Amazon, we really love your cloud service, AWS. But we can't understand why you haven't learned your lesson after S3 went down earlier this month and spurred a flurry of frustrated tweets. On the heels of that outage comes another major outage, one that shut down the likes of Reddit, Netflix, Airbnb. This outage knocked out Kissmetrics and Lighthouse — tools we depend on to get our jobs done and help our own customers.
We understand outages happen. What we don't understand is why you aren't more upfront with your customers about it. After all, the last time S3 went down, your status board showed all green except for a tiny icon that indicated something might actually be amiss. Just a reminder:
After the guff you took from the last outage, you'd think that you'd be a little bit more upfront with customers (i.e. the internet). But there were no Twitter updates (the last one tweet was two days beforehand). At first that pesky green light with the little icon popped up again:
However, we have to give you kudos for eventually changing the green light to yellow, indicating a performance issue. But it seems the issue was downplayed. We clicked the icon and got this message:
"Downgraded performance" really understates half the internet seemingly being down. A red light might have been more accurate. And you did a good job updating your status board throughout the outage, however, saying that folks would be getting a lot of error messages:
But those affected by the outage had more of a presence, reaching out to their customers about what was going on.
We've said it before and we'll say it again, building customer trust is dependent on transparency. Downplaying a problem doesn't really help anyone. You'll just end up with more and more frustrated, angry users. We hope the next time this happens, you'll be a bit more upfront with your customers.