[jira] Commented: (DDLUTILS-208) PostgreSQL: RecreateTableChange loses autoIncrement

["Thomas Dudziak (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/DDLUTILS-208?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12644849#action_12644849 ] 

Thomas Dudziak commented on DDLUTILS-208:
-----------------------------------------

I'm not sure I understand the problem. RecreateTableChange will cause the table to be recreated, and the order of DROP TABLE and DROP SEQUENCE statements makes sure that the table id dropped before the sequence, so how could that cause errors on columns using the sequence ?

> PostgreSQL: RecreateTableChange loses autoIncrement
> ---------------------------------------------------
>
>                 Key: DDLUTILS-208
>                 URL: https://issues.apache.org/jira/browse/DDLUTILS-208
>             Project: DdlUtils
>          Issue Type: Bug
>          Components: Core (No specific database), Core - PostgreSql
>         Environment: PostgreSQL 8.3.1
>            Reporter: Rijk van Haaften
>            Assignee: Thomas Dudziak
>             Fix For: 1.1
>
>
> PlatformImplBase.processChange(RecreateTableChange) does not migrate (autoincrement) sequences, but does call SqlBuilder.dropTables. PostgreSqlBuilder.dropTables explicitly drops the sequences:
>         for (int idx = 0; idx < columns.length; idx++)
>         {
>             dropAutoIncrementSequence(table, columns[idx]);
>         }
> (On the other hand, HsqlDbBuilder does not drop the sequence for instance.)
> The sequence is created again, but that causes 'duplicate' errors on the column using the sequence.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.