A key challenge for Sameroom—a technology that deals with dozens of external services—is resilience.
If some subset of external chat services misbehaves, Sameroom users of the healthy set should remain blissfully unaware of any trouble. When a misbehaving service comes back online, Sameroom has to detect this and re-enable all affected integrations, as quickly as possible.
We've spent quite some time tuning Sameroom to comply with these requirements, with pretty good results.
Next, we had to figure out a way to visualize the overall status, to give our customers a way to conduct first-line-of-defense troubleshooting in case something goes wrong.
Today, we're launching the first version of our Provider Health Status page:
The Provider Health Status page is home to a graph with three pieces of information: the types of services in play (including the mysterious Other), the relative size of each service's slice out of the overall pie (all active—at least one live Tube—accounts registered with Sameroom), and the ratio of failed-to-healthy connections within each slice.
Our next task is to separate the various IRC networks into independent providers (they've historically been lumped into an umbrella "IRC" one) and to gather user feedback.
If you have any questions or suggestions about this new functionality, drop us a line.