[cracell]

That's just stupid. Heroku rarely gives a proper technical explanation of their outages and under reports the length and severity of them drastically.

I assume this is to maintain their SLA. We really need independent third parties to record uptime for SLAs instead of trusting hosts to do it themselves.

This outage may be the last straw with Heroku for me. They've also previously stated years ago that they would end their dependance on AWS East and yet today shows that obviously hasn't happened.