Using a per file activity ratio to optimally relocate data between volumes
A method for identifying data for relocation in a multivolume file system. The method includes generating a file location map, the file location map containing a list of the locations of files that occupy space on each of a plurality of volumes of the file system, wherein The file system comprising least a first volume and a second volume. The method further includes updating the file location map in accordance with changes in a file change log for the file system, and identifying data residing on the first volume of the file system by scanning the file location map. Using the identified data, a ratio of per-file activity during a first time period relative to overall file system activity over a second time period is calculated to derive a file activity ratio for each of the files of the identified data. Files are then selected for relocation based on the file activity ratio.
1. A method for identifying data for relocation in a multivolume file system, comprising:
generating a file location map, the file location map containing a list of the locations of files that occupy space on each of a plurality of volumes of the file system, the file system comprising at least a first volume and a second volume;
updating the file location map in accordance with changes in a file change log for the file system;
identifying, using at least one computer processor, data residing on the first volume of the file system by scanning the file location map; using the identified data, calculating a ratio of per-file activity during a first time period relative to overall file system activity over a second time period to derive a file activity ratio for each of the files of the identified data; and
selecting files for relocation based on the file activity ratio, wherein the file change log is scanned in a sampling interval to identify files active during the sampling interval in response to at least one action.
2. The method of claim 1 , wherein the file change log is scanned to identify the files active during the sampling interval in response to a received a user command.
3. The method of claim 1 , wherein the file change log is scanned to identify the files active during the sampling interval in response to a predetermined schedule.
4. The method of claim 1 , wherein the file change log is scanned to determine which files are active during the sampling interval and wherein the file location map is used to determine which files are located on a volume of interest independent of activity against said which files.
5. The method of claim 1 , wherein files of the identified data having a file activity ratio below a low threshold are relocated from the first volume to the second volume, and files of the identified data having a file activity ratio above a high threshold are relocated from the second volume to the first volume, wherein the first volume is of a higher performance tier than the second volume.
6. The method of claim 5 , wherein the first volume comprises SSD (solid-state disk) based storage and the second volume comprises rotating disk media based storage.
7. The method of claim 6 , wherein files of the identified data having write activity above a high write activity threshold are relocated from the first volume to the second volume.
8. A non-transitory computer readable storage medium having stored thereon computer executable instructions that, if executed by a computer system, cause the computer system to perform a method comprising:
generating a file location map, the file location map containing a list of the locations of files that occupy space on each of a plurality of volumes of the file system, the file system comprising at least a first volume and a second volume, wherein the first volume is of a higher performance tier than the second volume; updating the file location map in accordance with changes in a file change log for the file system;
identifying data residing on the first volume of the file system by scanning the file location map;
using the identified data, calculating a ratio of per-file activity during a first time period relative to overall file system activity over a second time period to derive a file activity ratio for each of the files of the identified data, wherein the file change log is scanned in a sampling interval to identify files active during the sampling interval; and
selecting inactive files for relocation based on the file activity ratio, wherein the file change log is scanned in a sampling interval to identify files active during the sampling interval in response to at least one action.
9. The computer readable storage medium of claim 8 , wherein the file location map is used to determine which files are located on a volume of interest independent of activity against said which files.
10. The computer readable storage medium of claim 8 , wherein files of the identified data having a file activity ratio below a low threshold are relocated from the first volume to the second volume, and files of the identified data having a file activity ratio above a high threshold are relocated from the second volume to the first volume, wherein the first volume is of a higher performance tier than the second volume.
11. The computer readable storage medium of claim 8 , wherein the first volume comprises SSD (solid-state disk) based storage and the second volume comprises rotating disk media based storage.
12. The computer readable storage medium of claim 11 , wherein files of the identified data having write activity above a high write activity threshold are relocated from the first volume to the second volume.
13. The computer readable storage medium of claim 11 , wherein files of the identified data having random read activity above a high write activity threshold are relocated from the second volume to the first volume.
14. The computer readable storage medium of claim 11 , wherein the file change log is scanned to determine which files are active during the sampling interval without scanning file system metadata in substantially its entirety.
15. A multivolume file system, comprising:
a computer system having a processor coupled to a computer readable storage media and executing computer readable code which causes the computer system to:
generate a file location map, the file location map containing a list of the locations of files that occupy space on each of a plurality of volumes of the file system, the file system comprising at least a first volume and a second volume, wherein the first volume is of a higher performance tier than the second volume;
update the file location map in accordance with changes in a file change log for the file system;
identify data residing on the first volume of the file system by scanning the file location map;
using the identified data, calculate a ratio of per-file activity during a first time period relative to overall file system activity over a second time period to derive a file activity ratio for each of the files of the identified data, wherein the file change log is scanned in a sampling interval to identify files active during the sampling interval; and
select files for relocation based on the file activity ratio, wherein the file change log is scanned in a sampling interval to identify files active during the sampling interval in response to at least one action.
16. The multivolume file system of claim 15 , wherein the file location map is used to determine which files are located on a volume of interest independent of activity against said which files.
17. The multivolume file system of claim 15 , wherein files of the identified data having a file activity ratio below a low threshold are relocated from the first volume to the second volume, and files of the identified data having a file activity ratio above a high threshold are relocated from the second volume to the first volume, wherein the first volume is of a higher performance tier than the second volume.
18. The multivolume file system of claim 15 , wherein the first volume comprises SSD (solid-state disk) based storage and the second volume comprises rotating disk media based storage.
19. The multivolume file system of claim 15 , wherein files of the identified data having write activity above a high write activity threshold are relocated from the first volume to the second volume and files of the identified data having random read activity above a high write activity threshold are relocated from the second volume to the first volume.