Stability is also about building for the future. This means we have upgrade our whole software stack both on the server side and on the client as well as the mobile apps. This way we support the newest phones and their features as well as keep up with any type of security measures and other things. The same was we work through and optimise our code so does the Thord part libraries and solutions we rely on. So just by keeping them up to date we automatically can gain stability and performance.

Another part of stability is refactoring code. Over time your code and feature does things it wasn’t designed to do in the first iteration, you add stuff on top and on the side. Refactoring code makes it more manageable and also provide for more flexibility to build new features.

We also need to keep up to date with guidelines from Google and Apple for their platforms as they require certain things that didn’t exist when the apps were originally launched. This necessarily doesn’t giv you any benefits but it’s just to follow their policies. If you choose not to do so they might remove your apps from the App Store and Google Play.

It’s not all about code thou. The database is one of the key parts of a data driven platform. We have optimised the table layout, created better indexes that has improved certain operations from seconds down to a tenth of a second. Another way to improve performance is to optimise and tweaks the queries made to the database, and sometimes caching data to not hit the database at all. We have also increased memory on the database server and configured it to hold more of the data in memory which makes all operations much faster. 
The database are backed up up to 30 days back so we can roll back up to a month, we also have backups on the images and files that are uploaded. These jobs and routines also take time to setup and configure but if everything work as expected you should never need it. This is like being an insurance, time and money spent that doesn’t add any value except feeling comfortable and safe when any accident happens.

The last piece is the physical hardware. House of RC runs on services on Amazon Web Services deployed both in US and Europe to scale and be close to the users. We also have content replicators so static files can be served as close as possible to the user. We have upgraded our EC2 instances on AWS to have more CPUs and more memory. It was part of this upgrade and a move of out domain registrar that caused the domain problem during Christmas when the domain was down for 5 days. That is something we don’t want to ever happen again.

For some people this might be technical mumbo jumbo but we hope it gives you a feeling of how many things that we need to think about and take care of to host and operate a platform like House of RC. We would love to just focus on writing new a great features for you to benefit from, but if the infrastructure and the software stack is not stable and performant we wouldn’t be able to do that.

Next time we will post about one of the biggest new areas of House of RC 2.0, personalization and customization.