Hi Roman - 

----- Original Message -----
> From: "Roman Shaposhnik" <[email protected]>
> To: [email protected], [email protected]
> Cc: [email protected]
> Sent: Tuesday, June 11, 2013 3:02:41 PM
> Subject: Re: Bringing Hadoop to Fedora
> 
> On Tue, Jun 11, 2013 at 8:54 AM, Tim St Clair <[email protected]> wrote:
> > Greetings Hadoop Development Community,
> >
> > As some of you are aware, we've been looking to bring the upstream 2.X
> > series into the Fedora channels.
> 
> That's pretty awesome! Now, here's a couple question I've got:
>    1. is your focus exclusively on Hadoop or would you want to bring
>    Hadoop ecosystem projects into Fedora as well?
> 

At this stage "I'm" primarily focused on Hadoop, and will defer to matt(cc'd) 
on ecosystem scope.

>    2. since Hadoop is fundamentally a distributed application what
>    are you plans wrt. deployment mechanism? Do you guys envision
>    anything along the lines of juju?

It's too early at this point for us to determine best distribution+deployment 
practices for hadoop.  We've had plenty of experience dealing with other 
distributed applications at very large scales, but we would be all ears to hear 
the experiences of others. 

> 
>    3. can you elaborate a little bit on how do you see this effort leveraging
>    Apache Bigtop?

We've been looking @ numerous rpms, including BigTop, and have steadily been 
making modifications to adhere to Fedora packaging guidelines, and systems 
integration.  We would be happy to chat sometime and here your thoughts.  

> 
> Thanks,
> Roman.
> 

Reply via email to