AWS vs Azure vs Google (GCP) - Slides
- 3. Welcome to something a little different Continuation of our DevOps event series Very interactive format ! Lean Beer™ style Feedback very welcome ! Future topics might include: Sponsors Puppet vs Ansible Docker vs Packer Prometheus vs ELK ELK vs Splunk
- 5. ● Overview - 60 minutes ○ 20 - AWS ○ 20 - Azure ○ 20 - GCP ● Breakout into groups to discuss - 30 minutes ○ Positives ○ Negatives ○ Decisions ● Each group shares their key findings with the others - 30 minutes ● Vote & declare the winner, then beers. What are we going to do?
- 6. Pillars of Cloud Services Scalability Availability Resiliency Management Security Infrastructure Compute Storage Networking
- 7. AWS - STEVEN HARPER 20 Years of IT / Development / Operations / Design / Architecture ● Started AWS with a Digital Migration of Europe's Biggest E-Service Platform ● Designed and deployed Hybris Retail systems into AWS ● OAuth open connect V2.0 platform into AWS ● AWS Journey 2013 -> Current ● What’s completely new ○ Lambdas, Step Functions, Cloudformation
- 8. AWS Products
- 9. Regions & Availability Zones Frankfurt (3) Ireland (3) London (3) Paris (3) São Paulo (3) Gov (US-West) (3) N. Virginia (6) Ohio (3) N. California (3) Oregon (3) Mumbai (2) Seoul (2) Singapore (3) Sydney (3) Tokyo (4) Osaka-Local (1) Canada Central (2) Beijing (2) Ningxia (3) Bahrain Hong Kong SAR, China Sweden Gov (US-East)Over 120 Edge locations
- 10. Who uses AWS All Sizes of customers - over 1,000,000 customers
- 11. Security & Compliance CloudTrail - Compliance Audits Trusted Advisor - Real-time insight Enterprise Support - Technical Account Manager (TAM) KMS - Key Management Service Cloud HSM - Hardware Security Module
- 12. ● Managed Components ○ AMI’s ○ CloudFormation Stacks ● Flexible Licensing Models ○ BYOL ○ Pay per hour
- 13. Supports Many different Languages - Lambda ○ 4.3 ○ 6.10 ○ 8.10 ○ 2.7 ○ 3.6 ○ 1.0 ○ 2.0 ○ 2.1 + Triggered by events ○ 8 ○ 1.X Amazon S3 Amazon DynamoDB Amazon Kinesis Data Streams Amazon Simple Notification Service Amazon Simple Email Service Amazon Simple Queue Service Amazon Cognito AWS CloudFormation Amazon CloudWatch Logs Amazon CloudWatch Events AWS CodeCommit Amazon CloudWatch Events AWS Config Amazon Alexa Amazon Lex Amazon API Gateway AWS IoT Button Amazon CloudFront Amazon Kinesis Data Firehose Invoking a Lambda Function On Demand
- 14. Databases - Choices ● RDS ○ MariaDB, Amazon Aurora, Oracle, SQL Server, PostgreSQL, MySQL ● Non Relational ○ DynamoDB ● Graph ○ Amazon Neptune ● In Memory ○ AWS Elasticache ● Data Warehousing ○ Redshift ● Migration
- 15. Instance Families . Placement Groups - Allows reduced network distance . ● Burstable: T (10% capacity, burst to 100%) - Free Tier . ● General purpose: M . ● Compute optimized: C (CPU Focused) . ● Memory optimized: R, X, Z (RAM Focused) . ● Storage optimized: D, H ,I (IO Focused) . ● Accelerated computing: F, G, P (Hardware Acceleration Focused) . t3.nano (1 vCPU, 512 MiB) <------> u-12tb1.metal (448 vCPU, 12 TiB) .
