US 12,353,370 B2
Elastic, ephemeral in-line deduplication service
Saeed Ghanbari, New Haven, CT (US); Barry Patrick Benight, San Jose, CA (US); Deepak Raghu Kenchammana-Hosekote, Mountain View, CA (US); and Shiva Chaitanya, Watertown, MA (US)
Assigned to NetApp, Inc., San Jose, CA (US)
Filed by NetApp Inc., San Jose, CA (US)
Filed on Nov. 30, 2022, as Appl. No. 18/071,790.
Application 18/071,790 is a continuation of application No. 16/807,636, filed on Mar. 3, 2020, granted, now 11,537,573.
Application 16/807,636 is a continuation of application No. 14/866,683, filed on Sep. 25, 2015, granted, now 10,621,151, issued on Apr. 14, 2020.
Prior Publication US 2023/0086414 A1, Mar. 23, 2023
This patent is subject to a terminal disclaimer.
Int. Cl. G06F 16/215 (2019.01); G06F 16/174 (2019.01); G06F 16/2458 (2019.01)
CPC G06F 16/215 (2019.01) [G06F 16/1748 (2019.01); G06F 16/2471 (2019.01)] 20 Claims
OG exemplary drawing
 
1. A method, comprising:
in response to receiving a write request targeting a data unit, dividing the data unit into sub-units according to a sub-unit size;
determining, by a redirector, a number of deduplicator instances to instantiate for deduplicating the sub-units based upon a deduplication service policy specifying a threshold amount of data that can be processed by a single executing deduplicator instance;
creating a data unit manifest for the data unit with an indication of an order and count of the sub-units, wherein the data unit manifest is populated with paths to the sub-units according to a hierarchical namespace or a flat namespace where a path is a namespace identifier used to obtain data of a constituent sub-unit; and
requesting deduplication for the sub-units by the number of deduplicator instances using the paths within the data unit manifest.