You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@ignite.apache.org by "Pavel Kuznetsov (JIRA)" <ji...@apache.org> on 2018/10/19 13:00:00 UTC

[jira] [Commented] (IGNITE-9947) JDBC thin: Batch update is not performed if streaming state changed before executeBatch()

    [ https://issues.apache.org/jira/browse/IGNITE-9947?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16656778#comment-16656778 ] 

Pavel Kuznetsov commented on IGNITE-9947:
-----------------------------------------

Problem seems to be caused that we have 2 batch storages: for streaming on and off modes. 
See org.apache.ignite.internal.jdbc.thin.JdbcThinConnection.StreamState#streamBatch
and org.apache.ignite.internal.jdbc.thin.JdbcThinStatement#batch

We should either use one queue for batches or handle both queues in the org.apache.ignite.internal.jdbc.thin.JdbcThinStatement#executeBatch

[~tledkov-gridgain] What do you think?

> JDBC thin: Batch update is not performed if streaming state changed before executeBatch()
> -----------------------------------------------------------------------------------------
>
>                 Key: IGNITE-9947
>                 URL: https://issues.apache.org/jira/browse/IGNITE-9947
>             Project: Ignite
>          Issue Type: Bug
>          Components: sql
>    Affects Versions: 2.6
>            Reporter: Pavel Kuznetsov
>            Priority: Major
>
> Thin driver is affected.
> How to reproduce:
> 0) Create table 
> 1) Create PreparedStatement "INSERT INTO TAB VALUES (?, ?)"
> 2) Set statement's args and call .addBatch()
> 3) Turn on streaming
> 4) call .executeBatch()
> 5) Turn off streaming to flush streamer.
> After that table should contain batched data, but it doesn't



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)