- 16. VPC Separation & Low level network control ● Direct Connect ● VPN Connections ● VPN Peering ● VPC Separation ● Subnets (Private/Public) ● Security Groups ● NAT Gateways ● Virtual Private Gateways ● Network Access Control Lists ● Elastic IP’s ● Elastic Load Balancers
- 17. SQS, Kinesis, Firehose Simple Queue Service ● FIFO - only once ● Best Order - at least once Kinesis ● Massively, Scalable & Durable real-time data streaming service Firehose ● Delivery of kinesis into storage SNS, SES SES ● Email services SNS ● Topics and Subscriptions push notifications
- 18. Tooling & API’s AWS - CodeStar, CodeCommit, CodeBuild, CodeDeploy, CodePipeline Terraform / CloudFormation AWS Tools for Azure DevOps Ansible, Puppet, Chef AWS - CLI/SDK
- 19. Ease of Entry . AWS Training APN Partner Training AWS Certification AWS Academy AWS Educate
- 20. The Good & The Bad GOOD ● Capability to build anything ● Ease to automate ● Constantly innovating ● Language & OS agnostic ● Huge user base ● Great support BAD ● Small components, difficult to start ● Out of Free Tier, everything costs ● No Event Bus ● AI Offering & Machine learning poor compared to other offerings ● AWS Users / Server Roles
- 21. Azure - Sherin Mathew Current - Azure Analytics Architect, IBM. GBS Background - Telecom Application Dev and BI Dev, BI Programme Delivery, Project Management, Solution Architect, Principal Consultant. Experience - near 14 years in Telecom, Healthcare, ,Housing groups and Higher Education Sector along with Police forces and ICO. My Azure Journey - Last 4 years feel like 20! ➢ Healthcare - 2015-16 ➢ Higher Education 2016-18 ➢ Housing 2017-18 ➢ ICO 2017-2018 Volunteer - AI-Tech.UK, AI Tech User Group, AI Tech North 2019, SQL Pass @SMdisrupt “Disrupt the disruption!”
- 22. Azure More than 95 percent of Fortune 500 companies use Azure 120,000 New Azure Customers per month 30+ years Over three decades of experience serving enterprise customers 68,000+ partners 70+ certifications Cloud Leader since 2015
- 23. 54 Regions
- 24. Services Compute Web & Mobile Data & Storage Analytics Integration (Hybrid and Cloud ) Internet of Things Networking Hybrid Operations Security & Management Developer Services Machine Learning and AI Support
- 26. Azure Compute Virtual Machines Windows Virtual Machines Linux VMs Virtual Machine Scale Sets Azure Batch Azure Container Service Container Instances Azure Container Service Kubernetes Service Azure Container Registry Cloud Services Service Fabric Service Fabric Mesh Azure Functions Azure Batch AI DSVM
- 29. Azure Data Storage and Analytics ➢ Machine Learning* ➢ Data Lake Analytics* ➢ HDInsight ➢ Azure Databricks ➢ Analysis Services* ➢ Azure SQL Database ➢ Azure Data Warehouse* ➢ Azure Data Lake Store ➢ Azure Storage ➢ Azure Cosmos DB* ➢ Azure Data Factory* ➢ Event Hub ➢ Event Grid ➢ IoT hub ➢ Stream Analytics ➢ Data Catalog ➢ Data Explorer*
- 30. Azure Analytics in Action
- 31. Azure Networking ➢ Virtual Network ➢ Load Balancer ➢ Azure DNS ➢ Traffic Manager ➢ Data Gateway ➢ VPN Gateway ➢ ExpressRoute ➢ Application Gateway ➢ Network Watcher ➢ CDN
- 32. Azure Hybrid Integration Azure Managed Instance * DMA/DMS Azure Backup Azure Recovery Azure Storsimple Azure Sync Azure Data box* Azure Gateways AD Connect Containers Service Azure Stack
- 33. Machine Learning and AI Tools ➢ ML Studio ➢ VS Code ➢ ML.Net ➢ Azure ML Packages ➢ AI Toolkit for IoT ➢ MML Spark Frameworks ➢ CNTK ➢ ONNX ➢ TensorFlow ➢ Pytorch ➢ MXNet ➢ Caffe2 Infrastructure ➢ Azure Batch AI / DSVM ➢ AKS ➢ ADLS ➢ Databricks ➢ Spark for Azure HDinsight ➢ Edge ➢ SQL
- 34. Azure Advanced Analytics Pattern
- 35. Azure Dev and Ops Support ➢ Visual Studio Code ➢ Azure SDK ➢ Command-line tools ➢ Azure PowerShell ➢ Azure Command-Line Interface (Azure CLI) ➢ Storage Explorer ➢ Visual Studio Tools for Azure ➢ Visual Studio for Mac ➢ Docker Tools ➢ Azure Portal! ➢ DevOps ➢ OMS and log analytics Build cloud apps for Java, Node, Python, Ruby, PHP or .NET T-SQL!
