The CEO to Gryphon connector was unstable this evening from about 10:30pm ET to 11:45 ET resulting in publishing delays.
The cause of the failure was traced to the connector system running out of file allocations and unable to continue to serve connections. While the emergency failover did trigger, it did not trigger fast enough to handle the connections from the failing service. We have adjusted the timeout for the failover to be a bit more aggressive under similar conditions.
No data was lost during this window but stories published directly during this time may not appear on your live site until published again.