[
https://issues.apache.org/jira/browse/IMPALA-14417?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Pranav Yogi Lodha updated IMPALA-14417:
---------------------------------------
Description:
Currently, the JDBC external data source integration issues an additional
COUNT(*) query in the {{prepare()}} phase to estimate the number of rows in a
JDBC table. While this helps the planner choose a better parallelization
strategy for complex queries (e.g. joins or aggregations), it also introduces
an extra JDBC round trip.
To give users more control, we propose introducing a query option that enables
or disables row count estimation for JDBC tables. When enabled, Impala will
issue the COUNT(*) query during prepare. When disabled, Impala will skip this
step, avoiding the additional JDBC round trip.
was:
Currently, the JDBC external data source integration issues an additional
{{COUNT(*)}} query in the {{prepare()}} phase to estimate the number of rows in
a JDBC table. While this helps the planner choose a better parallelization
strategy for complex queries (e.g. joins or aggregations), it also introduces
an extra JDBC round trip.
To give users more control, we propose introducing a query option that enables
or disables row count estimation for JDBC tables. When enabled, Impala will
issue the {{COUNT(*) }}query during prepare. When disabled, Impala will skip
this step, avoiding the additional JDBC round trip.
> Query option for JDBC row count estimation
> ------------------------------------------
>
> Key: IMPALA-14417
> URL: https://issues.apache.org/jira/browse/IMPALA-14417
> Project: IMPALA
> Issue Type: New Feature
> Reporter: Pranav Yogi Lodha
> Priority: Major
>
> Currently, the JDBC external data source integration issues an additional
> COUNT(*) query in the {{prepare()}} phase to estimate the number of rows in a
> JDBC table. While this helps the planner choose a better parallelization
> strategy for complex queries (e.g. joins or aggregations), it also introduces
> an extra JDBC round trip.
> To give users more control, we propose introducing a query option that
> enables or disables row count estimation for JDBC tables. When enabled,
> Impala will issue the COUNT(*) query during prepare. When disabled, Impala
> will skip this step, avoiding the additional JDBC round trip.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