- 36. Security and Compliance ➢ Azure AD ➢ Subscription Security - IAM and RBAC ➢ Operational Security - Advanced Threat Detection, Auditing and Logging, OMS ➢ Network Security ➢ Compute Security ➢ Key Security ➢ Data Security ➢ Identity and Information Protection ➢ Security Centre ➢ 70+ Global, Government, Industry and Regional Compliances inc GDPR
- 37. Azure - What's Different Regions Nature of Integration Hybrid Development platform Reporting platform Support TCO PaaS Power Easier - Everything Documentation Community IP Ethos and Philosophy
- 38. Azure - Good and Bad Positives ● True Trusted Enterprise ● Dominance in Paas, ML, Analytics, IoT, PowerBI ● Maturity - Cloud Leader since 2015 ● Great Dev and Test support and tools ● Works well in Windows or Microsoft environment ● Agility and commitment ● Great knowledgeable support ● 54 regions and Data box! ● Clear consideration for .NET. ● Open Source integration ● Hybrid consistency Negatives ● Maturity in Prod and Enterprise is low ● Cross platform ● Devops for some services can be tricky ● Free and Preview services ● Quick Releases, changes and updates
- 39. Google Cloud Platform - James Heggs Tech Returners - https://techreturners.com Background 10+ years Software Development SaaS Discovered the world of DevOps for improving delivery of our operations One of those ‘ooh shiny new things’ types of Techies Experience Software development largely in web application and SaaS platforms Migration from on-premise virtualisation to cloud based environments Big fan of containerisation and Kubernetes orchestration Very big fan of Continuous deployment and deployment pipelining CNCF/Open source adoption tooling and infrastructure automation GCP Journey Used in anger only really the past 6-12 months Decided to try a largely PaaS based approach for the construction of software for a startup
- 40. 40The Products logos contained in this icon library may be used freely and without permission to accurately reference Google's technology and tools, for instance in books or architecture diagrams. Google Cloud Platform Icons Compute Big Data BigQuery Cloud Dataflow Cloud Dataproc Cloud Datalab Cloud Pub/Sub Genomics Cloud AI Cloud Machine Learning Engine Cloud Vision API Cloud Speech-to-Text Cloud Natural Language API Cloud Translation API Cloud Jobs API Cloud Dataprep Cloud Video Intelligence API Advanced Solutions Lab Compute Engine App Engine Kubernetes Engine GPU Cloud Functions Container- Optimized OS Identity & Security Cloud IAM Cloud Resource Manager Cloud Security Scanner Key Management Service BeyondCorp Data Loss Prevention API Identity-Aware Proxy Security Key Enforcement Cloud AutoML Cloud Text-to-Speech Cloud TPU Dialogflow Enterprise Edition Data Transfer Transfer Appliance Cloud Composer Cloud Security Command Center
- 41. 41The Products logos contained in this icon library may be used freely and without permission to accurately reference Google's technology and tools, for instance in books or architecture diagrams. Networking Virtual Private Cloud Cloud Load Balancing Cloud CDN Dedicated Interconnect Cloud DNS Cloud Network Cloud External IP Addresses Cloud Firewall Rules Cloud Routes Cloud VPN Management Tools Stackdriver Monitoring Logging Error Reporting Trace Debugger Cloud Deployment Manager Cloud Console Cloud Shell Cloud Mobile App Cloud Billing API Cloud APIs Cloud Router Partner Interconnect Cloud Armor Standard Network Tier Premium Network Tier Profiler Google Cloud Platform Icons
