[jira] [Updated] (DERBY-5881) Limitation: Record of a btree secondary index cannot be updated or inserted due to lack of space on the page. (improve error message)

["Kim Haase (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/DERBY-5881?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Kim Haase updated DERBY-5881:
-----------------------------
    Component/s:     (was: Documentation)

Removing Documentation from the affected components, since the doc parts of the issue were spun off into a separate issue, which was fixed.

> Limitation: Record of a btree secondary index cannot be updated or inserted due to lack of space on the page. (improve error message)
> -------------------------------------------------------------------------------------------------------------------------------------
>
>                 Key: DERBY-5881
>                 URL: https://issues.apache.org/jira/browse/DERBY-5881
>             Project: Derby
>          Issue Type: Bug
>          Components: Store
>    Affects Versions: 10.8.2.2, 10.9.1.0
>         Environment: Red Hat Enterprise Linux Server release 6.1 (Santiago) -x86_64
> and
> OpenSuse 11.3 -x86_64
>            Reporter: Tony Brusseau
>              Labels: derby_triage10_11
>
> I'm trying to load a large database. After about 18 million inserts I always get the following error:
>  Exception: java.sql.BatchUpdateException: Limitation: Record of a btree secondary index cannot be updated or inserted due to lack of space on the page.  Use the parameters derby.storage.pageSize and/or derby.storage.pageReservedSpace to work around this limitation.
> Error Code: 20000
> I'm already using the maximum pageSize as stated in the documentation (32768) and I've bumped up the reserved space to 25%. I've also gotten rid of every non-essential index for loading so this is probably related to a primary key index (I had to leave most of them around). Our loading process is very non-trivial and batched so I'm not even sure which insert/index is causing the problem.
> Problems:
> 1. The error message doesn't give any information about which tables/columns/indexes are causing this problem so I have no way of knowing if i can work around this problem somehow.
> 2.The database shouldn't die saying to adjust some magic numbers with no guidance on how to adjust them and by how much. Even if I do adjust them, what is to guarantee I won't get the same error at a later date...maybe even sometime during production which would be extremely bad. I believe an automated correction to this limitation needs to be created.



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