Skip to content

devops analysis

maefranco edited this page Nov 20, 2017 · 10 revisions

About

An analysis of several cloud environment options for deployment of the toro-net website.

Scores

Metric AWS Azure DigitalOcean Google App Engine Heroku IBM Cloud
Support 5.0 4.5 2.5 TBD 3.0 3.0
Features 5.0 3.5 4.0 TBD 3.5 4.0
Uptime TBD 2.0 5.0 TBD TBD TBD
Value / Cost 3.0 3.0 3.0 4.0 3.0 3.0
DB Style 4.0 2.0 3.0 TBD 3.0 4.0
Storage 3.5 4.0 4.0 3.0 3.0 4.0
UI / Usability 4.0 5.0 4.0 4.0 4.0 3.0
Migration TBD 4.5 TBD 2.0 TBD 4.0
Deployment 4.0 4.0 TBD 3.0 5.0 4.0
Average TBD 3.6 TBD TBD TBD TBD

Details

AWS

Metric Notes
Support Great support
Supported platforms .NET, Ruby, NodeJS, Go, Docker, PHP, Python
Features Various deployment options and ability to roll back to previous version; Full control over app health or servers changes with email notifications; Quick restart of app servers with a single command; Adaptive load balancing spreads out incoming app traffic across instances automatically; Automatic scaling of your web application based on its particular needs and defined conditons.
Cost Free for small projects to start with; As part of the AWS Free Usage Tier, you can get started with Amazon S3 for free. Upon sign-up, new AWS customers receive 5 GB of Amazon S3 storage in the Standard Storage class, 20,000 Get Requests, 2,000 Put Requests, and 15 GB of data transfer out each month for one year.
Target clients Startups, Medium Businesses, Large Enterprise
DB style RDS, RedShift, SimpleDB, DynamoDB
Storage First 50 TB / month $0.023 per GB $0.0125 per GB $0.004 per GB
Scalability Scale and flexibility are bound by the size of budget
UI / Usability Really good and simple console
Deployment Can use the automated deployment, or roll your own

Azure

Metric Notes
Support 33,378 stackoverflow questions, 1 conference per year, 100 webinars and podcast per year, holds 9% of the cloud computing market.
Features Marketplace has large number of items providing features ranging from virtual machine images, virtual machine extensions, APIs, applications, and Machine Learning services.
Uptime 10 issues documented this year, 39.77hrs downtime in 2017
Value Running 1VM at 3.5GB RAM and 50GB temp storage is $0.077/hr, $0.055 per write, $0.004 per read, $2.08 for 100GB storage. Usage bills per minute, SQL storage cheaper, noSQL DB pricing more complex and potentially expensive, possible discount for running linux.
DB style Supports 2 DB nativity, extensive support and features for SQL, only 2 noSQL engines, can run MongoDB but initially more complicated.
Storage type Supports restAPI, batch imports, has blob storage units 1TB in size.
Compute instances Supports 9 types of VMs, 52 instance types with 2843 disk images but only 2 Ubuntu 16.04 instances.
Security Achieves same security standard as AWS, has a security development life cycle, exposing security at every step, builds on OSA SQL security knowledge. Does not allow controlled access at the file level only folder
UI / Usability Good UI for developers, good for those that do not prefer the console.
Migration Provides tool and a road map for easy migration and optimization. https://www.youtube.com/watch?v=ZybZrdpWaKY&feature=youtu.be
Deployment Easy deployment, additional deployment support for windows technologies (can easily push for production using powershell or Visual studios), requires less involvement for DevOps, good for continuous deployment.

