[nknealk]

Thankfully the redshift outage was just on APIs, not existing machines. Our cluster was fine today, but external schema which rely on glue/athena did time out.