User-based data tiering
Techniques are provided for user-based data tiering. In an example, a computer maintains a first-in-first-out queue that logs a finite set of users that have most-recently modified a file. This queue can be maintained in an extended attribute of an inode that corresponds to a file. A computer can also maintain a policy that defines how to perform storage tiering on a file based on which users have accessed the file. When a tiering operation is performed, the files specified by a corresponding tiering policy can be evaluated for which users have recently accessed them. When a user specified by the tiering policy has recently modified a file, the file can be placed in tiering queue for tiering.
1. A system, comprising:
a processor; and
a memory that stores executable instructions that, when executed by the processor, facilitate performance of operations, comprising:
identifying a file stored on a first storagepool;
determining that the file has a first extended attribute that is maintained as a first-in-first-out (FIFO) queue of user identities associated with having written to the file, wherein an inode of the first storagepool identifies a first data block of the first storagepool that stores at least part of the file, and a second data block of the first storagepool that stores the first extended attribute;
determining that a first user identity is identified in the first extended attribute, and that the first user identity is identified in a first tiering policy, the first tiering policy specifying a criterion for transfer of files between the first storagepool and a second storagepool, the first storagepool providing a different performance level than the second storagepool; and
moving the file from the first storagepool to the second storagepool based on the first tiering policy.
2. The system of claim 1 , wherein the first user identity is one of a first group of user identities identified in the first extended attribute, wherein the first user identity is one of a second group of user identities identified in a second extended attribute, and wherein the operations further comprise:
moving the file based on the first tiering policy where any of the user identities in the first group of user identities matches any of the user identities in the second group of user identities.
3. The system of claim 1 , wherein the operations further comprise:
processing the file for movement from the first storagepool to the second storagepool based on a group of tiering policies that comprises the first tiering policy and a second tiering policy.
4. The system of claim 3 , wherein the operations further comprise:
performing the processing of the file for movement from the first storagepool to the second storagepool in response to determining that the file meets a criteria of the first tiering policy and fails to meet a criteria of the second tiering policy.
5. The system of claim 3 , wherein the file is analyzed according to the first tiering policy based on the first extended attribute, and wherein the file is analyzed according to the second tiering policy based on the first extended attribute.
6. The system of claim 1 , wherein the first tiering policy indicates a user-aware heat score for moving the file from the first storage pool to the second storagepool, the user-aware heat score being based on a number of accesses of the file and user priorities of user identities of the FIFO queue of user identities.
7. The system of claim 6 , wherein the operations further comprise:
determining that the file is not identified in a list of files for which tracking is not to be performed; and
determining that the first user identity is not identified in a list of user identities for which tracking is not to be performed.
8. A method, comprising:
identifying, by a system comprising a processor, a first file stored on a first storagepool;
determining, by the system, that a first user is identified in a first extended attribute that corresponds to the first file, and that the first user is identified in a first tiering policy, the first tiering policy identifying a criterion for transferring files between the first storagepool and a second storagepool, the first storagepool providing a different performance level than the second storagepool, and wherein an inode of the first storagepool identifies a first data block of the first storagepool that stores at least part of the first file, and a second data block of the first storagepool that stores the first extended attribute; and
moving, by the system, the first file from the first storagepool to the second storagepool based on the first tiering policy.
9. The method of claim 8 , further comprising:
determining, by the system, to refrain from move a second file from the first storagepool to the second storagepool according to the first tiering policy in response to determining that no user identified in the first tiering policy is found in a second extended attribute of the second file that corresponds to the first tiering policy.
10. The method of claim 8 , wherein at least one data block of the first file is identified in an inode of a file system, the inode identifying at least one data block of the first extended attribute.
11. The method of claim 8 , wherein the first extended attribute stores metadata about the first file.
12. The method of claim 8 , wherein the first tiering policy defines a user priority threshold for performance of tiering.
13. The method of claim 12 , further comprising:
performing, by the system, the moving of the first file from the first storagepool to the second storagepool based on determining that the first user has a priority below the user priority threshold.
14. The method of claim 8 , wherein the first extended attribute that is maintained as a first-in-first-out (FIFO) queue of users associated with user devices that have written to the first file.
15. A non-transitory computer-readable medium comprising instructions that, in response to execution, cause a system comprising a processor to perform operations, comprising:
identifying a file stored on a first storagepool;
determining that a first user identity is represented in a first extended attribute that corresponds to the file, and that the first user identity is represented in a first tiering policy, wherein a data structure of the first storagepool identifies a first data block of the first storagepool that stores at least part of the file, and a second data block of the first storagepool that stores the first extended attribute; and
moving the file from the first storagepool to a second storagepool based on the first tiering policy.
16. The non-transitory computer-readable medium of claim 15 , wherein the first tiering policy identifies a criterion for transfer of files between the first storagepool and the second storagepool.
17. The non-transitory computer-readable medium of claim 15 , wherein the first storagepool provides a different performance level than the second storagepool.
18. The non-transitory computer-readable medium of claim 15 , wherein the first tiering policy defines a user priority threshold for performance of tiering.
19. The non-transitory computer-readable medium of claim 18 , wherein the operations further comprise:
performing the moving of the file from the first storagepool to the second storagepool based on determining that the first user identity has a priority specified by the user priority threshold.
20. The non-transitory computer-readable medium of claim 15 , wherein the first user identity is one of a first group of user identities identified in the first extended attribute, wherein the first user identity is one of a second group of user identities identified in a second extended attribute, and wherein the operations further comprise:
moving the file based on the first tiering policy where any user identity in the first group of user identities matches any user identity in the second group of user identities.