PA200
Overview of contemporary public
cloud providers
RNDr. David Gešvindr
MVP: Data Platform | MCSE: Data Platform | MCT
gesvindr@fi.muni.cz
@gesvindr
Outline
1. Introduction to Public Cloud
2. Compute Services
3. Storage Services
4. Monitoring of applications
5. Other Services
Outline
1. Introduction to Public Cloud
2. Compute Services
3. Storage Services
4. Monitoring of applications
5. Other Services
Cloud definition
„Cloud computing is a model for enabling ubiquitous, convenient, ondemand
network access to a shared pool of configurable computing
resources (e.g., networks, servers, storage, applications, and services)
that can be rapidly provisioned and released with minimal
management effort or service provider interaction.“
- National Institute of Standards and Technology
Key features of the cloud
▪ On-demand self service
▪ Broad network access
▪ Resource pooling
▪ Rapid elasticity
▪ Measured service
Service Models
▪ Software as a Service (SaaS)
▪ Platform as a Service (PaaS)
▪ Infrastructure as a Service (IaaS)
▪ Public cloud
▪ Private cloud
▪ Hybrid cloud
Deployment Models
Contemporary Cloud Providers
Amazon Web Services Microsoft Azure Google Cloud
Oracle Cloud IBM Cloud
Public Cloud Adoption
Amazon Web Services (AWS)
▪ Launched in July 2002 (major redesign in 2006)
▪ Initially Amazon was providing access to spare compute capacity in
their datacenters hosting their e-commerce systems
− Their goal was to have a fully automated management of all resources
− Their E-commerce solution was later redesigned to be hosted in AWS
▪ Offers over 100 services (IaaS, PaaS and SaaS)
▪ 22 geographic regions
▪ You can start using AWS Free Tier
Microsoft Azure
▪ Launched in February 2010
▪ Public cloud provided by Microsoft
▪ Flagship of Microsoft products and services
− Very intensive development of provided services with a goal to become
leader in cloud industry
− Offers 58 regions worldwide (https://azure.microsoft.com/en-us/regions/)
▪ Offers over 100 services (IaaS, PaaS and SaaS)
− Great tooling support for developers (not only Microsoft/.NET developers)
▪ You can start using Azure for Students
− https://azure.microsoft.com/students
Google Cloud
▪ Launched in April 2008
▪ Google Cloud unites all Google Cloud Services
− Google Cloud Platform is a public cloud provided by Google
▪ 22 geographic regions
▪ Offers over 100 services (IaaS, PaaS and SaaS)
▪ You can start using Google Cloud Platform Free Tier
− https://cloud.google.com/free
Outline
1. Introduction to Public Cloud
2. Compute Services
3. Storage Services
4. Monitoring of applications
5. Other Services
Virtual Machines (IaaS)
▪ IaaS service which allocates a virtual machine in cloud provider‘s
datacenter
▪ Provides a full access to a virtual machine with Windows or Linux OS
preinstalled
▪ It is important to select type of VM based in application‘s
requirements
− Different ratio CPU/RAM, performance of storage, GPU availability
− General Purpose / Compute / Memory / Storage Optimized
▪ Services:
− Azure Virtual Machine + Azure Storage Managed Disk
− Amazon Elastic Compute Cloud (Amazon EC2) + Amazon Elastic Block Store (EBS)
− Google Compute Engine + Google Persistent Disk
Managed application hosting (PaaS)
▪ Web/application servers managed by the cloud provider (PaaS)
▪ Developer selects required runtime (.NET, PHP, Java, NodeJS, etc.)
