activemq-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Timothy Bish (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (AMQ-6317) ActiveMQ createSchemaStatements are not executed on init if a previous createSchemaStatement failed on execution
Date Thu, 16 Jun 2016 16:57:05 GMT

    [ https://issues.apache.org/jira/browse/AMQ-6317?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15334147#comment-15334147
] 

Timothy Bish commented on AMQ-6317:
-----------------------------------

Looks like the patch still uses the uk.org.lidalia instead of the DefaultTestAppender as Gary
has pointed out previously. 

> ActiveMQ createSchemaStatements are not executed on init if a previous createSchemaStatement
failed on execution
> ----------------------------------------------------------------------------------------------------------------
>
>                 Key: AMQ-6317
>                 URL: https://issues.apache.org/jira/browse/AMQ-6317
>             Project: ActiveMQ
>          Issue Type: Bug
>          Components: JDBC
>    Affects Versions: 5.12.3, 5.13.3
>         Environment: PostgreSql, Linux, Mac OS X
>            Reporter: Jeroen Bastijns
>
> On init the DefaultJDBCAdapter.doCreateTables-method is executed. This provides the tables
needed for ActiveMQ in persistent mode.
> The createSchemaStatements are executed within one SQL Statement. 
> When one of the createSchemaStatements throws an SQLException (table already exists)
the SQL Statement's transaction is aborted and all following createSchemaStatements are ignored.
> This is unwanted behaviour as the comment on the code block states that new statements
like for example 'ALTER TABLE' (introduced in new versions of activemq) should be executed
if the tables already exist.
> We had this issue when adding an extra createSchemaStatement for a JobScheduler table.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Mime
View raw message