My first computer had a 30 MB (no, this is not a typo) hard disk. Back then, with that much available space, I thought I’d never run out. Well, I did run out of space. And actually much sooner than I thought. And this same thing still happens today to most everyone. No matter how much disk space you have, you will eventually run out of it. This is due to our use of data which is ever increasing. And just as more data takes up more storage capacity, more data requires more network capacity. Network capacity planning is the science — some call it an art — of never running out of network bandwidth and it is the subject of today’s discussion.
While network capacity planning is not particularly complex, several steps are involved. We’ll start off our discussion by explaining what it is. We’ll explain the difference in applying capacity planning to a whole network or just a single segment. Then, we’ll have a look at how network capacity planning is typically done. Our approach is a three-component process which takes the current potential capacity, the current utilization, and the estimated growth into account to evaluate future needs. More about that in a minute. Then, we’ll have a look at some of the best tools one can use to assist with network capacity planning. They mostly help with the second component of the analysis and will also be very handy to measure the benefits of capacity management.
Network Capacity Planning In A Nutshell
Network capacity planning is the process of planning a network for bandwidth utilization, availability, and other network capacity constraints. It is a way to ensure that enough capacity is always available despite the ever-increasing or often varying requirements.
Concretely, networks are not an infinite resource. Even with modern networks commonly having bandwidths of 1 Gbps, 10 Gbps, and more, with the increasing demand caused by regular growth (such increasing number of users or increasing amounts of data), their full capacity will eventually be reached. Network capacity planning will help you see it coming and react proactively.
Once you’ve gone through the steps of network capacity planning, what you do with the results is up to you. As the name implies, this is just a planning activity. How you choose to implement the plan can vary. For instance, if it is determined that a certain circuit is about to reach its full capacity, you could either decide to increase it. Alternatively, you could choose to reduce its utilization. You could also use some bandwidth management techniques to try to mitigate the need to increase bandwidth.
Whole Network Vs Specific Segment
Network capacity planning can be a network-wide activity. It can be used, for instance, to determine what equipment will replace the current one which is showing signs of age. It can help choose between 1 Gbps or 10 Gbps switches, for instance.
Network capacity planning is also often applied to specific segments of a network. It is, for instance, commonly used on WAN circuits which, due to their relatively higher costs, tend to have limited bandwidth.
How It Is Done
There are many ways one can go about when it comes to network capacity planning. As far as we’re concerned, we see it as a multi-step process. Note that this process is meant to be used on existing networks. Planning capacity for a new network that has yet to be built is somewhat different.
We’ll go into the details of each step momentarily but for now, let’s just say that we first need to evaluate the current capacity. Then, we need to measure how much of that capacity is actually used. The next step is estimating what will be the future requirements. Finally, by combining the data from the first three steps, a plan will surface.
Evaluating Current Potential Capacity
The first step is the easiest. In fact, it is more than likely already done with most networks. We need to know the potential capacity, which is the total available bandwidth. For a 1 Gbps switch, it will be 1 Gbps on each port. Likewise, if you’ve installed a 3 Mbps WAN circuit between two sites, this is its potential capacity. Capacity is fixed regardless of utilization.
If your network in well-documented, this information should be readily available. If it is not, perhaps it is a good time to document it. Tools are available to assist with mapping and documenting your network.
Measuring Current Utilization
The second step is the most critical. It has to do with measuring the current utilization of the circuits. It is important and somewhat hard to do but, fortunately, this is where you can find lots of tools to assist. In fact, all the tools we’re suggesting below can be used to measure network utilization.
There are two types of utilization measurements that are commonly available. It can be purely quantitative. Such measurement will show you how much of a circuit is used regardless of how it is used and by whom. Tools that measure bandwidth utilization will typically use the Simple Network Management Protocol (SNMP) to perform their magic. This is the type of measurement that can prove particularly helpful when the planning is done in the light of planning the capacity of WAN circuits.
Other tools provide a qualitative analysis of network utilization and will also show you with various amounts of detail what the utilization is. It can, for instance, show you which user, which application, or which type of traffic is using the available bandwidth. These tools typically rely on the NetFlow protocol to gather detailed usage data from networking equipment.