DigitalOcean

Metric Notes
Support In DigitalOcean site: 18,114 questions and 1761 tutorials; 72 meetup groups with 32,014 members; 2 upcoming meetups this month worldwide;Not much support outside DigitalOcean site: only about 2532 search results on stackoverflow.com
Features All-SSD cloud; uses standard HTTP requests and OAuth for authentication; 8 datacenter locations worldwide(2 US, 1 Canada, 3 Europe, 1 India, 1 Singapore); marketplace is limited but offers features such as MEAN deployment, machine learning, wordpress, LAMP
Uptime 99.99% Uptime SLA
Value $5/mo ($0.007/hr) for 512MB memory, 1 vCPU, 20GB SSD disk, 1TB transfer
DB style Can use MongoDB, MySQL
Storage type SSD-based; Increase storage from 1GB - 16TB

Google App Engine

Metric Notes
Service type http
Ease of use heavy on command line tool
Scalability Scalable; automatically add more capacity as workloads increase; same basic performance scalability as AWS
Pricing Free for 500MB; Charges depending on data size; $.10-$.12 per CPU core hour; $.15-$.18 per GB month of storage; $.11-$.13 per GB outgoing bandwidth; $.09-$.11 per GB incoming bandwidth
Supported languages Python, Java, PHP; Node.js available in the flexible environment
Storage Uses Google’s Big Table which is really good for scaling and availability but data export is very limited
Load balancing Superior to AWS, handles sudden traffic spikes better
Security Automatic security scanning: scans and detects vulnerabilities automatically
Tools Works with popular engines and IDEs: Eclipse, Pycharm, Maven, Git, Jenkins
Deployment Seems a bit complicated, according to cloud.google.com, there are few steps: 1. Change to the app engine-java-guestbook-multiphase-master/final directory, and invoke Maven; 2. Wait for the upload to complete. When the upload succeeds, you see a message similar to this one; 3. http://[PROJECT-ID].apspot.com
Usability Easy to use; allows high-traffic web applications development without managing high-traffic infrastructure
Migration An app can be migrated from one cloud to another but with limitation but there is not much information available on that

Heroku

Metric Notes
Support Heroku provides several support channels, typically response back within 24 hours
Supported platforms Runs environments for Ruby, Python, PHP, Clojure, Go, Java, Scala, and Node.js apps
Features Capacity Provisioning, Database Rollback, Application Rollback, Manual Vertical and Horizontal Scaling, App Health Monitoring, Full GitHub Integration
Value Free dyno contains 512MB RAM,100 MB storage space, professional version is available from 25$ to 500$ depending upon the dyno type
DB style Heroku Postgres is a managed SQL database service provided directly by Heroku and can access a Heroku Postgres database from any language with a PostgreSQL driver.It also provides web dashboard to share queries
Storage Different dyno sizes offer different amounts of maximum RAM and can provide up to 14GB
UI / Usability Heroku is way simpler to use as it is equipped with a ready runtime environment and application servers, seamless integration with various development instruments, redundant servers. Moreover it’s easy to add other services through add-ons
Migration Will be difficult to move from PostgreSQL, setting up SSL can be trivial
Deployment Deploy your app by simply Pushing your code into git repo provided by Heroku and its easy, requires no involvement for DevOps

IBM Cloud

Metric Notes
Support IBM offers a ticketing system with 24 hour responsiveness. It provides a developers forum
Programming languages Java, Node.js, Go, PHP, Python, Ruby on Rails, Swift
Features Continuous delivery, Alerts and notifications, Application performance monitoring, Auto scaling
DB We can choose the storage option. It ranges from RDBMS to NoSql. IBM cloud provides security, backups, database logs and auto scaling in regards to database
Value The free version for the public has no time limit, offers a 256 MB of Cloud Foundry memory. Offers a pay as you go service. IBM cloud offers an academic initiative option for students. This plan is for 6 months and is free
Storage Can scale as per our demand. It is easy to manage and is flexible. Provides object, block and file storage services
Security Allows us to have complete control over our firewall settings and other security features. Provides an on-demand security service
Migration Can migrate quickly via a CLI. It is free of cost
UI / Usability Easy to use
Deployment Easy and fast deployment

Clone this wiki locally