[
https://issues.apache.org/jira/browse/HIVE-23965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17176300#comment-17176300
]
Zoltan Haindrich commented on HIVE-23965:
-----------------------------------------
* the description clearly describes that the metastore data is a composition of
questionable quality stuff...so we are running our planning tests against some
weird metastore content....
* I don't think adding more tests will increase test coverage - in this case we
are talking about queries which are already run 2 times already - I've seen
people updating q.out's like crazy....so adding an extra 100 q.out-s will not
neccessarily increase coverage...
* the independence from having docker setup is a great - the new approach uses
docker - but if that's a problem we could try to come up with some other
approach - I'm wondering about using an archived derby database with metastore
data
* the metastore content lodader approach is quite unfortunate - IIRC once I had
to fix up something in the loader once... because I made some changes to the
column statistics
I think we should remove the old approach...and run tests against the new;
more-realistic schema.
> Improve plan regression tests using TPCDS30TB metastore dump and custom
> configs
> -------------------------------------------------------------------------------
>
> Key: HIVE-23965
> URL: https://issues.apache.org/jira/browse/HIVE-23965
> Project: Hive
> Issue Type: Improvement
> Reporter: Stamatis Zampetakis
> Assignee: Stamatis Zampetakis
> Priority: Major
> Labels: pull-request-available
> Time Spent: 1h 10m
> Remaining Estimate: 0h
>
> The existing regression tests (HIVE-12586) based on TPC-DS have certain
> shortcomings:
> The table statistics do not reflect cardinalities from a specific TPC-DS
> scale factor (SF). Some tables are from a 30TB dataset, others from 200GB
> dataset, and others from a 3GB dataset. This mix leads to plans that may
> never appear when using an actual TPC-DS dataset.
> The existing statistics do not contain information about partitions something
> that can have a big impact on the resulting plans.
> The existing regression tests rely on more or less on the default
> configuration (hive-site.xml). In real-life scenarios though some of the
> configurations differ and may impact the choices of the optimizer.
> This issue aims to address the above shortcomings by using a curated
> TPCDS30TB metastore dump along with some custom hive configurations.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
