Facebook recently made headlines when the company announced it performed an in-house stress test on its core network infrastructure. It shut down an entire data center for a day to see how standing safeguards would actually respond. The result? Overall, applications remained available and the system persevered.
As Facebook’s Jay Parikh, global head of engineering, explained during a keynote presentation at the @Scale conference in San Francisco, the experiment was no small operation. According to Parikh, it required turning off tens of megawatts of power.
While Facebook did not say which data center it shut down, the scope of the experiment included an entire region of end users. Ultimately the test verified the resiliency of Facebook’s network, proving that it is capable of maintaining uptime in the event of an emergency. But it begs the question: What would have happened if the network was unable to respond as expected? At the least, Facebook would be in a major rut trying to get systems back online—nevermind restoring any lost data caused by the system failure.
If you are concerned about the resiliency of your network, we encourage you to stay away from “do it yourself” network stress tests. Remember that Facebook is a massive organization that has the resources and expertise to pull off such a stunt. But trying something like this in your data center could have disastrous consequences.
Instead, invest in a root-cause troubleshooting solution that will provide you with quality network performance information. Using Total Network Visibility from PathSolutions, you can see exactly how your network is performing in real-time. You can spot minor glitches before they become big problems, which means your business will never have to worry about its network crashing.
Click here for more information about how PathSolutions can provide you with the peace of mind you are looking for in your data center.