Teaching Cloud Computing and Windows
Azure in Academia
Domenico Talia
UNIVERSITA’ DELLA CALABRIA & ICAR-CNR
Italy
[email protected]
Faculty Days 2010 – September 16, 2010 - ROME
1
Sommario
• Obiettivi del Corso
• Struttura e Contenuti
• Concetti di Base e Sistemi
• Sistemi Cloud Commerciali e Open Source
• Windows Azure
• Alcune attività di Ricerca all’UNICAL
2
Obiettivi del Corso
• Il materiale didattico é principalmente volto a
favorire l'introduzione dei concetti di base e le
architetture dei sistemi Cloud.
• Alcuni corsi che potrebbero includere il materiale
didattico:
•
•
•
•
Cloud Computing
Sistemi Paralleli/Calcolo Parallelo
Sistemi Distribuiti
Modelli e Architetture per il Web
3
Contenuti del Corso
Argomento
1. Cloud computing: definizioni e concetti
Durata
2h
2. Web services, Grid e Cloud
2-3 h
3. Modelli service-oriented di Cloud computing
2-3 h
4. Sistemi Cloud commerciali e open source
2-3 h
5. Il sistema Azure: architetture e servizi
2h
4
Contenuti del Corso
• Il materiale offre una introduzione generale a tutti gli
argomenti trattati e poi descrive con maggior dettaglio i
concetti principali del Cloud computing e i dettagli tecnici ed
architetturali dei sistemi descritti.
• Come possibili progetti di fine corso il docente può creare uno
o più progetti basati
• sull’uso di sistemi Cloud commerciali come Azure, Google o Amazon il
cui costo di accesso ed utilizzo è molto limitato,
• oppure basati sull’uso di uno o più dei sistemi Cloud open source che
sono scaricabili ed istallabili anche su computer di limitate dimensioni.
5
Lezione 1: Cloud Computing: Definizioni e Concetti
6
Lezione 2: Web Services, Grid e Cloud
7
Lezione 3: Modelli Service-Oriented per il Cloud
8
Lezione 4: Cloud Commerciali e Open Source
9
Lezione 5: Windows Azure: Architetture e Servizi
10
Materiale sul Web
https://www.facultyresourcecenter.com/curriculum/pfv.aspx?ID=8469&Login=&wa=wsignin1.0
11
Distributed Discovery Services
• Exploiting the SOA model it is possible to define basic services
for supporting distributed data mining tasks/ knowledge
discovery applications in large scale distributed systems for
science and industry (from a private Cloud to Interclouds).
• Those services can address all the aspects that must be
considered in data mining and in knowledge discovery
processes
• data selection and transport services,
• data analysis services,
• knowledge models representation services, and
• knowledge visualization services.
12
Collection of Services for Distributed
Knowledge Discovery
• It is possible to define services corresponding to
Data Mining Applications or KDD processes
This level includes the previous tasks and patterns composed
in a multi-step workflow.
Distributed Data Mining Patterns
This level implements, as services, patterns such as collective learning,
parallel classification and meta-learning models.
Single Data Mining Tasks
Here are included tasks such as classification, clustering, and association
rules discovery.
Single KDD Steps
All steps that compose a KDD process such as preprocessing,
filtering, and visualization are expressed as services.
13
Knowledge Discovery Services
• This collection of data mining services can constitute an
Open Service Framework for Grid-based Data Mining
Open Service Framework for Cloud-based Knowledge Discovery
• Allowing developers to program distributed KDD processes
as a composition of single and/or aggregated services
available over a Cloud.
• Those services should exploit other basic Cloud services for
data transfer, replica management, data integration and
querying.
14
Knowledge Discovery Cloud Services
• By exploiting the Cloud services features it is possible
to develop knowledge discovery services accessible
every time and everywhere (remotely and from small
devices).
• This approach may result in
•
•
•
•
Service-based distributed data mining applications
Data mining services for communities/virtual organizations.
Distributed data analysis services on demand.
A sort of knowledge discovery eco-system formed of a
large numbers of decentralized data analysis services.
15
Grazie
16
Scarica

Distributed Data Mining Patterns as Services Programming Issues in