I don’t enjoy waiting. When I’m watching the football game on Sunday afternoon, I expect as little delay between the actual touchdown and my viewing of it as possible. When I’m buying something on Amazon, I expect to know immediately if something is out of stock so I can select an alternative. In our personal lives, we expect real-time media consumption and online experiences. And the same applies to our network management. As we’re keeping watch over our networks 24/7, we want a steady stream of real-time data on network performance so that if something is wrong, we can address it immediately.
But if you talk to software vendors about their “real-time” network monitoring tools, you’ll discover that technically speaking, “real-time” should be “near real-time.” Most network monitoring solutions poll network devices on a very short interval (for most network devices, the unofficial standard I see is between 30 seconds to 1 minute, but it can be shorter for critical devices and longer for non-critical). This is fast, but if we’re being totally accurate, it’s not truly real-time.
Is it misleading, then, to say that solutions provide “real-time network monitoring” when they don’t? And is short interval-based monitoring what administrators want and need?
The goal of network monitoring is to bring back quick snapshots of network performance at regular intervals so you can see trends over time—e.g. how is this switch performing now compared to several hours ago? What network teams want is frequently updated data and little-to-no waiting time. They want to see the network as it is now instead of as it was 10 minutes ago. For 99% of network administrators, 30-second polling is as close to real-time monitoring as you can get, and it’s perfectly acceptable. If you can find a tool that polls at 30-second intervals and has a relatively small footprint on your system, you’ve done well.
To summarize the first half of this blog post: real-time network monitoring usually means very, very fast and frequent network polling—which is close enough to real-time that we don’t bother to add caveats around it.
5 benefits of real-time network monitoring
Fast, frequent polling gives network teams near real-time information to help with troubleshooting problems and making decisions. Here are five reasons why a real-time network monitoring solution is so imperative to maintaining healthy network performance:
1. You get faster notifications.
As every network professional knows, setting up the right alerts for your team is half science and half art. Nobody wants notification overkill. The trick is to set up the alerts for your core technology and for events that you need to know about immediately, such as bandwidth overload or device failure, while setting up exceptions for notifications regarding planned outages or outstanding issues that are already being worked on. And having mapping functionality allows for the immediate at-a-glance view, which combined with alerting functionality, ensures that you're getting immediate notification of potential isues on your network.
2. Your troubleshooting is faster.
The faster and more frequent your network polling is, the quicker you’re aware of issues when they hit. When a router is exceeding its normal traffic levels, for instance, you can click into status windows and trust that the metrics you’re seeing on packet loss and response time are current—not yesterday’s data—and that accuracy makes troubleshooting infinitely easier.
3. You save money.
From maintaining employee productivity to meeting customer needs, everyone knows time is money in the IT world. When something goes down, the financial consequences start immediately. If you can be ahead of serious issues with real-time network monitoring, you have a better chance of preventing major productivity and revenue loss.
4. You build valuable historic data.
A database of past performance is gold to a network team. Not only can you identify patterns and investigate issues more effectively, but you can dig into past and present capacity to recommend hardware and software upgrades as needed.
5. You can better meet SLAs.
If you’re a network consultant or a managed service provider, frequent polling and real-time data coming back about client networks is the ticket to delivering on desired uptime and availability levels.
One caveat: the more aggressive your polling periods get, the higher your system resources will be impacted. However, choosing a lightweight monitoring tool can help lessen the traffic impact. Some applications will do an initial discovery of the network and in subsequent polling, just look for deltas based on that original data, helping to reduce impact on device performance. Finding the right solution will go a long way to making sure network performance isn’t negatively impacted by your real-time network monitoring tool.
Real-time network monitoring may not mean instant, but it’s so close that the benefits are virtually the same. Network teams who can take advantage of it will reap the benefits.