Cloud Computing:

Computing as a Service
Prof. Raffaele Montella, Ph.D.
http://raffaelemontella.it raffaele.montella@uniparthenope.it
Outline - Computing in the Cloud
● Introduction

● Computing as a service

● Virtualization and Containerization

● Advanced services

● Serverless computing

● Conclusion
Introduction

● Infrastructure as a Service.
● Instant deployable computing power.
● There is no queue: the resources are allocated and
consumed:
○ More CPUs than in the user availability.
○ Expensive high-end GPGPU devices.
○ Diverse and different environments for product testing.

● Access computing power from anywhere, anytime, any device.

Compute as a Service

● CPUs, RAM, storage, GPUs, FPGAs on demand

● Elasticity

● Scalability

● Pay as you go
Compute as a Service

● Pay how much you can: make a bid

● Private clouds

● Public clouds

● Cloud burst
Virtualization

● Computing on the cloud means not using servers directly.

● A virtual machine hosts an operating system.

● The custom software runs on this machine.

● A virtual machine is a software image of a complete

machine that is loaded on the compute server.
Virtualization

● The compute server runs a software component called

“hypervisor”.

● The hypervisor grants the server resources to the virtual

machines.

● Different operating systems are packed of the hypervisor.

● The computing servers are managed in a homogeneous

way.
Virtualization

● Components:
○ Cloud Management System
○ Fabric Controller
○ Cloud Monitor

● An instance of a virtual machine is a running virtual

machine started from its image.
Virtualization

● Many instance of the same virtual machine on the same

(or different server).

● Instance recovery in case of crash.

● Instance health.

● Instance load.

● Balance between costs and performance.

Containers

● Are similar to virtual machines.

● Based on a different virtualization technology.

● Layered on top of the host operating system.

● Operating system level virtualization.

Containers

● More effective, instant deployment.

● An application is packaged in a container.

● All dependencies are packaged.

● Atomic unit simple to manage.

Containers

● Example:
○ Container 1: web server
○ Container 2: database server

● Persistent disk can be hosted on the physical machine.

● Example:
○ Package: Sweep parameters application.

● Multiple containers’ instances with different parameters.

Containers

● Lightweight.

● Applications start in an instant.

● A container shares libraries with other containers.

● On a given compute server:

○ more containers
○ less virtual machine instances
Containers

● Packaging an application in a container is a simple task.

● Container definition script language.
● Requirements:
○ Libraries
○ Dependencies
○ Source files
○ Data
● The script runs on a laptop for testing and on a remote
cloud compute server for production.
Containers

● Containers are portable between different clouds.

● It is possible, but not really easy move virtual machine

images from one cloud to another.

● Virtual machines take minutes to be up and running.

● Virtual machines are more isolated.

Containers

● Managing networking in containers is tricky.

● Security in containers could be an issue.

● More containers share the same host OS instance.

Virtual Machines vs Containers

Cloud Fabric Controller

Cloud Fabric Controller

App App

App

App

App

App

App
Bins & Bins &
Libs Libs

Docker
Bins & Bins &
Libs Libs
Guest OS Guest OS
Guest OS

Hypervisor Hypervisor

Server Server

● Containers are more cloud-native

● Virtual Machines are more secure.
Virtual Machines vs Containers
Virtual Machine Containers
Heavyweight Lightweight
Fully isolated Process-level isolation
No automation in configuration Automated configuration
Slow deployment Fast deployment
Easy networking Abstract networking
Custom images Portable
● Containers are more cloud-native
● Virtual Machines are more secure.
Advanced Computing Services

● Infrastructure as a Service is just the foundation of the

computing in the cloud.

● Cloud vendors (Amazon, Microsoft, Google) have

dedicated services.

● Data analysis clusters.

● Tools for massive streams of events.
● Machine learning tools.
Advanced Computing Services

● The problem of computing is the scaling.

● Virtual machines and containers make easy to virtualize a

single machine image.

● The most part of today applications are data massive.

● Run parallel tools is not the exception, but the regular

way.
Advanced Computing Services

● High performance computing using Message Passing

Interface (MPI)
● Virtual clusters (Amazon, Microsoft Azure)

● Many task parallelism (MTP) : massive number of

independent task to run in a embarassing parallel way.
Advanced Computing Services

● Bulk synchronous parallelism (BSP)

○ MapReduce (Apache Hadoop)
○ Spark (interactive using Jupyter)

● Cloud Datalab (Google) - based on Jupyter

● Cloud Business Intelligence (Microsoft Azure)
Advanced Computing Services

● Scaling containers:
○ Docker Swarm
○ Kubernetes container management (Google)

● High Throughput Condor (1985… still widely used, and

developed)
○ Class-AD - Resource specification language
○ Used in European grid computing for decades
alongside Globus 2 and GridFTP
Advanced Computing Services

● MesOS: distributed operating system with web interface.

● Dataflow and workflows:
○ Galaxy / Globus Galaxy: bioinformatics
https://www.globusgenomics.org
○ Galaxy / Globus Galaxy-ES: earth science
https://onlinelibrary.wiley.com/doi/abs/10.1002/cpe.354
0
○ DAGonStar: http://github.com/dagonstar
● More:
https://github.com/meirwah/awesome-workflow-engines
Serverless Computing

● Having up and running a virtual machine instance on the

cloud could be costly if the most part of the time is in idle.

● The effort needed to configure a service is not trivial.

● The security have to be enforced.

● The scalability is an issue.

Serverless Computing

● A service can be deployed and bounded to a web route

without the need to be hosted to a dedicated machine.

● The service is billed for the amount of the resources for

the time that are used.

● The services are hosted in lightweight containers.

● Serverless computing is cloud native/continuously

upgraded.
Conclusions

● Good things about hosting computation in the cloud:

○ Cost
○ Scalability
○ Access
○ Configurability
○ Variety
○ Security
○ Upgradability
○ Semplicity
Conclusions

● Bad things about hosting computation in the cloud:

○ Cost
○ Variety
○ Security
○ Dependence

● Private clouds
● Initial cost of investment
● Hybrid clouds
● Total cost of ownership
● Cloud burst
● Time to marked