Transparents

Transcription

Transparents

:
nouveaux problèmes, nouvelles solutions
Hélène Renard,
I3S - UMR 6070 UNSA / CNRS -
École polytechnique universitaire de Nice-Sophia Antipolis
23 février 2011 - Journée “Des grilles aux clouds”
Modalis
Journée
Des grilles aux clouds
Plan de la présentation
1. Évolution ou révolution ? (Thierry Priol)
Virtuous side
Dark side
Conclusion
2. Sky Computing on FutureGrid and Grid’5000 with Nimbus
(Pierre Riteau)
Introduction to Sky Computing
The Nimbus Project
Large-Scale Sky Computing Experiments
Conclusion
3. Des grilles vertes aux nuages verts : vers des systèmes efficaces en énergie (Laurent Lefèvre)
4. A Telecommunications Operator View on Cloud Computing
(Thierry Coupaye)
5. Questions éventuelles
Hélène Renard
2
Évolution ou révolution ?
Virtuous side
Dark side
Conclusion
(Pierre Riteau)
The Nimbus Project
Conclusion
(Thierry Coupaye)
Hélène Renard
3
Virtuous side
Grilles vs Clouds
Définition de la grille
A fully distributed, dynamically reconfigurable, scalable and
autonomous infrastructure to provide location independent, pervasive,
reliable, secure and efficient access to a coordinated set of services
encapsulating and virtualizing resources (computing power, storage,
instruments, data, etc.) in order to generate knowledge... from the
CoreGRID NoE
Hélène Renard
4
Virtuous side
Grilles vs Clouds
Définition du cloud
An emerging computing paradigm where applications, data and
infrastructures are provided as a service that can be ubiquitously
accessed from any connected devices over the internet.
Hélène Renard
5
Virtuous side
What is behind Cloud
Google cluster 1997
_
_
Datacenters as the reincarnation of the mainframe concept
The end of the PC/Mac era ?
Just a web browser is needed
The network is the computer,
Hélène Renard
thin client, ...
6
Virtuous side
Why Cloud now and not before ?
_
Internet !
_
PC is not anymore the central device for personal computing
_
Network performance has been improved dramatically the last 15
years
Nearly always connected to the Internet (anytime, anywhere)
MP3, SmartPhone, Tablets, Set-top box, PCs, ...
How to get access to my personal data anywhere/anytime and from
any devices ?
Cost
Oversized systems to meet peak demand (both in the private and
public sector)
Outsourcing (labor cost is much higher that computing cost)
Hélène Renard
7
Virtuous side
Computing as a utility : a brief history
Hélène Renard
8
Virtuous side
Cloud Acronyms
_
PaaS - Platform/People as a Service
_
SaaS - Software/Search as a Service
_
IaaS - Infrastructure as a Service
_
DaaS - Data as a Service
_
CaaS - (composition/communication
/composite) as a Service
_
HaaS - Human as a Service ... just your
shared agenda ;-)
_
KaaS - Knowledge as a Service
_
...
Hélène Renard
9
Virtuous side
Cloud : how to escape from the jungle
Hélène Renard
10
Virtuous side
Infrastructure as a Service
_
Get access on demand to a large number of highly virtualized resources
_
Dynamicity, elasticity.
Concept of OS Virtualization
OS does not matter anymore !
OS are just software libraries and does not play a central role !
Concept of virtual machines to host instances of OS ;
Physical resources are shared by several virtual machines.
Hélène Renard
11
Virtuous side
Let’s take an example... Amazon !
Hélène Renard
12
Virtuous side
Platform as a Service
_
_
_
_
An application development, deployment and management fabric.
User programs web service front end and computational & Data
Services
Framework manages deployment and scale out
No need to manage VM images
Hélène Renard
13
Virtuous side
Software as a Service
Hélène Renard
14
Dark side
What is the dark side of Cloud computing ?
We have only seen the virtuous side !