and then deploys its application using supported deployment
method
− Common support for Continuous Integration and Deployment
▪ Support for vertical and horizontal scalability (autoscale)
▪ Services:
− Azure App Service
− AWS Elastic Beanstalk
− Google App Engine
Managed Containers (PaaS)
▪ Cloud provier provisions and manages compute infrastructure for
deployment and operations of applications in containers
▪ Developer creates a Docker container with application and stores it
in a container registry
▪ The service downloads the application container and runs it
▪ Services:
− Azure Container Instances, Azure Kubernetes Service
− Amazon Elastic Container Service, Amazon Elastic Kubernetes Service
− Google Cloud Run
Serverless Computing (PaaS/FaaS)
▪ Function as a Service (FaaS) is a new trend, when developers
deploy not applications but functions (methods with code) that are
triggered based in selected trigger
▪ The function is automatically hosted and also scaled
▪ Commonly billed based on consumed CPU time and memory
▪ Supported triggers: HTTP call, time, message in queue, new file
▪ Services:
− Azure Functions
− AWS Lambda
− Google Cloud Functions
Outline
1. Introduction to Public Cloud
2. Compute Services
3. Storage Services
4. Monitoring of applications
5. Other Services
Object / Blob storage (PaaS)
▪ Highly scalable object/blob storage service
▪ Commonly used as a primary storage service for data
▪ Supported types of data
− Files (blobs) with block or random access, Data Lakes
− Messages in a queue
− Records in a table (key-value storage, no SQL access)
▪ Services:
− Azure Storage
− Amazon Simple Storage Service (Amazon S3), Amazon Simple Queue Service
− Google Cloud Storage
Relational database (PaaS)
▪ Relational database engine hosted as a service (fully managed by
the cloud provider offering high availability, managed backups, etc.)
▪ Commonly used database engines are offered as a service by cloud
providers
▪ New database services are built by the cloud providers to achieve
high scalability
▪ Services:
− Azure SQL Database (Microsoft SQL Server, PostgreSQL, MySQL, MariaDB)
− Amazon RDS, Amazon Aurora
− Google Cloud SQL
NoSQL database
▪ Various NoSQL databases provided as a managed service (PaaS):
− Document Database (Mongo DB)
− Column-family (Apache Cassandra)
− Key-value (Redis)
− Graph (Gremlin)
▪ Services:
− Azure CosmosDB, Azure Storage, Azure Cache for Redis
− Amazon DynamoDB, Amazon DocumentDB, Amazon Neptune, Amazon
Timestream, Amazon Managed Apache Cassandra Service
− Google Cloud Bigtable, Google Cloud Firestore, Google Memorystore
Big Data analytics
▪ Data are stored in Data Lakes (eg. CSV files, NoSQL databases)
▪ Analytical Services can connect to these data lakes and run
analytical queries, computations and machine learning
− Apache Hadoop, Apache Spark
▪ Services:
− Azure Storage, Azure Synapse Analytics, Azure Databricks
− Amazon S3, Amazon Athena, Amazon EMR
− Google Cloud Storage, Google Cloud Bigtable, Google BigQuery, Google
Cloud Dataproc
Outline
1. Introduction to Public Cloud
2. Compute Services
3. Storage Services
4. Monitoring of applications
5. Other Services
Monitoring of applications
▪ Cloud applications should be actively monitored not just by basic
tools that are part of the hosting service
▪ Additional services can be used for collection of exceptions,
outages and their intelligent analytics
▪ Services:
− Azure Application Insights
− Amazon CloudWatch
− Google Cloud Logging / Monitoring / Trace / Profiler…
Outline
1. Introduction to Public Cloud
2. Compute Services
3. Storage Services
4. Monitoring of applications
5. Other Services
Artificial Intelligence
▪ Services providing a simple API to access advanced AI models and
services to:
− Analyze images
− Analyze texts
− Analyze anomalies
▪ Services:
− Azure Cognitive Services
− Amazon ML Services, Amazon AI Services
− Google AI Hub, Vision AI, Video AI, Natural Language, Translation, AutoML
Questions
RNDr. David Gešvindr
MVP: Data Platform | MCSE: Data Platform | MCT
gesvindr@fi.muni.cz
@gesvindr