You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@cassandra.apache.org by "Aleksey Yeschenko (JIRA)" <ji...@apache.org> on 2018/09/24 15:03:00 UTC
[jira] [Comment Edited] (CASSANDRA-14772) Fix issues in audit /
full query log interactions
[ https://issues.apache.org/jira/browse/CASSANDRA-14772?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16625918#comment-16625918 ]
Aleksey Yeschenko edited comment on CASSANDRA-14772 at 9/24/18 3:02 PM:
------------------------------------------------------------------------
This is too narrow of an interpretation of a freeze.
What can't and shouldn't go in is new features (that are not testing related).
Sometimes you will have issues with code that made it in before the freeze that needs to be addressed, and doing a full-on revert is too drastic a response.
Sometimes you'll have small features that are required strictly for testing purposes - think logging extra information in FQL to enable replay testing, or returning extra metadata from the server to enable cluster fuzz testing and validation.
Certain cleanup and maintenance also easily falls under stabilisation umbrella.
Again, you have to be reasonable about it and exercise judgement on case-by-case basis.
was (Author: iamaleksey):
This is too legalistic of an interpretation of a freeze.
What can't and shouldn't go in is new features (that are not testing related).
Sometimes you will have issues with code that made it in before the freeze that needs to be addressed, and doing a full-on revert is too drastic a response.
Sometimes you'll have small features that are required strictly for testing purposes - think logging extra information in FQL to enable replay testing, or returning extra metadata from the server to enable cluster fuzz testing and validation.
Certain cleanup and maintenance also easily falls under stabilisation umbrella.
Again, you have to be reasonable about it and exercise judgement on case-by-case basis.
> Fix issues in audit / full query log interactions
> -------------------------------------------------
>
> Key: CASSANDRA-14772
> URL: https://issues.apache.org/jira/browse/CASSANDRA-14772
> Project: Cassandra
> Issue Type: Bug
> Reporter: Marcus Eriksson
> Assignee: Marcus Eriksson
> Priority: Major
> Fix For: 4.0
>
>
> There are some problems with the audit + full query log code that need to be resolved before 4.0 is released:
> * Fix performance regression in FQL that makes it less usable than it should be.
> * move full query log specific code to a separate package
> * do some audit log class renames (I keep reading {{BinLogAuditLogger}} vs {{BinAuditLogger}} wrong for example)
> * avoid parsing the CQL queries twice in {{QueryMessage}} when audit log is enabled.
> * add a new tool to dump audit logs (ie, let fqltool be full query log specific). fqltool crashes when pointed to them.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@cassandra.apache.org
For additional commands, e-mail: commits-help@cassandra.apache.org