What is the dark side of Cloud Computing ?
Hélène Renard
15
Dark side
Some research issues with Cloud Computing
_
Reliability / Resilience / Fault-tolerance ;
_
Trust, Security and Privacy ;
_
New economical models for computing ;
_
Service Level Agreement / Quality of Service ;
_
Building cloud-aware applications from legacy applications ;
_
Energy management ;
_
Data management ;
_
Cloud federation ;
_
Autonomic behaviors / Self-* ;
_
Brokering / Scheduling ;
_
Programming models (MapReduce, ...) ;
Interactions between legal aspects (laws) and computer science :
_
privacy and liability
Hélène Renard
16
Dark side
Are Cloud infrastructures less secure than
non-Cloud ones ?
_ One
of the fastest and easiest ways to access corporate data is through unprotected
PDAs that are lost or stolen, as they contain business names and addresses, spreadsheets and other corporate documents http://www.theregister.co.uk/2004/09/01/
pda sec
_ 60%
of corporate data resides unprotected on PC desktops and laptops (IDC analyst
Cynthia Doyle, Business Continuity in 2002 : It’s Not Business as Usual, April 2002)
_
Read from
10%
90%
49%
57%
80%
73%
_
66% of USB thumb drive owners report losing them, over 60% with private corporate
data on them !
http://www.nationalpost.com/
of laptop computers will be stolen within the first 12 months of purchase.
of stolen laptops are never recovered.
of companies have had laptops stolen with the last 12 months.
of corporate crimes are linked to stolen laptops.
of computer crime consists of ”inside jobs” by disgruntled employees.
of companies had no specific security policies for their laptops in 2003.
Hélène Renard
17
Conclusion
Conclusion
_
Cloud is becoming a buzzword... a lot of hype around it :
_
Not the swiss knife for distributed computing ;
More an evolution than a revolution ;
Less ambitious than Grid but there is an increasing public and business demand.
But there are new opportunities for research :
Reliability / Resilience / Fault-tolerance ;
Trust, Security and Privacy ;
New economical models for computing ;
Service Level Agreement / Quality of Service ;
Building cloud-aware applications from legacy applications ;
Energy management ;
Cloud federation ;
Autonomic behaviors / Self-* ;
Brokering / Scheduling (performance, energy, ...) ;
Programming models (MapReduce, ...) ;
Interactions between legal aspects (laws) and computer science.
Hélène Renard
18
Sky Computing on FutureGrid and Grid’5000 with Nimbus
Virtuous side
Dark side
Conclusion
(Pierre Riteau)
The Nimbus Project
Conclusion
(Thierry Coupaye)
Hélène Renard
19
Outline
_
_
The Nimbus Project
_
_
Conclusion
Hélène Renard
20
Infrastructure-as-a-Service
_
Basic features
_
All operations accessible through an API
_
Autonomic infrastructure management
Business model
_
Run VM from VM image
Modify + save VM image
Terminate VM
Pay for CPU time + network traffic in/out + storage
Initially targeting web service hosting
Also HPC now (Amazon Cluster Compute instances)
Hélène Renard
21
Sky Computing
_
Federation of multiple clouds
_
Creates large scale infrastructures
_
Allows to run software requiring large computational power
Hélène Renard
22
The Nimbus Project
The Nimbus Project
_
_
Started in 2005 by Kate Keahey (Argonne/UC)
Cloud computing Toolkit
_
_
_
Open source IaaS implementation
Amazon EC2-compatible (WSDL and Query APIs)
Cumulus (Amazon S3-compatible storage cloud)
Targets Clouds for Science
Unique features
Context Broker
Workspace Pilot
LANTorrent
Spot instances
Hélène Renard
23
The Nimbus Project
The Nimbus Architecture
Hélène Renard
24
The Nimbus Project
Context Broker
_
Service to configure a complete cluster with different roles
_
