During upgrades to a legacy service today, a tooling version mismatched caused a conflicting dependency to be installed that impacted a neighboring service. Unfortunately this legacy service isn't yet containerized, and an outage was immediately detected at 11:44am PST. As our team worked to fix the issue, it took us slightly longer than normal to repair the damage caused by the upgrade to restore service. At around 12:10pm service was fully restored and traffic was flowing again as normal.
We're reviewing our deploy practices, and we're prioritizing containerizing the legacy services involved to avoid this type of issue going forward.
Nov 16, 13:16 PST
The API is back up and accepting requests successfully. Our team is actively monitoring the situation
Nov 16, 12:12 PST
We are currently investigating downtime in our public API
Nov 16, 11:50 PST