[jira] [Updated] (CASSANDRA-4816) Broken get_paged_slice

["Sylvain Lebresne (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/CASSANDRA-4816?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Sylvain Lebresne updated CASSANDRA-4816:
----------------------------------------

    Attachment: 4816.txt

My bet would be that this is due to the memtable iterator. The way we implement get_paged_slice, we reset the start of the QueryFilter used after the first row have been read. Which means that for each row, we need to make sure the filter is not used until we've processed the preceding row. This is the case for the sstable iterator, but not for the memtable one. Attaching patch to change that (I haven't tested the patch though).
                
> Broken get_paged_slice 
> -----------------------
>
>                 Key: CASSANDRA-4816
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-4816
>             Project: Cassandra
>          Issue Type: Bug
>    Affects Versions: 1.1.6
>            Reporter: Piotr Kołaczkowski
>            Priority: Blocker
>         Attachments: 4816.txt
>
>
> get_paged_slice doesn't reset the start column filter for the second returned row sometimes. So instead of getting a slice:
> row 0: <start_column>...<last_column_in_row>
> row 1: <first column in a row>...<last_column_in_row>
> row 2: <first column in a row>...
> you sometimes get:
> row 0: <start_column>...<last_column_in_row>
> row 1: <start_column>...<last_column_in_row>
> row 2: <first column in a row>...

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira