On Sunday January 24 at 2:15am ET, a server that routes requests between our co-location facility and one of our cloud regions crashed. This region is used by Yext’s Professional Services group to run ETL jobs that load data from customer systems into the Yext Knowledge Graph, and the crash resulted in jobs scheduled for Sunday morning to fail to run. On Monday morning, we noticed the issue and restored service at 10:26am ET, and we ran the jobs which had failed on Sunday. This incident delayed affected customers' data updates propagating to their Yext Knowledge Graph, but no data was lost.
We are adding a paging-severity alert to monitor the health of this router, bringing the monitoring up to parity with the infrastructure used to serve our product.