How to build a Hyper-Available Infrastructure – Part 1: Overview
In today’s digital world we highly depend on services and their corresponding data to be available 24/7. We are frantic about our digital life and demand the underlying infrastructures to be Hyper-available.
While traveling through the region to visit customers and partners, I am highly depending on a lot of those online services and corresponding data it holds for my travel. My air-tickets, parking arrangement and hotel are booked pre-travel. I get my confirmation and tickets digitally by email, app and/or text message. Calendar entries are automatically updated with travel information.
On travel day Waze looks at my calendar and keeps an eye out on traffic jams so I get notified if I need to leave my house and start moving towards the airport earlier. At the airport my car license plate gets scanned and matched to the reservation in the parking system. After parking the car, I move swiftly through the airport. The security gates doom up and I need my boarding pass before I can enter and greet the security officers this morning who are awaiting me for a thorough inspection.
Crucial in this whole travel ritual is that all services and data are hyper-available to me. My boarding pass sits in the KLM app on my mobile phone, if there are any changes like delays, boarding times, gates and such I get notified right away. A lot of things can go haywire in this whole process which can ruin my day. Think about what could happen! The mobile data network going down, KLM app not notifying me of any changes or a parking system not recognising reservations and such. While it is uncomfortable that services we depend on are offline, but what about those systems being online but the data they work with has been tempered with? Wouldn’t that scare you more then not being able to use the systems?!?
As you can see we need confidence that data we use is integer and that the services we use need to be always online or also called be hyper-available! Our digital life depends on data coming from a multitude of services and underlying systems. The infrastructure these services run on is the critical underlying foundation. Ask yourself how would you design an infrastructure that is hyper-available?
To answer that question, we must know the level of insurance the organisation and users need/are willing to pay for. This way we can design the infrastructure accordingly. Always design with failure domains in mind, where resiliency to failure are key factors. Make sure data is protected at all times, across multiple systems, but also use that data so you do not end up with a reactive insurance policy.
Levels of insurance
How do you measure the levels of insurance the organisation needs? This can be done by classifying the services and corresponding data with Recovery Time Objectives (RTO) and the maximum amount of data loss that is acceptable to the organisation measured with Recovery Point Objectives (RPO).
When discussing these metrics with the people in your organisation they first will tell you that they do not want to lose any data (RPO = 0) and that services need to be always online (RTO=0). Just make sure their data and services are hyper-available! It is important to map the different services and data to get a correct insight in terms of impact on the organisation. Often not all services need the same amount of protection and speed of recovery. Employees often can do their work when a service is back online in a basic form, where for instance historical data is not yet available but will be restored to the service a bit later.
Designing Hyper-Available systems is, and will always be, a balancing act about managing expectations, limitations and available money to get the best possible solution. By approaching it per service in your organization you will have the correct levels of availability setup, while full-filling the defined availability requirements. Data being used by these services is most critical but also is exploding in a frantic rate to keep up with.
Maximize return on investment
To optimize the use of resources for business operations, I will dive deeper in possible solutions by combining innovative software solutions from VMware and Veeam and how they work together in tandem to deliver a Hyper-Available foundation. Protecting and managing against loss, outage, theft, integrity of services and data over multiple platforms.
In the next parts of this blog series we will dive deeper into the multitude of possibilities by combining VMware and Veeam.