Sam & Ashish, co-founders of Ring.io here. Please pass this note to anyone at your company who should be aware of what happened with Ring.io this week.

As you're getting to wrap up your week we wanted to write to personally apologize for the disruption that Ring.io caused to your business and your team and to take a minute to explain what happened this week that caused 2 outages, and what we're doing to fix it going forward.

What happened
Ring.io is a SAAS company built on cloud services like most SAAS companies. We rely on specialized hosting providers to run the software that powers our service. 

On Wednesday, Aspect Cloud, the provider that runs our telephony servers, had an outage that disabled all processing in their North American data centers.

Ring.io offers a service for routing calls around an outage called Backup Number.

Unfortunately, Backup Number relies on the availability of Aspect Cloud. This is a design flaw that we simply didn't anticipate. Aspect Cloud has been our provider for 6 years and until 2016 they had a stellar record running our application.

When we published on Wednesday our root cause analysis for that outage we committed ourselves to improving our Backup Number functionality to still work even if Aspect Cloud is down. Work on this enhancement has started.

Then, today, Aspect Cloud has a different kind of outage, but one that affected a large percentage of our calls. Our operations team worked with their operations team to restore the service - but the restoration was complex and involved cycling clusters in and out of rotation - meaning that calls would be processed, then not processed for a while.

We were able to route calls using Backup Number but not in all circumstances - whenever Aspect Cloud was hard down, there's nothing we could do.

The timeline of the Wednesday outage and of today's outage are available on our status website.

What we're doing to fix this going forward

• We are improving the Backup Number functionality so that: a) it works even if Aspect Cloud is down, b) it works not just for inbound calls, but also for outbound calls
• We are accelerating the plan to have more than one voice processing provider