[
https://issues.apache.org/jira/browse/PHOENIX-5224?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16816673#comment-16816673
]
Lars Hofhansl commented on PHOENIX-5224:
----------------------------------------
Is that theoretical based on general JDBC best practices, or do you have some
numbers? How much is the runtime of these tests actually improve?
I'm not opposed to the change, it's a nice improvement. Just wondering about
the claim that this "will greatly damage the performance" in Phoenix.
> Iteratively executing the same SQL query will greatly damage the performance
> ----------------------------------------------------------------------------
>
> Key: PHOENIX-5224
> URL: https://issues.apache.org/jira/browse/PHOENIX-5224
> Project: Phoenix
> Issue Type: Bug
> Reporter: bd2019us
> Priority: Major
> Labels: pull-request-available
> Attachments: 1.patch
>
> Time Spent: 1h 20m
> Remaining Estimate: 0h
>
> Locations:
> (1)
> phoenix-core/src/it/java/org/apache/phoenix/end2end/index/IndexRebuildIncrementDisableCountIT.java:170
> (2) phoenix-core/src/it/java/org/apache/phoenix/tx/TxCheckpointIT.java:108
> (3) phoenix-core/src/it/java/org/apache/phoenix/end2end/DeleteIT.java:794,858
> (4)
> phoenix-core/src/it/java/org/apache/phoenix/end2end/UpsertSelectAutoCommitIT.java:240
> (5) phoenix-core/src/it/java/org/apache/phoenix/end2end/DropTableIT.java:35
> At these locations, the java.sql.Statement is intensively used in a loop and
> executing the same SQL query, which will highly damage the performance of
> database access. In such cases, java.sql.PreparedStatement should be used to
> enable the database compiling the SQL command ahead of time, which has a much
> better performance compared with Statement.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
