSP;

responsive to determining that the I/O transaction LBA does not match a hash table LBA, regenerating a stripe associated with the I/O transaction before processing the I/O transaction.

15. An information handling platform, comprising:

a logical unit number (LUN) comprising:

a plurality of hard disk drives; and

a storage controller configured to expose a virtual disk;

an information handling system, comprising:

a central processing unit (CPU);

a storage cache device for use in providing a host storage cache for the virtual disk; and

a memory including processor-executable storage instructions, that, when executed by the CPU, cause the CPU to perform operations comprising:

receiving a degrade state signal indicating a DEGRADED state of the virtual disk;

responsive to receiving the degrade signal, performing rebuild-flush operations comprising:

writing, to the virtual disk and to a hot spare drive (HSP) associated with the virtual disk, valid-modified data, in the host storage cache, associated with the virtual disk; and

writing valid-unmodified data, in the host storage cache, associated with the virtual disk to the HSP;

rebuilding the virtual disk, said rebuilding including, for each virtual disk logical block address (LBA):

skipping to a next virtual disk LBA responsive to determining that a current virtual disk LBA is cached in the host storage cache; and

regenerating a stripe corresponding to the virtual disk LBA responsive to determining that the current virtual disk LBA is not cached in the host storage cache.

16. The information handling platform of claim 15 , wherein the storage controller is configured to expose a RAID LUN to the host.

17. The information handling platform of claim 16 , wherein the RAID LUN comprises a RAID 5 LUN including N HDDs where N is an integer greater than 2.

18. The information handling platform of claim 17 , further comprising:

maintaining a valid-unmodified hash table indicating virtual disk LBAs having corresponding valid-unmodified entries in the host storage cache;

maintaining a valid-modified hash table indicating virtual disk LBAs having corresponding valid-modified entries in the host storage cache; and

locking storage cache lines associated with the virtual disk to prevent eviction.

19. The information handling platform of claim 15 , further comprising:

generating a rebuild-flush I/O request corresponding to the virtual disk LBAs, said generating including:

setting, in an I/O structure corresponding to the rebuild-flush I/O request, a rebuild-flush flag indicating whether a corresponding virtual disk LBA is a valid-unmodified LBA; and

submitting the rebuild-flush I/O request.

20. The information handling platform of claim 19 ,

processing, by a RAID storage driver, the rebuild-flush I/O request in accordance with the flush-rebuild flag, said processing including:

writing valid-unmodified virtual disk LBAs to the HSP; and

writing valid-modified virtual disk LBAs to the virtual disk and the HSP.