[
https://issues.apache.org/jira/browse/CALCITE-2943?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Jesus Camacho Rodriguez resolved CALCITE-2943.
----------------------------------------------
Resolution: Fixed
Fix Version/s: 1.20.0
Fixed in
[ecc100ea23bdb5dd1a4c34633c3a376856b64970|https://github.com/apache/calcite/commit/ecc100ea23bdb5dd1a4c34633c3a376856b64970].
> Materialized view rewriting logic calls getApplicableMaterializations each
> time the rule is triggered
> -----------------------------------------------------------------------------------------------------
>
> Key: CALCITE-2943
> URL: https://issues.apache.org/jira/browse/CALCITE-2943
> Project: Calcite
> Issue Type: Improvement
> Components: core
> Reporter: Jesus Camacho Rodriguez
> Assignee: Jesus Camacho Rodriguez
> Priority: Major
> Labels: pull-request-available
> Fix For: 1.20.0
>
> Attachments: Screen Shot 2019-03-21 at 2.33.01 PM.png
>
> Time Spent: 0.5h
> Remaining Estimate: 0h
>
> {{RelOptMaterializations.getApplicableMaterializations}} is called each time
> the rule is triggered.
> {code:java}
> ...
> // Obtain applicable (filtered) materializations
> // TODO: Filtering of relevant materializations needs to be
> // improved so we gather only materializations that might
> // actually generate a valid rewriting.
> final List<RelOptMaterialization> applicableMaterializations =
> RelOptMaterializations.getApplicableMaterializations(node,
> materializations);
> ...
> {code}
> When I implemented the rule, I assumed (incorrectly) that
> {{getApplicableMaterializations}} was a lightweight call and hence would help
> discarding materialized views extracted from the planner quickly. It turns
> out that the method can quickly become the most time consuming part of the
> rule execution; I assume the method was just supposed to be used once per
> query.
> !Screen Shot 2019-03-21 at 2.33.01 PM.png!
> Since the prefiltering that we do right now is rather simple and we already
> extract the tables used by queries and materializations within the rule, we
> can just skip the materialization over there.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
