Beginning around 2:30pm ET, we began seeing abnormally heavy loads on the load-balanced event app servers. Initial analysis indicated a high number of requests from the eP Collaboration tool (Chime app). We were seeing users still able to access the virtual platform services, though receiving intermittent errors.
We were able to implement a series of mitigations to address the short-term issue. Services were fully back up and running a bit before 4pm ET.
We take availability seriously. As such, the team is currently reviewing both the cause and mitigations put in place to determine if they are sufficient and how best to alleviate these types of issues in the future. The Development/Engineering team will be providing updates as needed and will post a post-mortem within the next day or two.