Do you wish you had a checklist to make sure you’ve taken every possible step to maximize your network health?
Every network is different. Depending on your organization’s needs and infrastructure, the way you configure and monitor it will vary. But following best practices will help you know how to improve network infrastructure. Here are ten steps you can take to improve your network monitoring.
1. Map out your current architecture and ensure its accuracy
The first step of network monitoring is to know what’s in your network. How many routers and switches do you have? How many endpoints? Along with a total device inventory, you need to know where each one is physically located and how it connects to the rest of your network. Without a complete picture of your network hardware, you can’t effectively monitor it.
To save you the pain of manually documenting this information, network mapping software can create an automatic diagram in a matter of minutes. Once you have an up-to-date map in your hands, you can tweak the layout and icons to fit your preferences. Most network mapping tools also make it easy to see which devices are up and which are down, so you know where to start troubleshooting.
2. Determine what information is most critical to monitor and collect
For many organizations, losing devices in the middle, like a core switch stack, would result in no email and internet access for users. It’s crucial that you know which devices could cause major operational bottlenecks for your users if they fail. You also need to know which metrics to watch that will signal when a device is going off the grid. Make sure you have the ability to get these metrics from your mission-critical devices.
Most network monitoring software uses SNMP or other protocols to capture common device metrics but will also give you the capability to create your own custom device probes to pull exactly what you need from your hardware.
3. Configure exception-based notifications with a clear escalation plan
Notifications are only helpful if they come at the right time, for the right reason, to the right person. You want to know when you have a potential network problem on your hands, but you don’t want false alarms or false positives filling up your inbox.
Simplify your network monitoring by only setting up notifications about the information you need. Create exception-based notifications that will help you quickly solve real problems.
Then, make sure you have an escalation plan—and that everyone on your IT team knows who will be getting alerts and which point, how they need to respond, and what should happen at each stage. Solidifying your plan helps ensure that alerts are effective and that any severe network issues are dealt with correctly. This is especially important if you have SLAs to meet.
4. Identify and monitor network performance baselines
Before you take steps to improve your network health, you need to know how it is behaving now and therefore, how to monitor network performance. Your current performance will serve as a baseline. Then you can set appropriate threshold ranges for acceptable device performance. That way, when thresholds are exceeded, you know you’re actually dealing with poor performance and not just a false alarm.
Also make sure you’re aware of regular, scheduled exceptions to normal performance levels. Running backups at night, for example, could cause an unexpected jump in traffic levels, but if you know that high utilization during a specific timeframe at night is to be expected, you can tell your software not to alert you when these anticipated spikes occur.
5. Configure SNMP on network devices
You can gather a huge amount of information about network performance by using SNMP, one of the most common Internet protocols. Configure your core infrastructure to support SNMP to help you monitor metrics beyond simply up or down status.
I recommend using SNMPv3 so you can ensure network security.
6. Make upgrades to your core hardware to support performance monitoring protocols
When it comes to your core network infrastructure, your time will be better spent replacing outdated hardware instead of optimizing it. If the core switches, routers, and firewalls in your network don’t support modern performance monitoring protocols, the type of performance data you want to capture will be limited—which could inhibit your ability to effectively monitor, respond, and mitigate network problems in a timely manner.
By upgrading these devices, you ensure you have the visibility you need.
7. Use NetFlow to capture and monitor bandwidth utilization on routers and switches
You can analyze NetFlow, a Cisco protocol that collects IP traffic information, to discover who’s consuming bandwidth and how it’s being consumed across your network. NetFlow and its related counterparts sFlow, jFlow, and cFlow can be analyzed with a network traffic monitoring tool. These tools capture bandwidth data and let you easily graph or filter it to find the answers you need.
Using a network traffic monitoring tool to collect information on traffic flowing across your network can help you more easily identify the source of bottlenecks, as well as plan how much bandwidth capacity you’ll need to support a growing network.
8. Use protocols to support Windows and Linux Servers
Using WMI and PowerShell give you a simple and secure way to allow connections on remote Windows machines. WMI and PowerShell can also be used to monitor Windows applications and systems events. SSH, also known as Secure Shell, is similar to WMI and PowerShell but mainly used to provide secure connections to the Linux environment. It also allows a Linux network monitoring solution to be able to securely connect to the Linux environment and monitor specific events.
9. Keep probe and MIB architecture updated to ensure monitoring of standard and non-standard devices
Networks with a variety of standard and non-standard equipment require probe and MIB architecture that is constantly up-to-date.
Most capable network monitoring solutions will provide within the product a database of industry SNMP MIBs that enables capturing standard performance statistics for network components. There is, however, vendor-provided MIB information that provides and supports more non-standard information that could be crucial to device performance. This information is constantly getting updated based on new vendor hardware and software releases. Any custom probes or MIBs using this information should also be updated to ensure accurate information is captured.
10. Implement continuous monitoring (24/7)
Monitoring 8-5 while you’re in the office is the first step to optimal network performance, but enabling ongoing monitoring will help support healthy network operations at all hours of the day—especially in industries like finance where you need 24/7 availability of your products and services. Monitoring the network in an ongoing fashion doesn’t have to require more work or office time. A network monitoring solution will keep tabs on device statuses and send you real-time alerts via email or text anytime network issues occur.
Follow these ten steps to get maximum benefits—a healthy network, improved response time, and increased visibility—from your network monitoring.
... then our Ultimate Buyer's Guide will help answer all your questions. Compare vendors, learn about trends, and find out what features you need.