Welcome to Simply Serverless. This site aims to be a hub from which you can learn to create and manage exciting and dynamic websites on Amazon AWS, Google Cloud or Microsoft Azure. .
AD
AD
There are many things you can build using serverless technologies. We're going to look at building a complex dynamic website. You can then pick and choose the components you need to utilise for your project.
| Generic Capability | Amazon AWS | Google Cloud | Microsoft Azure | IBM BlueMix |
|---|---|---|---|---|
| Static Storage | S3 | Cloud Storage | Azure Storage | Object Storage |
| DNS | Route 53 | Public DNS | Azure DNS | |
| Microservices | Lambda | Functions | Functions | OpenWhisk |
| API management | API Gateway | Cloud Endpoints | API Management | |
| CDN | Cloudfront | Cloud CDN | Azure CDN | |
| Persistence | DynamoDB | DataStore | DocumentDB | |
| Logging | CloudWatch | Cloud Logging |
AD
AD
The internet age can be loosely defined as having three ages - we'll look at each and see how development practices have evolved to the swarm. This paper from CERN (the birthplace of the Web) seems to have originated the naming strategy
Eary internet servers were well known. They may have sat under your desk, or in a rack in a nearby building. You knew what machines ran what services and if a machine broke the service went away until you lovingly replaced the faulty memory, disk or what ever. People really cared about those machines!
The big problem with lovingly caring about your pets is that it does not scale well. If you have hundreds or thousands of servers then on the law of averages you'll have some that are poorly (or dead!). If you wish to tend them all you'll be very busy. The battle cry went out to "treat them as cattle not pets!" The basic idea here is to script the build and deployment of the servers. if one is unwell, kill it off and bring up another with the same configuration. A lot of software emerged to enable this automation of infrastructure, cf Chef, Puppet and Ansible
Now we're looking at the datacentre as the computer. Your functions can be run across many machines, without you needing to be aware of the issues involved in scaling it out, mitigating against failure etc. (Well, almost)