1. Home
  2. Systems design
Systems design

Designed for the day something fails.

A system is only as good as its worst day. We start with your recovery objectives and design the network, identity and workloads around them.

Principles

Four rules we design by.

Design for failure

Every component will fail eventually. We decide in advance what happens when it does, and build for that.

Identity is the perimeter

In the cloud, who can sign in matters more than which network they are on. Identity is designed first, not bolted on.

Everything as code

If it isn't in code, it can't be reviewed, repeated or rebuilt. Networks, policies and servers all live in Terraform.

Write the decisions down

Each significant choice is recorded with its reasons and alternatives, so the next engineer understands why.

Users reach the primary region. Azure Site Recovery keeps the secondary ready.The primary region goes dark. Traffic moves to the secondary, which was ready the whole time.

See it work

Two regions, one plan.

A typical design runs in a primary Azure region and replicates to a secondary with Azure Site Recovery. Switch to a region outage to see what changes for your users.

  • Architecture designDiagrams and written decisions
  • Recovery objectivesRPO and RTO agreed per workload
  • Terraform modulesThe design, deployable
  • Failover runbooksTested, step by step
In plain language

The terms we’ll use.

Recovery planning has its own vocabulary. These are the words that will come up in our first conversation.

RPORecovery point objective
How much data you can afford to lose, measured in time. An RPO of 15 minutes means losing at most the last 15 minutes of changes.
RTORecovery time objective
How long a system can be down before it hurts the business. It sets how fast recovery has to be.
Failover
Moving a running workload to a second region when the first is unavailable.
Failback
Returning the workload to its home region once that region is healthy again, without losing what changed in the meantime.
Landing zone
The prepared Azure foundation of subscriptions, network, policy and identity that workloads are deployed into.

Plan for your
worst day.

info@kinstacktech.com Book a consultation