Estimating New Requirements
For this next step, you are on your own. Or are you really? In fact, if you’ve been monitoring network utilization for some time, you can readily get some data on the evolution of the utilization. The past historical data can be used to extrapolate the future evolution.
Of course, you also need some intimate knowledge of upcoming projects, deployments, and trends. No matter how good your historical data is, it won’t know–much less tell you–that the deployment of a distributed big data application is about to take place.
Estimating requirements is by no means an exact science yet it is possibly the most critical part of network capacity planning. And unfortunately, this is also something that can’t really be automated.
Putting It All Together
The most important part of network capacity planning is what comes next. That is making sense of all the information gathered in the previous steps and making a plan. Unfortunately, this is another step that is best performed by humans. There are simply too many factors to consider.
Let’s see how it’s done on an example WAN circuit. This is a 10 Mbps MPLS WAN circuit and bandwidth monitoring has shown that its current average utilization is about 4 Mbps and it’s been steadily growing at a rate of 0,5 Mbps each month. This means that you’ll reach 70% utilization in just 6 months. And since you should never allow average utilization to exceed 70%, you can conclude that you’ll either need to increase the circuit or find a way to reduce bandwidth usage—by using WAN optimization techniques, for example—in the next 6 months.
The Best Network Capacity Planning Tools
So, as we stated before, the best tools you can use for network capacity planning are those that will let you monitor traffic. The tools we’ve picked for inclusion on our list are all tools—or toolsets—that allow both bandwidth usage measurement and network traffic analysis. It is important because, although bandwidth usage will show you which circuit or segment has or will have congestion, knowing what traffic causes it can help you make a better decision as to what the best course of action is.
Here is our list with the best tools for network capacity planning.
If you’ve been in network administration for any length of time, chances are you’ve heard of SolarWinds. The company has been around for over almost 20 years and is known to make some of the best free tools targeting specific tasks of network administrators. Among others, they have a great subnet calculator. But SolarWinds’ is even better known for making some of the best bandwidth monitoring tools and the best NetFlow collectors and analyzers. These two tools are known as the SolarWinds Network Performance Monitor and the SolarWinds NetFlow Traffic Analyzer. They can be purchased together as the SolarWinds Network Bandwidth Analyzer Pack and together, they form an excellent toolkit for measuring network utilization, both quantitatively and qualitatively.
The first component of the bundle, the Network Performance Monitor or NPM is a feature-rich SNMP network monitoring system. It has an easy to use web-based dashboard, is easy to set up. It also boasts customizable alerting features that can notify administrators of specific metrics exceeding a predefined value.
The NPM can also build intelligent maps of your network—a useful feature in the initial documentation of your network—and can display the critical path between any two points on the network. The software also features wireless network monitoring and management. Network Insights, a feature that is available for Cisco Nexus and ASA devices as well as F5 load balancers, provides advanced monitoring of several parameters that are specific to these devices.
The second component of this bundle is the NetFlow Traffic Analyzer or NTA, an add-on module to the Network Performance Monitor. This tool provides detailed information on what the observed traffic is. It will, for instance, report on what type of traffic is more frequent or what user is using the most bandwidth. Several different views are available on the tool’s dashboard such as top applications, top protocols or top talkers.
The NTA uses the NetFlow protocol and receives detailed usage information from your NetFlow-enabled network devices. Although originally developed by Cisco, NetFlow is included in equipment from many vendors. It may bear a different name such as J-Flow on Juniper devices but all variants are supported by the NTA. The NetFlow protocol has even been standardized by the IETF as IPFIX, or IP Flow Information Exchange which is, of course, also supported by the NTA.
Prices for the SolarWinds Network Bandwidth Analyzer Pack start $4 910 for monitoring up to 100 elements. Prices go up from there based on the number of monitored devices. If you prefer to try the product before buying it, a free 30-day trial can be downloaded from SolarWinds.
2. Paessler PRTG
The Paessler Router Traffic Grapher, or PRTG, is a multi-function monitoring system. At its base, it is a bandwidth utilization monitoring system. It uses SNMP to read usage data from networking devices. But PRTG doesn’t stop at bandwidth monitoring and it will use other technologies to monitor systems. This tool relies on sensors to do its magic. For bandwidth monitoring, every interface is a sensor. But there are many other types of sensor. There is, for instance, a NetFlow sensor that will provide detailed usage data.
