hadoop-common-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Yoram Arnon" <yar...@yahoo-inc.com>
Subject RE: release 0.1
Date Thu, 30 Mar 2006 17:54:01 GMT

-----Original Message-----
From: Konstantin Shvachko [mailto:shv@yahoo-inc.com] 
Sent: Wednesday, March 29, 2006 5:44 PM
To: hadoop-dev@lucene.apache.org
Subject: Re: release 0.1

Just one thing. Let's do it as Doug proposed earlier hadoop dfs -check This
is a good piece of code, but too far from what the real fsck should do to
deserve the name :-) I mean sooner or later somebody will write a real fsck,
and we will have to call it something else.

Andrzej Bialecki wrote:

> Doug Cutting wrote:
>> Andrzej Bialecki wrote:
>>> I would like to commit DFSck before the release, if that's ok.
>> That sounds fine to me.
>> This week, pre-release, we should be careful: (1) not to make changes 
>> that might destabilize things; and (2) we should not add new public 
>> APIs that have not been well-vetted, since once these are in a 
>> released version it will be harder to modify them.  But DFSck is 
>> primarily a new class that prints something informative, right?  So 
>> it shouldn't break anything, and it doesn't add any public API other 
>> than DFSck().main(), right?  (I can't get through to Jira right 
>> now...)
> Right, it doesn't touch anything in the API, just adds a shortcut 
> "hadoop fsck" to bin/hadoop.

View raw message