UTCS Colloquium: Kevin Kane and Brian A. LaMacchia/Microsoft: Cyclotron: Building a Cycle-Stealing Computing Grid Using Virtualization and Constrained Delegation ACES 2.402 Tuesday November 18 2008 11:00 a.m.

Contact Name: 
Jenna Whitney
Date: 
Nov 18, 2008 11:00am - 12:00pm

There is a signup schedule for this event (UT EID req

uired).

Type of Talk: UTCS Colloquium

Speaker(s)/Affiliation:
Kevin Kane and Brian A. LaMacchia/Microsoft

Date/Time: Tuesday N

ovember 18 2008 11:00 a.m.

Location: ACES 2.402

Host: Jim
Browne

Talk Title: Cyclotron: Building a Cycle-Stealing Computing

Grid Using Virtualization and Constrained Delegation

Talk Abstract:<

br>Cycle-scavenging computing grids appeal to organizations
with large

numbers of work-stations that remain idle outside
of working hours but
security and isolation issues that come
with the use of non-dedicated

resources have slowed the
adoption of grids in the enterprise. In this

talk we present
Cyclotron a prototype cycle-stealing grid solution wit

h
enterprise-friendly security and isolation properties. Cyclotron
leverages virtualization to provide isolation between cycle-
scavenging
jobs and donors''jobs see a virtual machine pre-
configured with their

environmental requirements and donor
machines receive a guarantee that

scavengers cannot interfere
with their machines. Additionally Cyclotr

on leverages the
SecPAL declarative security policy language to provide
fine-
grained access control and constrained delegation services
ac

ross the grid.

The Cyclotron project is one of a number of ongoing
incubation
efforts within the Office of the Chief Research and Strategy

Officer (CRSO) at Microsoft. CRSO Incubation projects are
part ap

plied research part advanced development their goal
is to aid in the t

ransfer of basic research technology from the
laboratory to product tea

ms. Successful incubations often result
in the direct delivery of feat

ures to product organizations but
sometimes they deliver proof-of-conc

ept prototypes that demon-
strate the feasibility of an approach. Occas

ionally (and like basic
research) incubations succeed through failure a

nd help product
organizations avoid pursuing unprofitable technology pa

ths.

Speaker Bio:
Kevin Kane is a Software Development Engineer

in the Security
Incubations team in the CSRO Incubations group at Micro

soft.
His research interests include security issues in distributed en

viron-
ments including grid computing and virtualized environments. He

received his B.S. degree from the University of Maryland in 2000

and his M.S. and Ph.D. degrees from the University of Texas at
Austin i

n 2005 and 2006 respectively where his research focused
on access con

trol in decentralized distributed systems under the
supervision of Pro

f. James Browne.

Brian A. LaMacchia Bal to his friends is one of a h

andful of applied cryptographers at Microsoft. As Software Architect and G

roup Mgr.
for Security Incubations within the Office of the Chief Resear

ch and
Strategy Officer Brian’s current focus is on cryptographi

c applications
and security architectures for next-generation manycore

-friendly client
desktop platforms. Brian is also a founding member of
the Microsoft
Cryptography Review Board and consults on security and cr

yptography
architectures protocols and implementations across the compa

ny.
Before moving to CRSO Incubations Brian was the architect for cryp

to-
graphy in the Windows Security group. Prior positions Brian has hel

d at
Microsoft include Development Lead for .NET Framework Security and

Program Manager for core cryptography in Windows 2000. In addition to his responsibilities at Microsoft Brian is an affiliate faculty membe

r of
the Department of Computer Science and Engineering at the Universit

y
of Washington. Brian received S.B. S.M. and Ph.D. degrees in Elect

rical
Engineering and Computer Science from MIT in 1990 1991 and 1996

respectively.