Profile / optimize SM (MCOL-3417)

[MCOL-3418] optimize around cpimport's inefficiencies during uncompressed import Created: 2019-08-08  Updated: 2019-08-14  Resolved: 2019-08-14

Status: Closed
Project: MariaDB ColumnStore
Component/s: ?
Affects Version/s: None
Fix Version/s: 1.4.0

Type: Sub-Task Priority: Major
Reporter: Patrick LeBlanc (Inactive) Assignee: Patrick LeBlanc (Inactive)
Resolution: Fixed Votes: 0
Labels: None


 Description   

Importing part table from a 500gb dbt3 dataset. When the target table is compressed, observed behavior is expected and good (19k calls to write()). When uncompressed, the opposite (750k calls to write()).

Small writes make the overhead parts in SM dominate. In particular, what rises to the top of the profile are functions related to json parsing.

The first thing I will do is implement a parsed-json cache, then reevaluate.



 Comments   
Comment by Patrick LeBlanc (Inactive) [ 2019-08-14 ]

Cut end-to-end runtime 40-50% for this test case. Took more changes than I expected.

Generated at Thu Feb 08 02:42:35 UTC 2024 using Jira 8.20.16#820016-sha1:9d11dbea5f4be3d4cc21f03a88dd11d8c8687422.