Data storage backup with lessened cache pollution

Patent metadata

Field	Value
Publication number	US-9519549-B2
Application number	US-201213348395-A
Country	US
Kind code	B2
Filing date	Jan 11, 2012
Priority date	Jan 11, 2012
Publication date	Dec 13, 2016
Grant date	Dec 13, 2016

Abstract

Control of the discard of data from cache during backup of the data. In a computer-implemented system comprising primary data storage; cache; backup data storage; and at least one processor, the processor is configured to identify data stored in the primary data storage for backup to the backup data storage, where the identified data is placed in the cache in the form of portions of the data, and where the portions of data are to be backed up from the cache to the backup storage. Upon backup of each portion of the identified data from the cache to the backup storage, the processor marks the backed up portion of the identified data for discard from the cache. Thus, the backed up data is discarded from the cache right away, lessening cache pollution.

First claim

What is claimed is:

1. A method for backing up data in a data storage system comprising primary data storage, cache, and backup data storage, the method comprising steps of:
identifying, in a list, portions of data stored in the primary data storage for backup to the backup data storage, the portions of data comprising subsets of data;
prefetching downward from the primary data storage the portions of data identified in the list;
caching the prefetched portions of data identified in the list in the cache;
backing up the prefetched portions of data identified in the list from the cache to the backup data storage; and
from the list, upon backup of each of the prefetched portions of data identified in the list from the cache to the backup data storage, marking each backed up portion of the prefetched portions of data identified in the list for accelerated Least Recently Used (LRU) discard from the cache;
wherein the backing up step and the marking step both comprise data blocks, and the backing up step comprises generating a backup access pattern, wherein the backup access pattern is generated from an exact pattern of access from a backup application, and the list is arranged in a same order specified by the backup access pattern such that the prefetching step prefetches, in the same order specified by the backup access pattern, data blocks of prefetched portions of data identified in the list for caching before backup to the backup storage.

2. The method of claim 1 , wherein said identifying step identifies in said list an advance subset of data stored in said data storage for backup.

3. The method of claim 1 , wherein said identifying step identifies in said list a complete set of files of data stored in said data storage for said backup.

4. The method of claim 1 , wherein said identifying step additionally comprising generating said list of physical READs of said data stored in said data storage for said prefetching step and for said marking step.

5. A data storage subsystem comprising:
primary data storage;
cache;
backup data storage; and
at least one processor configured to perform steps of:
identifying, in a list, portions of data stored in the primary data storage for backup to the backup data storage, where the portions of the data comprise subsets of data, and where the portions of data identified in the list are prefetched downward from the primary data storage and cached in the cache;
backing up the prefetched portions of data identified in the list from the cache to the backup data storage; and
from the list, upon backup of each portion of the prefetched portions of data identified in the list from the cache to the backup data storage, marking each backed up portion of the prefetched portions of data identified in the list for accelerated Least Recently Used (LRU) discard from the cache;
wherein the backing up step and the marking step both comprise data blocks, and the backing up step comprises a backup access pattern, wherein the backup access pattern is generated from an exact pattern of access from a backup application, and the list is arranged in a same order of the backup access pattern such that the prefetching step prefetches, in the same order specified by the backup access pattern, data blocks of prefetched portions of data identified in the list for caching before backup to the backup storage.

6. The data storage subsystem of claim 5 , wherein said identifying step of said processor identifies in said list an advance subset of data stored in said stat storage for backup.

7. The data storage subsystem of claim 5 , wherein said identifying step of said processor identifies in said list a complete set of files of data stored in said data storage for said backup.

8. The data storage subsystem of claim 5 , wherein said identifying step of said processor additionally comprises generating said list of physical READs of said data stored in said data storage for said prefetching step and for said marking step.

9. A computer program product for storing data with a data storage system comprising primary data storage; cache; backup data storage; and at least one computer-implemented processor, the computer program product comprising a non-transitory computer-usable storage medium having computer-usable program code embodied therein, the computer-usable program code comprising instructions to cause the processor to perform steps comprising:
identifying, in a list, portions of data stored in the primary data storage for backup to the backup data storage, where the portions of the data comprise subsets of data, and where the portions of the data identified in the list are prefetched downward from the primary storage and cached in the cache; and
backing up the prefetched portions of data identified in the list from the cache to the backup data storage; and
from the list, upon backup of each portion of the prefetched portions of data identified in the list from the cache to the backup data storage, marking each backed up portion of the prefetched portions of data identified in the list for accelerated Least Recently Used (LRU) discard from the cache;
wherein the backing up step and the marking step both comprise data blocks, and the backing up step comprises a backup access pattern, wherein the backup access pattern is generated from an exact pattern of access from a backup application, and the list is arranged in a same order of the backup access pattern such that the prefetching step prefetches, in the same order specified by the backup access pattern, data blocks of prefetched portions of data identified in the list for caching before backup to the backup storage.

10. The computer program product of claim 9 , wherein said computer-usable program code for said processor identifying step comprises code to identify in said list a complete set of files of data stored in said data storage for said backup.

11. The computer program product of claim 9 , wherein said computer-usable program code for said processor identifying step comprises code to identify in said list a complete set of files of data stored in said data storage for said backup.

12. The computer program product of claim 9 , wherein said computer-usable program code for said processor identifying step additionally comprises code to generate said list of physical READs of said data stored in said data storage for said prefetching step and for said marking step.

Assignees

• Blount Lawrence Carter
• Fiske Rahul
• Jones Carl Evan
• Roy Subhojit
• IBM

Inventors

• Blount Lawrence Carter
• Fiske Rahul
• Jones Carl Evan
• Roy Subhojit

Classifications

• G06F11/1456

  Hardware arrangements for backup · CPC title

• G06F2212/6026

  Prefetching based on access pattern detection, e.g. stride based prefetch · CPC title

• G06F2212/6024

  History based prefetching · CPC title

• G06F11/1458 Primary

  Management of the backup or restore process · CPC title