It is a 'rumour network' so my guesses are:
1. Software update which hasn't gone through the agreed change control process, or
2. Third party hosting provider outage, or
3. Hardware issue at a 'single point of failure', or
4. Undetected software bug.
All mitigatable given adequate resources.