Challenges, not problems
The system needed to take emails and templates in various formats, process them, and send them out. It needed to keep track of each email, with exact status of what happened to it. It needed to scale by just adding servers. And it needed detailed ad hoc reports.
Solutions, not ideas
We built a system which was horizontally scalable from Day 0, automatically distributing mail traffic and log processing workloads across all nodes in the cluster. Every event on every email is tracked and updated in system logs. A Solr cluster holds the logs, pushing out archives to AWS S3.
Now that’s value
Broadside can stress both network bandwidth and CPU power of all servers it runs on. It is a superb example of solving a serious challenge at scale, processing 10s of millions of mails a day.