Works with a cluster distributed on multiple clouds (e.g. Nimbus
and Amazon EC2)
VMs contact the context broker to
_
_
Learn their role
Learn about other VMs in the cluster
Ex. : Hadoop master + Hadoop slaves
Hadoop slaves configured to contact the master
Hadoop master configured to know the slaves
Hélène Renard
25
_
Nimbus
_
ViNe
_
Resource management
Contextualization
All-to-all connectivity
Hadoop
Task distribution
Fault tolerance
Resource dynamicity
Hélène Renard
26
ViNe
_
Project of the University of Florida (M. Tsugawa et al.)
_
High performance virtual network
_
All-to-all connectivity
Hélène Renard
27
Hadoop
_
Open-source MapReduce implementation
_
Heavy industrial use (Yahoo, Facebook...)
_
Efficient framework for distribution of tasks
_
Built-in fault-tolerance
_
Distributed file system (HDFS)
Hélène Renard
28
Grid’5000 Overview
_
Distributed over 9 sites in France
_
1500 nodes, 5500 CPUs
_
Study of large scale parallel/distributed systems
Features
_
Highly reconfigurable
_
Environment deployment over bare hardware
_
Can deploy many different Linux distributions
Even other OS such as FreeBSD
_
_
Controlable
Monitorable (metrics access)
Experiments on all layers
Network, OS, middleware, applications
Hélène Renard
29
FutureGrid : a Grid Testbed
_
NSF-funded experimental testbed
_
5000 cores
_
6 sites connected by a private network
Hélène Renard
30
Resources used in Sky Computing
Experiments
_
3 FutureGrid sites (US) with Nimbus installations
_
Grid’5000 sites (France)
_
UCSD (San Diego)
UF (Florida)
UC (Chicago)
Lille (contains a white-listed gateway to FutureGrid)
Rennes, Sophia, Nancy, etc.
Grid’5000 is fully isolated from the Internet
One machine white-listed to access FutureGrid
ViNe queue VR (Virtual Router) for other sites
Hélène Renard
31
Experiment scenario
_
Hadoop shy virtual cluster already running in FutureGrid
(SD, UF, UC)
_
Launch BLAST MapReduce job
Start VMs on Grid’5000 resources
_
_
Automatically extend the Hadoop cluster
_
_
With contextualization to join the existing cluster
Number of nodes increases
TaskTracker nodes (Map/Reduce tasks execution)
DataNode nodes (HDFS storage)
Hadoop starts distributing tasks in Grid’5000
Job completes faster !
Hélène Renard
32
Conclusion
Conclusion
_
_
Sky Computing to create large scale distributed infrastructures
Our approach relies on
_
Nimbus for resource management, contextualization and fast cluster
instantiation
ViNe for all-to-all connectivity
Hadoop for dynamic cluster extension
Provides both infrastructure and application elasticity
Hélène Renard
33
Des grilles vertes aux nuages verts
Virtuous side
Dark side
Conclusion
(Pierre Riteau)
The Nimbus Project
Conclusion
(Thierry Coupaye)
Hélène Renard
34
Energy : 1st challenge for large scale
systems ?
_
_
Future exascale platforms -> systems from 20 to 100MW (current
4-6 MW)
How to build such systems and make them energy sustainable /
responsible ?
Hardware can help (component by component)
Software must be adapted to be scalable but also more energy efficient
Usage must be energy aware
Hélène Renard
35
Power demand and Green IT explosion
_
IT : 2-5% of CO2 emissions
_
Green It -> reducing electrical consumption of IT equipments - CO2
impact depends on countries
_
Focus on usage : fighting un-used/over- provisioned plugged resources
_
Problem : grey energy (ecoinfo.org)
_
GreenIT scientific events (12/2010 - 15/2011 + tracks/issues greenitconferences.org)
Hélène Renard
36
Green-IT Leverages
_
Shutdown : reducing the amount of powered unused resources
_
Slowdown : adapting the speed of resources to real usage
_
Optimizing : improving hardware and software for energy reduction
purpose
_
Coordinating : using large scale approaches to enhance green leverages