- 42. 42The Products logos contained in this icon library may be used freely and without permission to accurately reference Google's technology and tools, for instance in books or architecture diagrams. Developer Tools Cloud SDK Cloud Source Repositories Maven App Engine Plugin Cloud Tools for IntelliJ Cloud Tools for PowerShell Cloud Tools for Visual Studio Container Registry Cloud Tools for Eclipse Cloud Build API Platform & Ecosystems API Analytics API Monetization Apigee API Platform Apigee Sense Cloud Endpoints Developer Portal Gradle App Engine Plugin IDE plugins Internet of Things Cloud IoT Core Storage & Databases Cloud Storage Cloud Bigtable Cloud Datastore Cloud SQL Cloud Spanner Persistent Disk Cloud Memorystore Google Cloud Platform Icons Cloud Filestore Cloud Test Lab Cloud IoT Edge
- 44. Network
- 45. Where is GCP going Totally ripped off this slide - Yoink! Thanks Steven <-- No Worries
- 46. Pricing
- 47. Always Free
- 48. Google Cloud Platform - Machine Learning and AI ● Developer ready API’s ○ Vision API - Image Content Analysis ○ Cloud speech to text ○ Cloud text to speech ○ Cloud natural language processing ● Firebase Predictions API ○ https://firebase.google.com/products/predictions/ ● DialogFlow ○ ChatBot (Conversational Interfaces and leverages Google ML) ○ Google Cloud Functions Ready ○ Multi Language Support ● More control - Cloud Machine Learning Engine or Cloud TPU’s ○ TensorFlow Model Creation ○ Access to high throughput/spec machines
- 49. Google Cloud Platform - Machine Learning - Cloud TPU
- 50. Google Cloud Platform - Databases ● Cloud SQL ○ MySQL or Postgres ● BigTable ○ Low latency ○ Useful for AdTech (instant response) ○ NoSQL ● Cloud Spanner ○ Mission Critical Relational Database ○ Transactional ○ Powers Google Search ● Lots of others like: ○ Cloud Datastore ○ Cloud Firestore ○ Firebase Realtime Database
- 51. Google Cloud Platform - Databases
- 52. Google Cloud Platform - Serverless ● https://cloud.google.com/serverless/ ● Cloud Functions ○ Less language coverage than AWS Lambda ○ Only recently out of Beta ● App Engine ○ Essentially bundle your apps (EG. Express) automatically in to Docker containers and deploy ● Cloud Pub/Sub ○ Like SNS - Message Architecture - Zero Config ● Kubernetes Engine Serverless AddOn ○ Run your own Kubernetes cluster whilst providing ability to deploy serverless apps to it ● Cloud Build ○ Serverless build and deployment pipelines
- 53. Google Cloud Platform - Kubernetes - OOTB + On-Prem
- 54. Google Cloud Platform - The Good and The Bad Good Bad GCP Product Strategy GCP Product Strategy Developer ready - PaaS, Serverless, Firebase Cloud functions language coverage Machine Deep Learning made (slightly) accessible Less documentation of lower level config like VPC, Security Serverless Offering Growing IAM and Service Accounts
- 55. Personal experiences ● Which services have you had good experiences with ● Where have you found gaps / poor experiences Please spread Azure & Google experience around the groups - use colour codes !
- 56. ● Everyone writes down what they want to talk about, writes questions they have, or ideas they want to discuss on a post-it and sticks it down ● A volunteer de-duplicates the post-its for a group ● Everybody puts 3 dots on the post-its they most want to talk about ● Start a 5-minute timer and start talking about the most-voted-for post-it, starting with the person who wrote the post-it ● When the timer beeps, everyone votes on whether they want another 2 mins ● If it’s time to move on, pick up the next most-voted-for post-it ● At the end of the session, each group decides what they’d like to share with everyone Lean Coffee
- 57. TIME TO VOTE Open in a browser: https://kahoot.it/ Prepare to enter an 8 digit PIN (number) Enter your name Wait…… Vote!