Past Incidents

Adult database (specifically a1) was having trouble keeping up with the traffic. We've offloaded some work to the slaves, but it didn't help much. Service was available all the time, but with considerable slowdown. Sysadmins will try to move the instance to faster hardware.

After deployment of a fix one (out of four) PHP machines had a failure resolving opcache, which caused roughly one quarter of requests towards frontend sites to fail. Issue is resolved.