accumulo-notifications mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "ASF GitHub Bot (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (ACCUMULO-626) create an iterator fuzz tester
Date Mon, 09 Nov 2015 19:41:11 GMT

    [ https://issues.apache.org/jira/browse/ACCUMULO-626?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14997227#comment-14997227
] 

ASF GitHub Bot commented on ACCUMULO-626:
-----------------------------------------

Github user joshelser commented on the pull request:

    https://github.com/apache/accumulo/pull/50#issuecomment-155168971
  
    > For iterators that can be run at major compaction, how can we exercise delete keys?
    
    Very good question! The completely mocked-out IteratorEnvironment is a sign that I haven't
thought past the scan case. I'm not even sure if the ColumnFiltering iterator with SortedMapIterator
is even a 100% valid in-memory reproduction of how it works in a TabletServer.
    
    I had left input entirely up to the user, so I'm not sure of a simple way to test deletions
without potentially messing up some expected table structure. There are some tricky edge cases
at compaction that I haven't yet considered how to encapsulate. Combiners on non-full compactions
are another good example.


> create an iterator fuzz tester
> ------------------------------
>
>                 Key: ACCUMULO-626
>                 URL: https://issues.apache.org/jira/browse/ACCUMULO-626
>             Project: Accumulo
>          Issue Type: New Feature
>          Components: test
>            Reporter: Eric Newton
>            Assignee: Josh Elser
>
> Users often write iterators without fully understanding its limits and lifetime. Accumulo
should have an iterator fuzz-tester which will take user data and run the iterator under extreme
conditions.  For example, it should re-create and re-seek the iterator with every key returned.
 It could automatically compare results of such a run with the naive run, which seeks to the
beginning and scans all the data.



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

Mime
View raw message