Most applications start small and at that point it’s generally easy to plan for an infrastructure. An old workstation in co-location or better yet, a virtual instance from a known cloud provider. Next step is maybe moving the database away from that. And then a message queue. And then an indexer. And so on. And on top of that all these have different operational characteristics: one will benefit from high memory, other will need high levels of iops, some will need high-throughput reads. Things are getting messy already and hopefully so! Your business is scaling.
Or maybe, your solution has large-scale requirements from the start. Maybe you’ll need to store petabytes, maybe you’ll process a million events per second. Bottom line is you’ll need to plan what kind of hardware and infrastructure you need, and of course in what amounts because later corrections might not be possible.
Our capacity planning services will help you understand the nature of your production needs, now and in the future. We start by analyzing the domain and first try to find the numbers the system will need to support, like the number of daily passengers in a subway system or visitors of a website. Then these numbers are converted to the technical requirements like a storage capacity, or number of transactions. It’s then needed to correct for seasonality, expected spikes, redundancy requirements and maintenance disruptions to get to the final design. All of these will require high levels of expertise that we have in plenty.