US 11,995,037 B2
Programmatically choosing preferred storage parameters for files in large-scale distributed storage systems
Murray M. Stokely, Mountain View, CA (US); and Arif Merchant, Mountain View, CA (US)
Assigned to Google LLC, Mountain View, CA (US)
Filed by Google LLC, Mountain View, CA (US)
Filed on Dec. 3, 2019, as Appl. No. 16/702,535.
Application 16/702,535 is a continuation of application No. 16/434,376, filed on Jun. 7, 2019, granted, now 11,113,242.
Application 16/434,376 is a continuation of application No. 15/271,739, filed on Sep. 21, 2016, granted, now 10,339,108, issued on Jul. 2, 2019.
Application 15/271,739 is a continuation of application No. 14/033,255, filed on Sep. 20, 2013, granted, now 9,477,679, issued on Oct. 25, 2016.
Prior Publication US 2020/0104281 A1, Apr. 2, 2020
Int. Cl. G06F 16/17 (2019.01); G06F 3/06 (2006.01); G06F 16/182 (2019.01)
CPC G06F 16/1727 (2019.01) [G06F 3/0617 (2013.01); G06F 3/0643 (2013.01); G06F 3/067 (2013.01); G06F 16/182 (2019.01); G06F 16/184 (2019.01)] 20 Claims
OG exemplary drawing
 
1. A method comprising:
grouping, by a processor, a subset of files from among a set of files within a distributed storage system based on using a common creation date of each file of the subset of files;
receiving, by the processor, a current file location for a respective one of the files of the subset of files within the distributed storage system;
receiving, by the processor, access information associated with a plurality of files of the subset of files, the access information generalizing access patterns for the plurality of files of the subset of files, each file of the plurality of files different from the respective one of the files of the subset of files;
determining, by the processor using the access information associated with the plurality of files of the subset of files and the common creation date of each file of the subset of files, an access pattern for the respective one of the files of the subset of files within the distributed storage system;
determining, by the processor, a new file location for the respective one of the files of the subset of files within the distributed storage system that minimizes a storage cost associated with storing the respective one of the files of the subset of files within the distributed storage system, the new file location based on:
the current file location of the respective one of the files of the subset of files within the distributed storage system; and
the determined access pattern for the respective one of the files of the subset of files within the distributed storage system; and
generating, by the processor, an alert for a user of the distributed storage system indicating the new file location for the respective one of the files of the subset of files within the distributed storage system.