Home

When building applications for scale, how important is QA testing?

Testing whatever you create is essential. We are in the process of adopting a strategy Netflix came up with: they simulate problems on the go, because scale can only be tested when there is a similar situation. You cannot take for granted that your application is built for that scale, because certain components may go wrong. The solution has to be resilient. We take care of latency for the applications that we build at Spice Money, because even 30 seconds matter in FinTech. And when you are talking about scale, data synchronization between the different systems becomes important because we want to have real-time data. While creating those architectures, list the probable points of failure and how you will address them. If one of your systems failed, your applications at scale will be a disaster if you have only created a single point of failure. Consider all the different pieces that can fail as well as the capacity you have built for. If even a single component in your entire application becomes a bottleneck, your entire architecture is impacted. That’s why it is important to invest a lot of time in thinking about and criticizing your own design, because you need to proactively point out all possibilities. And you need to run through whatever you have created with your operations team or the people who will be managing that particular system. What challenges are they facing? If something goes wrong in real time while your applications are implemented, what is the response time? How soon can you come back? If we have taken care of these areas, then we are pretty much home free. But if we do not put enough thought into them, it will be a difficult situation for everyone.

7 views
1 comments
0 upvotes
Related Tags
Anonymous Author
Testing whatever you create is essential. We are in the process of adopting a strategy Netflix came up with: they simulate problems on the go, because scale can only be tested when there is a similar situation. You cannot take for granted that your application is built for that scale, because certain components may go wrong. The solution has to be resilient. We take care of latency for the applications that we build at Spice Money, because even 30 seconds matter in FinTech. And when you are talking about scale, data synchronization between the different systems becomes important because we want to have real-time data. While creating those architectures, list the probable points of failure and how you will address them. If one of your systems failed, your applications at scale will be a disaster if you have only created a single point of failure. Consider all the different pieces that can fail as well as the capacity you have built for. If even a single component in your entire application becomes a bottleneck, your entire architecture is impacted. That’s why it is important to invest a lot of time in thinking about and criticizing your own design, because you need to proactively point out all possibilities. And you need to run through whatever you have created with your operations team or the people who will be managing that particular system. What challenges are they facing? If something goes wrong in real time while your applications are implemented, what is the response time? How soon can you come back? If we have taken care of these areas, then we are pretty much home free. But if we do not put enough thought into them, it will be a difficult situation for everyone.
0 upvotes