US 12,260,109 B2
Optimizing data storage with access windows
Konstantinos Nikoloudakis, San Francisco, CA (US); Sven Koehler, Sunnyvale, CA (US); Danyao Wang, Kitchener (CA); Sahand Saba, Kitchener (CA); Long Fei, Campbell, CA (US); Simon Tyler Wise, New Hamburg (CA); and David Halladay Schneider, Mountain View, CA (US)
Assigned to Google LLC, Mountain View, CA (US)
Filed by Google LLC, Mountain View, CA (US)
Filed on Jan. 20, 2024, as Appl. No. 18/418,253.
Application 18/418,253 is a continuation of application No. 17/656,645, filed on Mar. 26, 2022, granted, now 11,893,265.
Application 17/656,645 is a continuation of application No. 16/838,627, filed on Apr. 2, 2020, granted, now 11,307,978, issued on Apr. 19, 2022.
Application 16/838,627 is a continuation of application No. 16/194,115, filed on Nov. 16, 2018, granted, now 10,621,079, issued on Apr. 14, 2020.
Application 16/194,115 is a continuation of application No. 15/614,871, filed on Jun. 6, 2017, granted, now 10,133,658, issued on Nov. 20, 2018.
Prior Publication US 2024/0160375 A1, May 16, 2024
Int. Cl. G06F 3/06 (2006.01)
CPC G06F 3/0652 (2013.01) [G06F 3/0608 (2013.01); G06F 3/067 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A computer-implemented method when executed by data processing hardware causes the data processing hardware to perform operations comprising:
aggregating access data for a plurality of data objects stored at a storage environment;
for each respective data object of the plurality of data objects:
determining a corresponding access window associated with the respective data object based on the aggregated access data, the corresponding access window comprising an expiration date in the future; and
determining a corresponding future storage cost associated with the respective data object;
identifying a particular data object of the plurality of data objects based on:
a likelihood that the particular data object will not be accessed outside the corresponding access window of the particular data object; and
the corresponding future storage cost of the particular data object satisfying a predetermined cost threshold; and
generating an instruction to delete the particular data object upon expiration of the corresponding access window of the particular data object.