US 12,007,997 B2
Metadata search via N-gram index
Lin Chan, Bellevue, WA (US); Tianyi Chen, Kirkland, WA (US); Benoit Dageville, San Mateo, CA (US); Yiming Kang, Seattle, WA (US); Jun Luo, Bellevue, WA (US); Nithin Mahesh, Redmond, WA (US); Eric Robinson, Sammamish, WA (US); and Brian Smith, Hillsborough, CA (US)
Assigned to Snowflake Inc., Bozeman, MT (US)
Filed by Snowflake Inc., Bozeman, MT (US)
Filed on Mar. 14, 2023, as Appl. No. 18/183,483.
Application 18/183,483 is a continuation of application No. 17/457,010, filed on Nov. 30, 2021, granted, now 11,636,110.
Claims priority of provisional application 63/273,348, filed on Oct. 29, 2021.
Prior Publication US 2023/0214391 A1, Jul. 6, 2023
This patent is subject to a terminal disclaimer.
Int. Cl. G06F 16/00 (2019.01); G06F 7/14 (2006.01); G06F 16/22 (2019.01); G06F 16/23 (2019.01); G06F 16/2453 (2019.01); G06F 21/62 (2013.01)
CPC G06F 16/24549 (2019.01) [G06F 7/14 (2013.01); G06F 16/2228 (2019.01); G06F 16/2379 (2019.01); G06F 21/6218 (2013.01); G06F 21/6227 (2013.01)] 27 Claims
OG exemplary drawing
 
1. A method comprising:
storing a log associated with an operation performed on a source table;
creating or updating an index record for a N-Gram index associated with the source table to generate a new version of the N-Gram index based on the log using a background process;
receiving, from a user, a search request including a search string; and
in response to the search request, processing the search request by performing steps including:
loading, in a log cache, log data associated with new logs not yet reflected in the N-Gram index;
retrieving index records matching the search string;
storing the matched index records in an index cache;
merging the log data in the log cache with the matched index records in the index cache to generate merged data; and
generating results of the search request based on the merged data, wherein processing the search request is a separate process than the background process used for updating the new logs in the N-gram index.