This is the third post in the series of posts detailing Codechecker, specifically on scale, performance and security aspects. You can read all the posts here in our codechecker posts.
In October, we increased our number of servers compared to September or August. This was a simple experiment to see how additional servers would impact our numbers. We run a fleet of on-demand and spot instances to manage our codechecker infrastructure. As a result, we doubled the number of spot instances we run. This gave us the performance boost in our 99th percentile of the queue time. Now, we plan to run our fleet with the extra servers going forward.
