Optimum performance is something we’re always in search of in IT. When we’re not reinforcing the home of a super demanding workload, we’re ensuring that the chances of our company website or HR database crashing are as slim as possible.
It can feel a precarious place to be, on one hand hoping that tech stays up and running and then on the other hand hoping it’s working at its best. It doesn’t leave much margin for error. High availability is the opportunity to look across your business, the tech that supports it and consider the risks that could be threatening the reliability of particular workloads.
What is high availability?
Application High Availability (sometimes referred to as App HA or just HA) is a consultative philosophy that saves us from praying for the safety of our servers every night before bed, or crossing our fingers when we notice a major surge in traffic and server activity. It helps us look at server or application availability and mathematically diminish the risks involved in keeping our applications up and running.
Once you’ve got a big picture understanding of what poses a threat to uptime, we can take steps to circumvent any outage in the future. What you’re achieving by commissioning a consultant to carry out High Availability analysis is ensuring your system is configured to handle a variety of failures and loads with very little, or no, downtime.
Achieving high availability
As is often the case in the world of tech, throwing extra components, hardware and systems into the mix isn’t the way to get a more stable (and in turn, more available) application. The more components you have, the more things can potentially contribute to downtime. Through load balancing, we’re able to distribute workloads across a cluster to optimise resource usage, minimising response while maximising output, ensuring we don’t put too much of a strain on a system.
Application high availability is a consultative exercise so there are a variety of ways of improving resilience, each of which depend on your circumstances, business aims and application. Here, we’ll go into just a couple of ways availability can be bolstered in a little bit of detail…
1. Consider hosting location If an issue affects a whole data centre, and your application is hosted in that data centre, you’ve quite literally put all of your eggs in one basket. Natural disasters are a bit of an over the top thing to be thinking about all of the time, but the reality is they do happen, and your customers probably won’t care about a flood if they can’t access software their business relies on. Whole swaths of Amazon’s network have suffered outage in the past, affecting big dogs like Netflix, so hosting in multiple locations is a fairly easy place to start when it comes to improving application resilience.
2. Application servers One server running your application means that if a great load of traffic hits it all at once, (which, for the business, wouldn’t be a bad thing) users experience slow loading and crashes. Distributing the load of your application across various servers means that if one server needs to restart, requests are still served and users are still smiling. Going a step further, consider the cloud native development philosophy, where micro services and applications exist on individual servers, handling a specific part of an application only, thus improving the availability of the application by spreading the load and how the resource is used.
What’s the cost of high availability?
Finally, it isn’t cheap, which is important to shout out in this blog. Of course, downtime costs money, reputational damage and in the worst cases, company closure, but the cost of high availability means that it’s usually reserved for when extra performance is crucial to business success. It can be difficult to quantify, but if your business can financially justify the cost of extra performance (and so, uses workloads that gobble resource to generate revenue), this might just be the service for you.
If you’d like to know more about our application HA service, how we engage clients or what you can expect, check out our Application High Availability service page here. If you’re a little further along with your need for application HA, be sure to drop us a line using the form below. We’d love to hear from you.
Posted in Cloud Consulting on Aug 24, 2017