Enhanced redundant caching for shingled magnetic recording devices in data storage drive assemblies

Patent metadata

Field	Value
Publication number	US-10001932-B2
Application number	US-201514734574-A
Country	US
Kind code	B2
Filing date	Jun 9, 2015
Priority date	Jun 9, 2015
Publication date	Jun 19, 2018
Grant date	Jun 19, 2018

Abstract

A data storage array is presented that includes a plurality of storage drives each comprising storage media with a first storage region and a shingled magnetic recording (SMR) storage region. The data storage array includes a control system communicatively coupled to the storage drives and configured to receive a write operation over a host interface for storage of write data by the data storage array. Responsive to the write operation, the control system is configured to cache the write data in first storage regions of two or more of the storage drives prior to transfer of the write data into at least an SMR storage region of a storage drive associated with the write operation.

First claim

What is claimed is:

1. A data storage array, comprising:
a plurality of storage drives each comprising storage media having a first storage region and a shingled magnetic recording (SMR) storage region; and
a control system communicatively coupled to the plurality of storage drives and configured to:
establish one or more cache groups among the plurality of storage drives comprising redundant sets of the storage drives;
receive write operations over a host interface for storage of write data by the data storage array, wherein the write operations are directed to storage locations associated with at least an SMR storage region of a target storage drive;
for each of the write operations, select a least-busy cache group from among the one or more cache groups not presently servicing storage operations;
cache the write data in associated first storage regions of redundant storage drives in the least-busy cache group prior to commit of the write data into at least the SMR storage region of the target storage drive;
responsive to a cache commit event, perform a sequential burst transfer of the write data from selected first storage regions of the one or more cache groups into at least the SMR storage region of the target storage drive;
detect errors during the sequential burst transfer from the selected first storage regions; and
responsively select redundant first storage regions to retrieve the write data for transfer into at least the SMR storage region.

2. The data storage array of claim 1 , wherein the control system is further configured to:
monitor activity levels among the one or more cache groups; and
determine the least-busy cache group for each of the write operations based at least in part on the activity levels indicating at least one of the one or more cache groups having associated storage drives not presently servicing read operations or write operations.

3. The data storage array of claim 1 , wherein the control system is further configured to establish data redundancy for caching the write data across the first storage regions of each selected least-busy cache group by at least forming a redundant array of independent disks (RAID) arrangement among the first storage regions of each selected least-busy cache group.

4. The data storage array of claim 1 , wherein the control system is further configured to:
detect the errors during the sequential burst transfer by at least identifying that at least a portion of the write data cached in a first one of the first storage regions of an associated cache group has experienced an error; and
responsively use a second one of the first storage regions of the associated cache group to retrieve the write data during the sequential burst transfer into the SMR storage region.

5. The data storage array of claim 1 , wherein the control system is further configured to:
detect the errors during the sequential burst transfer by at least identifying that a first one of the storage drives of an associated cache group has experienced a failure; and
responsively use a second one of the storage drives of the associated cache group to retrieve the write data during the sequential burst transfer into the SMR storage region.

6. The data storage array of claim 1 , wherein:
the write operations are directed to the storage locations that correspond to at least the target storage drive; and
the selected least-busy cache groups comprise storage drives that do not comprise the target storage drive.

7. The data storage array of claim 1 , wherein:
the first storage regions each comprise random read-write regions; and
the SMR storage regions each comprise sequential write regions.

8. A method of operating a data storage array, the method comprising:
storing data in a plurality of storage drives each comprising storage media having a first storage region and a shingled magnetic recording (SMR) storage region;
establishing one or more cache groups among the plurality of storage drives comprising redundant sets of the storage drives;
receiving write operations over a host interface for storage of write data by the data storage array, wherein the write operations are directed to storage locations associated with at least an SMR storage region of a target storage drive;
for each of the write operations, selecting a least-busy cache group from among the one or more cache groups not presently servicing storage operations;
caching the write data in associated first storage regions of redundant storage drives in the least-busy cache group prior to commit of the write data into at least the SMR storage region of the target storage drive;
responsive to a cache commit event, performing a sequential burst transfer of the write data from selected first storage regions of the one or more cache groups into at least the SMR storage region of the target storage drive;
detecting errors during the sequential burst transfer from the selected first storage regions; and
responsively selecting redundant first storage regions to retrieve the write data for transfer into at least the SMR storage region.

9. The method of claim 8 , further comprising:
monitoring activity levels among the one or more cache groups; and
determining the least-busy cache group for each of the write operations based at least in part on the activity levels indicating at least one of the one or more cache groups having associated storage drives not presently servicing read operations or write operations.

10. The method of claim 8 , further comprising:
establishing data redundancy for caching the write data across the first storage regions of each selected least-busy cache group by at least forming a redundant array of independent disks (RAID) arrangement among the first storage regions of each selected least-busy cache group.

11. The method of claim 8 , further comprising:
detecting the errors during the sequential burst transfer by at least determining that at least a portion of the write data cached in a first one of the first storage regions of an associated cache group has experienced an error; and
responsively using a second one of the first storage regions of the associated cache group to retrieve the write data during the sequential burst transfer into the SMR storage region.

12. The method of claim 8 , further comprising:
detecting the errors during the sequential burst transfer by at least identifying that a first one of the storage drives of an associated cache group has experienced a failure; and
responsively using a second one of the storage drives of the associated cache group to retrieve the write data during the sequential burst transfer into the SMR storage region.

13. The method of claim 8 , wherein:
the write operations are directed to the storage locations that correspond to at least the target storage drive; and
the selected least-busy cache groups comprise storage drives that do not comprise the target storage drive.

14. The method of claim 8 , wherein:
the first storage regions each comprise random read-write regions; and
the SMR storage regions each comprise sequential write regions.

15. A data storage assembly, comprising:
a plurality of hard disk drives (HDDs) each comprising rotating magnetic storage media having a random read-write cache storage region and a sequential write shingled magnetic recording (SMR) storage region;
an enclosure comprising structural elements to mount the plurality of HDDs and at least one external connector for communicatively coupling a control system of the data storage assembly over a host link; and
the control system communicat

Assignees

• HGST Netherlands BV
• Western Digital Tech Inc

Inventors

• Hess Randall L
• Nash Berck E
• Reiser James M
• Roberson Randy L
• Stokes Kris B
• Yandell Jesse L

Classifications

• G06F3/0685

  Hybrid storage combining heterogeneous device types, e.g. hierarchical storage, hybrid arrays · CPC title

• G06F3/061

  Improving I/O performance · CPC title

• G06F3/0619 Primary

  in relation to data integrity, e.g. data losses, bit errors · CPC title

• G06F3/0647 Primary

  Migration mechanisms · CPC title

• G06F3/0665

  at area level, e.g. provisioning of virtual or logical volumes · CPC title