Hélène Renard
37
Methodology
_
Proposing a generic energy aware model able to be derivated onto
different scenario (Grids, Clouds, Networks)
_
Designing software solutions for infrastructures
_
Simulating and Validating at medium and large scale
Hélène Renard
38
Conclusion
_
At the beginning, the Grid was “green” (cycle stealing, old machines
usage...) like the Cloud (aggregation)
_
Energy aspects change the way we design Grids&Clouds applications, protocols, services and policies (i.e. load balancing is not
always the best solution)
_
Challenge : design energy proportional equipments and frameworks
(computing, memory or network usage)
Hélène Renard
39
A Telecommunications Operator View
Virtuous side
Dark side
Conclusion
(Pierre Riteau)
The Nimbus Project
Conclusion
(Thierry Coupaye)
Hélène Renard
40
Agenda
_
An industrial view on cloud computing
_
Orange & cloud computing
_
A glimpse of the future at Orange Labs
Drivers and barriers associated to
cloud computing
Stakes & current offers
The open cloud, the personal cloud, cloud networking...
_
The last word, cloud and grid ?
Hélène Renard
41
Drivers for users
_
Cost savings
_
Simplicity, Faster Time-to-market
_
Easy service experimentation before
service launch
Faster deployment, no need for servers
and software to launch a service
adFlexibility, Scalability
_
Reduced TCO : hardware, software, IT
staff
Pay-per-use (pay as you go), multitenancy
Automatic, transparent scale up and down
Improved availability and QoS
Although well-known breakdowns, cloud services have statistically
better availability than on site services... and they are improving
Hélène Renard
42
Barriers
_
Lack of technical maturity
_
Major risks of lock-in lock-in
_
_
Lack of standards (API, programming models)
Lack interoperability
Lack of portability (applications,mgt tools)
Legal Issues
_
SLA, auto-scaling/auto-sharing, performance, availability, dependability
PaaS : applications packaging, deployment, management, test, configuration management...
Storage
Network
Security, privacy withmultitenancy
Software licences
Data location (eg government, health)
Integration with legacy IT (IS)
Huge investments in data centers
building, hardware, cooling, energy
Hélène Renard
43
France Télécom / Orange
Hélène Renard
44
The Future of Cloud @ Orange Labs
Hélène Renard
45
The last word : “scalability”
_
Sclability in mainstream cloud is currently more
perceived :
_
as hundreds of thousand applications/users each
having a few virtual machines
than about few applications having each thousands of virtual machines
What about then :
Grid, HPC & cloud
Cloud for e-science ? Cloud market for e-science ?
Hélène Renard
46
Questions éventuelles
Virtuous side
Dark side
Conclusion
(Pierre Riteau)
The Nimbus Project
Conclusion
(Thierry Coupaye)
Hélène Renard
47
Le pire, quand on fait un discours, ce n’est pas de prendre conscience
que vos auditeurs regardent leur montre, c’est le moment où ils se
mettent à la secouer pour voir si elle n’est pas arrêtée. C’est à vous !
Hélène Renard
48

Transparents

Transcription

Documents pareils

Avec ces chaussures, il ne manque plus aux enfants qu`une

Cercle de la Terrasse - Oxford and Cambridge Club of Geneva

Power of Attorney - Ministère de l`Éducation et de l`Enseignement

[Descartes e il Seicento 5] Jean-Robert Armogathe, Vincent Carraud

Revue de presse ANGLAIS Semaine du 13 au 19 Février 2012

4. L` animal a... Le poisson a . Le cochon d`Inde a . L

4228 Holland Dr Des Moines, IA 50310 | Home for Sale | Real

Imagine you are working for the Scottish Tourist Board and you are

Nouveau feuillet T5008/ RL-18 Aux Courtiers ayant choisi Banque

Water Needs - Anglophonie