An ongoing increase in traffic caused an overall increase in latency across much of our infrastructure. Certain high traffic, high through actions, such as recording and querying workout logs caused latency (and ultimately errors) across must of the iFit application.
In response, we increased the total size of the infrastructure to handle the increase in load. We also offloaded certain actions, such as querying activity logs, to separate infrastructure. We also tweaked various caching strategies to improve overall performance.
From here, we will continue to monitor. We do expect traffic to increase further in the near future. We are taking various steps at both the infrastructure and application levels to improve performance and tolerance as traffic increases.