[
https://issues.apache.org/jira/browse/DERBY-4116?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12689454#action_12689454
]
Knut Anders Hatlen commented on DERBY-4116:
-------------------------------------------
The arguments to openScan() look correct to me. Since we don't actually scan
the conglomerate, many of them wouldn't make any difference anyway, I think.
Now we update the heap conglomerate inside the for loop, which means that we'll
update it once per index. Would it be enough to update it once, after we have
gone through the last index?
Nit: openScan() should be called outside the try block (if it fails, there's no
open scan to close). If it's moved out of the try block, the check for
heapSC!=null in the finally block can be removed too.
> SYSCS_UTIL.SYSCS_UPDATE_STATISTICS should update the store estimated row
> count for the table
> --------------------------------------------------------------------------------------------
>
> Key: DERBY-4116
> URL: https://issues.apache.org/jira/browse/DERBY-4116
> Project: Derby
> Issue Type: Bug
> Components: SQL
> Affects Versions: 10.5.1.0, 10.6.0.0
> Reporter: Kathey Marsden
> Assignee: Kathey Marsden
> Attachments: derby-4116_diff.txt
>
>
> When SYSCS_UTIL.SYSCS_UPDATE_STATISTICS is run, it doesn't update the store
> estimated row count. The program oldSelectivity.java attached to DERBY-3955
> shows that the statistics are not updated.
> See discussions in DERBY-3955
> https://issues.apache.org/jira/browse/DERBY-3955?focusedCommentId=12688789&page=com.atlassian.jira.plugin.system.issuetabpanels%3Acomment-tabpanel#action_12688789
> https://issues.apache.org/jira/browse/DERBY-3955?focusedCommentId=12688813&page=com.atlassian.jira.plugin.system.issuetabpanels%3Acomment-tabpanel#action_12688813
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
