- 020 8163 7876
- [email protected]
Established in 1987
The largest IT services provider to NHS England.
£200 Million per annum
During the organisations migration journey to AWS, there were a automation and skills gaps in the various teams. Our job was to implement well architected solutions by creating easy to use repeatable DevOps tooling & ways of working.
Migration to AWS was taking too long due to the initial complicated IaC.
Costs were very high. AWS Commercials and savings opportunities are being missed.
Using scheduled scaling and elastic scaling wasn't being used.
Many teams didn't have the correct DevOps tooling and mindset experience.
Legacy and ageing infrastructure operating systems.
Waterfall & Legacy ways of working.
No automation & lack of documentation.
Operational Costs are high for server patching.
Microsoft Licensing Costs are high.
Modernise servers/services where possible.
Our general approach with any infrastructure migration is to review each on premise capability to ensure we understand what we are moving.
This builds confidence and creates an easier approach to transfer the service to the AWS hosted capability.
Image Factory
Create an Image Factory with Hashicorp Packer. This will allow us to use our signature layered AMI build process. This ensures a patch process is put forward and common agents and configuration installed during the builds. Each server on premise is then audited and an AWS AMI created.
Using Spot instances to keep build costs low.
IaC
Using the AMI's we deploy the EC2 infrastructure using Terragrunt/Terraform. This allows continuous delivery of the server and ensure that immutable servers are hosted in ASG's and continuously updated by automated CI/CD.
A repeatable and reliable way to deliver infrastructure.
Copy & Paste Code makes it easy for the team to use.
For static servers, we plug in WSUS/Patch Management systems to ensure security compliance.
Database migration
Database solutions are complete by using data migration and RDS MS SQL Server. Further reducing operational cost by using AWS Serverless technology,
Modernisation
Ensuring that stateless servers are implemented with Auto Scaling Groups, means that we can reduce operational cost, and no longer require Active/Passive servers - thus reducing the hosting cost.
AWS
A well architected delivery for the datacentre migration.
Security: Delinea PAM access integration to allow users access to the EC2 instances.
Costs: Savings Plans/Reserved Instances all used to reduce commercial costs. Opensource tools used to migrate and create infrastructure, which keeps costs down. Automated infrastructure provisioning allows for transient environments which allows downtime when not in use.
Operations: We introduced Datadog for enhanced observability and cost management. A shift left patch process by building AMI's and releasing to infrastructure, reduced WSUS overhead.
Performance: A like for like deployment of VMWare to AWS EC2 instances ensures that the services continue to run at their best.
Sustainability: Our infrastructure provisioning automation allow non production environments to be created and destroyed when not in use, Using Autoscaling groups allow servers to scale up and down when required.
Reliability: Using our analysis of server capability we are able to host some servers in Auto scaling groups, which allows us to run 1 server, rather than an active passive setup. This helps reduce the cost by 50% when running a capability, due to the build it fault tolerance of ASG’s.
People
We help and trained multiple organisation units to modern ways of working on introduce a DevOps mentality to their way of working.
A collaborative way of working, sharing ideas and common code base helped many teams work more efficiently.
Processes
Many processes are automated, which provides a far more efficient way of working,
Products
Opensource tooling for the migration keeps the costs down without reducing quality.
The versatility and extensibility means we can automate more make the delivery of services repeatable, reliable and reproduceable.
A layered AMI approach to allow patching, application builds to be as fast and painless as possible.
Automated implementation using Terraform for observability metrics from Infrastructure, APM, Logs, Syntethic Montioring and custom metrics provided a single pane of glass to view the entire estate.
