Your users are complaining: some system is down or slow. You need to determine if the problem is under your control or if the fault lies with a third party, such as your ISP or a SaaS provider. The time it takes to figure that out is your MTTI: "Mean Time to Innocence."
At the recent O'Reilly Velocity show in New York City, my colleague, Phil Stanhope, talked about this topic. He pointed out a few important reasons why determining MTTI is so much more complex now than it was 10+ years ago. The Internet is increasingly complex and routinely experiences outages, instabilities, and attacks. While cloud providers, CDNs, and acceleration services may claim to be "always up," that doesn't mean that they're "always reachable." In fact, they are almost certainly experiencing a constant rate of low-level failure that is largely outside IT's control and is still affecting users. Therefore, getting to MTTI is harder than ever.
How can companies address reducing — or eliminating completely — their mean time to innocence? They key is recognizing the most important third parties that your systems depend on, monitoring them constantly, and alerting when things go bad. This advice might sound obvious, but have you considered these possible infrastructure dependencies?
- Monitor the critical third-party assets and services used by your web page. Your web site no doubt invokes on a slew of external assets from analytics providers, ad networks, social networks, payment processing, and various other SaaS providers. Problems with these can slow down initial page rendering and subsequent user interactions. Make sure you're tracking and alerting on these providers' performance.
- Do you use a content delivery network (CDN)? By definition, CDNs operate large, sprawling networks with devices in many places. That's a lot of infrastructure to keep running and failures are inevitable. And CDNs are just as subject to local network anomalies as anyone else. Invest in a monitoring service that can detect localized poor performance or micro outages that affect only a subset of your users.
- Are you using cloud providers? You're probably watching the health of systems and the applications that run on them carefully, but you also need to monitor the network performance between your network and the cloud provider.
- Do you own your own IP address space? If you're using provider-independent address space (i.e., network blocks you own rather than assigned by your ISP), you need to make sure no one else is using it. Unfortunately, there's no strong authentication in the Internet's BGP routing system, so route hijacking happens: others announce your address space and some or all of the traffic bound for your network goes to them. This monitoring is one to outsource, for sure. Subscribe to a route monitoring and alert service so you always know that no one else is using your address space.
Lowering your MTTI has its benefits. You eliminate the internal blame game. Rather than different organizations pointing fingers inside the company, you can focus your efforts on solving the external problem. All the monitoring data gives you end-to-end visibility of your systems and applications and lets you control and respond to problems quickly and efficiently. This also allows you to articulate the problem quickly to customers who reach out to complain, thereby improving your overall communication with customers.
So while determining MTTI is more complex than ever, there is also more data and technology available to companies to help mitigate these complexities. If you monitor the performance of third-party assets on your web page, you watch for outages and slowdowns with your CDNs and cloud providers, and you ensure you and only you are using your IP address space, you'll be on your way to lowering your mean time to innocence and solving problems for your users that much sooner.
This article is published as part of the IDG Contributor Network. Want to Join?