This incident has been resolved.
Starting Wed, October 7th around 07:20am (UTC), the primary ISP for our US-hosted datacenter was experiencing issues that may have caused intermittent connectivity. We failed over to the backup ISP services at that time.
Customers may have experienced symptoms including (but not limited to):
At 09:15am (UTC) the service was restored back to the primary ISP.
We are currently investigating this issue.