[PL #13410] Private Planetlab Broken

[jaffe at cs.huji.ac.il via RT]

Email Recipients (see http://www.planet-lab.org/Support)
       Owner: Nobody
       Requestor: jaffe at cs.huji.ac.il
       Ticket Ccs: jaffe at cs.huji.ac.il, kirk at cs.huji.ac.il

==================================================

Thanks for the update.   I believe that there has been a lack of 
communication on all sides (particularly on my side since I didn't tell you 
that it was live).  I'm looking forward to the May Workshop so that we can 
clear the air and get things moving again.

In the interest of sharing, We have got the following projects underway:

Gangali as a root service -- Done -- See www.everlab.org/ganglia
AFS Access for all nodes -- Working but not automatically installed
A Shared Condor Slice for compute intensive tasks -- Working but not yet 
packages in RPM's
Multi-process support -- Working based on the stock smp FC4 kernel in 3.2 
PlanetLab
Boot via PXEBOOT -- Working and deployed

Our biggest challenge is getting our researchers to use the system.  Until 
we get Condor and AFS up and stable, our compute intensive users are going 
to hold back.  We do seem to be stable and hope to add more nodes to the 
system within the next few weeks.

Elliot


----- Original Message ----- 
From: "Steve Muir via RT" <devel at planet-lab.org>
To: <jaffe at cs.huji.ac.il>
Sent: Sunday, April 16, 2006 12:07 AM
Subject: Re: [PL #13410] Private Planetlab Broken

> i think perhaps the biggest reason for us not keeping you more closely
> involved in various efforts is that we weren't aware that you were
> actually running a real system.  i personally was under the impression
> that you had built a prototype to demonstrate feasibility, but that it
> wasn't actually being used as an operational facility; if i had known this
> i certainly wouldn't have taken that approach that i did in making changes
> to NM (just as one example).
>
> cheers,
>
> steve
>
>