HYBRID MEDIA STORAGE SYSTEM ARCHITECTURE

• US 20110035548A1
• Filed: 02/11/2009
• Published: 02/10/2011
• Est. Priority Date: 02/12/2008
• Status: Active Grant

• Current Assignee: NetApp, Inc.
• Original Assignee: NetApp, Inc.
• Inventors: Miller, Steven C., Kimmel, Jeffrey S., Kleiman, Steven R.
• Granted Patent: US 9,134,917 B2
• Time in Patent Office: Days
• US Class Current: 711/114
• CPC Class Codes: G06F 11/108 Parity data distribution in semiconductor storages, e.g. in SSD
• CPC Class Codes: G06F 16/1805 Append-only file systems, e.g. using logs or journals to store data
• CPC Class Codes: G06F 16/1847 specifically adapted to static storage, e.g. adapted to flash memory or SSD
• CPC Class Codes: G06F 2211/109 Sector level checksum or ECC, i.e. sector or stripe level checksum or ECC in addition to the RAID parity calculation
• CPC Class Codes: G06F 3/061 Improving I/O performance
• CPC Class Codes: G06F 3/0616 in relation to life time, e.g. increasing Mean Time Between Failures [MTBF]
• CPC Class Codes: G06F 3/0643 Management of files
• CPC Class Codes: G06F 3/0647 Migration mechanisms
• CPC Class Codes: G06F 3/0685 Hybrid storage combining heterogeneous device types, e.g. hierarchical storage, hybrid arrays

Abstract

A hybrid media storage architecture has a log-structured file system configured to control a plurality of different storage media organized as hybrid storage media that cooperate to provide a total storage space of a storage system. The log-structured file system is configured to perform initial placement and migration of data, as well as fine-grain write allocation of the data, among storage space locations of the hybrid storage media to thereby improve the performance characteristics of the media. By defining and implementing heuristics and policies directed to, e.g., types of data, the file system may initially place data on any of the different media and thereafter migrate data between the media at fine granularity and without the need for manual enforcement.

Claims

1. A storage system comprising:

a plurality of storage media organized as hybrid storage media that cooperate to provide a total storage space of the storage system, the hybrid storage media including solid-state devices (SSDs) and hard disk drives (HDDs); and

a file system configured to control the hybrid storage media by performing initial placement and migration of data, as well as write allocation of the data, among storage space locations of the hybrid storage media to improve performance characteristics of the media.

2. The storage system of claim 1 wherein the file system is a log-structured file system.

3. The storage system of claim 1 wherein the file system is further configured to implement heuristics and policies that initially write the data on any of the media and thereafter move the data between the media at fine granularity and without the need for manual enforcement.

4. The storage system of claim 3 wherein the heuristics and policies are directed to types of data processed by the file system.

5. The storage system of claim 1 wherein the HDDs are apportioned into regions of related data in accordance with a data layout format of the file system.

6. The storage system of claim 5 wherein the file system implements placement of data within the regions of the HDDs by designating block locations on the HDDs where the data resides.

7. The storage system of claim 6 wherein the file system further applies log-structured techniques to improve random write performance of the SSDs and to move data in fine granularity between the SSDs and HDDs.

8. The storage system of claim 1 further comprising a parity protection module configured to cooperate with the file system to control storage operations to the SSDs and HDDs.

9. The storage system of claim 8 wherein the parity protection module is a RAID module.

10. The storage system of claim 9 further comprising a SSD controller of each SSD, the SSD controller configured to export geometry information to the RAID module, the geometry information including a size of the SSD in terms of device block numbers (dbns).

11. The storage system of claim 10 wherein a dbn is a logical address that the SSD controller presents to the RAID module and that is subject to translation mapping by translation logic of the SSD controller to a physical address of the SSD.

12. The storage system of claim 11 wherein the translation logic is configured to perform wear leveling to reduce failure of the SSD due to repeated erasure and writing of the SSD.

13. The storage system of claim 12 wherein, in response to accessing an erase block via a write operation, the translation logic moves the erase block to a new location on the SSDs and makes a previous location of the block available for reuse by a subsequent write operation.

14. The storage system of claim 8 wherein the RAID module is configured to organize the SSDs as one or more parity groups and manage parity computations and topology information used for placement of data on the SSDs of each group.

15. Apparatus comprising:

a plurality of storage media organized as hybrid storage media of a storage system, the hybrid storage media including solid-state devices (SSDs) of a SSD array and hard disk drives (HDDs) of a HDD array; and

a file system controlling the hybrid storage media, the file system having a first data layout format that provides write access to data containers stored on the SSDs to enable servicing of random data access operations directed to the SSD array, the file system further having a second data layout format that provides sequential read access to large data containers stored on the HDDs to enable servicing of data stored on the HDD array.

16. The apparatus of claim 15 wherein the file system further implements data layout techniques that improve read and write performance to the SSD and HDD arrays.

17. The apparatus of claim 16 wherein the data layout techniques include a first set of write anywhere techniques to enable placement of data anywhere in free, available space on the SSDs of the SSD array.

18. The apparatus of claim 17 wherein the data layout techniques include a second set of techniques to move data sequentially to the HDDs of the HDD array.

19. A method comprising:

processing a write operation at a file system of a storage system, the write operation directed to a data container stored on the storage system;

examining the write operation at the file system to discern a type of data associated with the write operation, the type of data including one of random write data and large sequential write stream data;

if the type of data is random write data, initially storing the random write data associated with the write operation on solid-state devices (SSDs) of a SSD array; and

if the type of data is large sequential write stream data, storing the large sequential write stream data associated with the write operation on hard disk drives (HDDs) of a HDD array.

20. The method of claim 19 further comprising:

moving selected data stored on the SSDs to the HDDs.

21. The method of claim 20 wherein moving selected data comprises:

retrieving the selected data from the SSDs into a memory of the storage system; and

storing the retrieved selected data to one or more regions of the HDDs.