Starting Thursday afternoon (12 June), the computers that make up the IEM webfarm started to experience high CPU loads. The primary symptom was timeouts or errors servicing Web Map Service requests. I thought something was wrong with the satellite imagery being used by the services, but that turned out to be a red herring. One of the internal filesystem shares was locked up and causing cascading failures. I finally figured it out this morning and got things repaired around 7 AM. Sorry for the troubles.