Network monitoring is more complicated than ever. Most large enterprises have internal networks including LANs and WANs, and cloud storage as well, in private clouds, public clouds or hybrid clouds. This new network architecture makes it possible for businesses to move faster. However, managing and monitoring complex networks can be a challenge.
Traditional network monitoring tools track the performance and availability of each piece of network equipment. In a single geographic location or a relatively simple network, that may be fine. However, enterprise networks need to be able to monitor network equipment as well as hybrid cloud environments, LANs and WANs. Unfortunately, gaining that holistic view can be a challenge.
Network Topography is Changing
We are moving from networks that were centralized inside the enterprise. These networks had distinct tiers and logical north-south traffic flow. Today, tiers are less distinct. And enterprises are using services-based infrastructure like Software-as-a-service and Platform-as-a-service. Enterprises are also dealing with a distributed user base. Users want to plug and play with their own devices and be able to connect seamlessly at work and at home.
Before, we had to monitor network devices, such as hosts, routers and switches. Now we have all that and the cloud. It brings with it virtual machines, hypervisors, containers, virtual switches and an abundance of APIs.
The Limitations of the Tools Available
Many enterprises use more than one cloud environment. Either public and private clouds with a single vendor or cloud resources across multiple vendors. Or network resources are co-located or held in a third-party data center.
Because of the need to monitor different vendor environments, often one tool can’t do the job. Monitoring may require many tools, such as:
- Application performance tools
- Open-source tools
- Network performance tools
- Network management tools
- Logging tools
- And more
While all these tools serve a purpose, they still may not provide the visibility needed. Device-centric network monitoring tools don’t always scale for the growing enterprise. Additionally, a device-centric model can’t provide visibility into cloud resources, like cloud storage or Software-as-a-service.
There are cloud-native tools offered by cloud providers, such as AWS’s Cloud Watch, Google Cloud’s StackDriver and Azure’s Monitor. However, they have their limits. They can often see and monitor application layers and infrastructure, but they may not integrate well with other tools. Some tools may lack the features needed or may not provide the desired level of visibility.
New Tools on the Horizon
It seems like cloud providers are beginning to recognize the gap caused by the lack of network-wide monitoring. Azure created Virtual Network Tap to continuously stream virtual machine traffic to an analytics tool or packet collector. AWS offers flow logs for their Amazon EC2, Amazon VPC, CloudWatch, and Amazon S3 consoles. It captures the IP traffic between network interfaces and integrates with their Cloud Watch tool.
There are also some new tools available that can monitor multi-cloud environments. These include ThousandEyes, Kentik and Opsview. There are also advanced application performance management tools, like Dynatrace and New Relic, that can integrate data from multiple sources.
The Future of Hybrid Network Monitoring
While some of these new tools can start to solve the problems of complex enterprise network monitoring, there is another star on the horizon. Or two, actually. Software-defined Networking (SDN) and API technology.
While APIs are already relied upon for many uses. When it comes to network monitoring, APIs can be used to pull data from all across the network. This ability to pull information from disparate systems can be used to create a holistic view of even the most complex network.
Software-defined Networking is vendor-neutral and directly programmable. So it can be used with other tools to automate network management tasks. It’s also possible to manage non-SDN networks using SDN gear in parallel with traditional switching. A centralized SDN network can view the entire network environment. Using SDN, it is easy to manage and install software-based security tools such as encryption, network monitoring and virtual firewalls.
The Importance of Visibility
Network monitoring might be more complicated than ever, but we have the tools that provide a complete, holistic view. Even if, for the time being, we need to rely on more than one monitoring tool to ensure that visibility.
If you’re in the market for an enterprise network monitoring solution, it may be best not to expect the world from a single tool. However, when used together with existing network monitoring solutions and cloud-native tools, some of the promising new tools may help you gain the holistic view you’re seeking.
And, keep an eye to the future to see how SDN and API technology might revolutionize network monitoring.