Server Monitoring Amazon: EC2’s Thursday Outage

We knew in 15 seconds.

Last Thursday Amazon’s EC2 services went down for about 20 minutes. Amazon didn’t announce it on the status board for 18 minutes…just before it came back online. Since the outage occurred in the middle of the night (EDT) you might think that delay wouldn’t be a big deal. But it could be, depending on what applications you’re running.

At Alertra, we knew EC2 had an issue in just 15 seconds. And that's when we started notifying our customers with servers in Amazon's cloud. How? Thanks to our trademarked Synapse™ server monitoring technology. No other monitoring service has it.

So what is it?

It starts with a tiny TCP SYN packet sent to your server every few seconds. If your server refuses the connection (returns an RST packet) or doesn't answer at all, we immediately begin a full protocol check from one of our server monitoring stations. This innovative process allows us to detect many kinds of outages within seconds.

The full protocol check is a rigorous RFC-compliant protocol test. This test can also be conducted at specified intervals in your account profile. If any issue is found we check the same server from multiple geographic locations. In this way we eliminate false alarms. You only get notified when there is a real issue that needs your attention. You can read more about it here.

Synapse™ saves you time, money, and customers.

If you’re not already an Alertra customer, why not give our technology a try? We offer free 30-day trial, no credit card required. What have you got to lose?

Or rather, what could you lose if it took 18 minutes before you knew about an outage?