PRTG runs on Windows, however, its user interface is web-based and can be accessed from any browser on any platform. There are also mobile apps for Android and iOS that you can install on your smartphone. One of PRTG’s main features is its speed of installation. Paessler claims you could be up and running within a couple of minutes. Whether this is true is open for debate but installation is indeed quick and easy, thanks in part to the product’s auto-discovery feature.
PRTG is available in two versions. The free version is limited to 100 sensors. For more than 100 sensors, you need one of the paid licenses which are available for 500, 1000, 2500, 5000, and unlimited nodes at prices varying from about $1 600 to slightly under $15 000. The free version will allow unlimited sensors for the first 30 days allowing you to test-drive the product.
3. ManageEngine OpManager Plus
You may already know the ManageEngine OpManager, a well-known bandwidth monitoring tool. A somewhat lesser-known product from ManageEngine, possibly due to its very similar name, is the ManageEngine OpManager Plus. The “Plus” in the product’s name is significant. This tool combines the network monitoring features of its little brother with bandwidth management using NetFlow and other similar protocols, making it a great product to use for network capacity planning.
It doesn’t stop there, though. The ManageEngine OpManager Plus also features server monitoring, network configuration management, firewall log analysis, IP address and switch ports management, and application management. It is one of the most complete suites of network administration tools. And the software has one of the best dashboards which is both intuitive and easy to use, thanks to its drill-down functionality. For mobile device users, client apps for iOS and Android are also available.
The network monitoring and bandwidth management features and the bandwidth analysis module of the ManageEngine OpManager Plus combine to give you all the data you need for network capacity planning. And in addition, you also get a ton of other useful tools.
The ManageEngine OpManager Plus prices start at $995 for up to 25 devices. Prices go up as you add more devices. Detailed pricing information for your specific situation can be obtained by contacting ManageEngine. Like most products it this range, a free 30-day trial is available.
4. Nagios XI
Nagios XI is a very good monitoring tool for networks and servers. It’s built upon the open-source Nagios Core engine. The tool will run on CentOS or RedHat Enterprise Linux. It could, of course, run as a virtual machine on a Windows host. Unlike the other network tools on this list, Nagios XI doesn’t use SNMP to monitor network devices. It has its own proprietary method built into the Nagios Core engine, which communicates with devices to gather status reports. The metrics that are gathered regularly by the monitor appear as live data in the dashboard and you can also opt to store data for analysis later. Overall, this makes for a rather powerful tool.
Despite its use of a different data gathering scheme, Nagios XI will provide all the data you need to perform your network capacity planning duties. It has a very comfortable dashboard where you can quickly view the status of your network. This product also has one of the best alerting systems to notify administrators when something odd is detected.
Nagios XI is available in a Standard edition starting at $1 995 and an Enterprise edition starting at $3 495. The latter has extended functionality and includes additional features to assist in large-scale configuration, network capacity planning, and scheduled reporting. Prices vary according to the number of devices to monitor. If you want to try the product, a 60-day free trial can be downloaded.
5. WhatsUp Gold with Network Traffic Analysis add-on
WhatsUp Gold from Ipswitch used to be just an up-or-down monitoring tool many years ago. It has evolved into one of the best system and network monitoring platform. And through multiple available add-ons, WhatsUp Gold’s functionality can be easily expanded. One such add-on is the Network Traffic Analysis add-on.
This add-on is really what makes WhatsUp Gold an excellent fit for network capacity planning. It will provide a wealth of information on existing traffic capacity and utilization. You can also run a test of new applications you plan to deploy on the network and use the add-on’s CBQoS and NBAR capabilities to track its traffic patterns. The traffic monitor presents data that can be sorted, filtered, and aggregated and shows you the capacity of each link and network device. Traffic patterns can be derived from live data as well as historical data.
The WhatsUp Gold Network Traffic Analysis add-on is available as an add-on to the Premium edition of WhatsUp Gold and is included with the WhatsUp Gold Total Plus edition. A free 30-day trial of WhatsUp Gold and its Network Traffic Analysis add-on is available.
Have you used any of these? Let us know in the comments section below.