You are viewing a plain text version of this content. The canonical link for it is here.
Posted to derby-dev@db.apache.org by "Mike Matrigali (JIRA)" <ji...@apache.org> on 2011/06/10 23:37:59 UTC

[jira] [Created] (DERBY-5273) enhance SYSCS_CHECK_TABLE to check every page in the database

enhance SYSCS_CHECK_TABLE to check every page in the database
-------------------------------------------------------------

                 Key: DERBY-5273
                 URL: https://issues.apache.org/jira/browse/DERBY-5273
             Project: Derby
          Issue Type: Improvement
    Affects Versions: 10.8.1.2
            Reporter: Mike Matrigali


The current consistency checker operates at a logical layer.  It checks that indexes and base user tables are consistent.  In doing so
it checks most of the pages in the database, but not all.  I don't think it checks blob/clob chains.  I know it does not check free pages.

One simple enhancement would be to just read every page of every file in the database into the cache.  This would at least check that
all pages were valid.   I believe code to do this can either be copied or shared with the backup code which at a low level reads each page
from disk into cache and then writes it to the backup directory.  

DERBY-5266 looks like a case where the consistency checker missed an invalid page that backup is subsequently finding.  It would
be great if anyone has an example of a database that they can post that has a similar problem so that we can test a real live problem, 
rather than making one up.  

--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Updated] (DERBY-5273) enhance SYSCS_CHECK_TABLE to check every page in the database

Posted by "Mike Matrigali (JIRA)" <ji...@apache.org>.
     [ https://issues.apache.org/jira/browse/DERBY-5273?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Mike Matrigali updated DERBY-5273:
----------------------------------

    Issue & fix info: [Newcomer]

I am marking this newcomer.   Might be good as a second project working on Derby.  Good for someone who wants to learn a bit
about the underlying storage structure.  Most of the work is figuring out what backup does and then work out a way to something
similar from a call to the consistency checker.  Probably means adding a new raw store interface that can be called from 
the consistency checker procedure similar to how backup is called.

> enhance SYSCS_CHECK_TABLE to check every page in the database
> -------------------------------------------------------------
>
>                 Key: DERBY-5273
>                 URL: https://issues.apache.org/jira/browse/DERBY-5273
>             Project: Derby
>          Issue Type: Improvement
>    Affects Versions: 10.8.1.2
>            Reporter: Mike Matrigali
>
> The current consistency checker operates at a logical layer.  It checks that indexes and base user tables are consistent.  In doing so
> it checks most of the pages in the database, but not all.  I don't think it checks blob/clob chains.  I know it does not check free pages.
> One simple enhancement would be to just read every page of every file in the database into the cache.  This would at least check that
> all pages were valid.   I believe code to do this can either be copied or shared with the backup code which at a low level reads each page
> from disk into cache and then writes it to the backup directory.  
> DERBY-5266 looks like a case where the consistency checker missed an invalid page that backup is subsequently finding.  It would
> be great if anyone has an example of a database that they can post that has a similar problem so that we can test a real live problem, 
> rather than making one up.  

--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Updated] (DERBY-5273) enhance SYSCS_CHECK_TABLE to check every page in the database

Posted by "Mike Matrigali (Updated) (JIRA)" <ji...@apache.org>.
     [ https://issues.apache.org/jira/browse/DERBY-5273?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Mike Matrigali updated DERBY-5273:
----------------------------------

    Component/s: Store
    
> enhance SYSCS_CHECK_TABLE to check every page in the database
> -------------------------------------------------------------
>
>                 Key: DERBY-5273
>                 URL: https://issues.apache.org/jira/browse/DERBY-5273
>             Project: Derby
>          Issue Type: Improvement
>          Components: Store
>    Affects Versions: 10.8.1.2
>            Reporter: Mike Matrigali
>
> The current consistency checker operates at a logical layer.  It checks that indexes and base user tables are consistent.  In doing so
> it checks most of the pages in the database, but not all.  I don't think it checks blob/clob chains.  I know it does not check free pages.
> One simple enhancement would be to just read every page of every file in the database into the cache.  This would at least check that
> all pages were valid.   I believe code to do this can either be copied or shared with the backup code which at a low level reads each page
> from disk into cache and then writes it to the backup directory.  
> DERBY-5266 looks like a case where the consistency checker missed an invalid page that backup is subsequently finding.  It would
> be great if anyone has an example of a database that they can post that has a similar problem so that we can test a real live problem, 
> rather than making one up.  

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira