What’s New with vRealize Network Insight 4.2

We’re excited to announce the release of VMware vRealize Network Insight 4.2, with another ton of new awesome features! With new data sources, network flow latency metrics, another open-source project, application discovery and public cloud enhancements, and a ton of small things that just makes you love using the product; let’s take a deep dive into 4.2!

Data Sources

Data is life to Network Insight. It ingests data from all kinds of different devices, virtual, physical, cloud, and then correlates all this data to make it useful. The more data sources it has, the more value it will have.

Fortinet Firewall Support

With Network Insight 4.2, we’re happy to add Fortinet to our ecosystem of data sources. Get full visibility of the Fortinet FortiManager management plane that manages the FortiGate firewalls; see it in context of the entire infrastructure, instantly see which firewall rules are attached to which VMs, get a clear picture of the network topology, and more.

OpenShift Container Platform

On the subject of data sources; 4.2 also adds support for the OpenShift container platform. We added support for container workloads with VMware Enterprise PKS and vanilla Kubernetes in 4.1, which brought network visibility, application security and migration planning, and network troubleshooting to the smallest workloads.

User Assisted Network Information (Custom Data Sources)

Pronounced U-AN-I (like in the song Runaway from Galantis), this custom data source option is pretty exciting. We’re opening up the data source support and allowing you to input your router or switch information, no matter the vendor. Network Insight has a supported data source list, and all the data sources on this list are developed and tested by the team. If there is a network device in your infrastructure that is not on that list, you can now add it to Network Insight anyway!

Adding UANI data sources will give you: VM-to-VM network paths, a dashboard for the router or switch with the network topology, a list of all interfaces, routes, VRFs, the works. It will be a regular router or switch.

I’ll do a more in-depth look into UANI in a next blog post. I just cannot leave out the exciting new open-source project around this, which can be found here: https://github.com/vmware/network-insight-sdk-generic-datasources

This is a Python library which helps to automate the collection of information (interfaces, routes, etc.) from any network device and translates it into a prescriptive format for UANI and then uploads it to Network Insight. The idea is to have the SDK running in a scheduled task, and updating these data sources on an interval (just like regular data sources).

Network Latency Metrics

Latency is one of the best indicators for the performance of an application; more latency is more waiting time for the end-user. Network Insight 4.2 introduces network latency metrics in 2 forms: TCP Round-Trip Time (RTT) that is attached to network flows and latency metrics between specific components (virtual and physical NICs in the vSphere host).

The TCP RTT metrics are attached to network flows and can be found in the dashboard of a flow (click on a specific flow) or you can use the search engine, for example Average Tcp RTT of flow where Average Tcp RTT > 10ms. This will result in a list of flows that have a higher RTT latency of 10ms.

Average RTT latency attached to network flowsThere is also a beautiful way to analyze the latency performance of the entire infrastructure by using the Flow Insights page and looking at the Network Performance diagram.

Use the latency Sankey chart to quickly discover problematic flowsThis diagram shows the overall latency statistics and quickly identifies any flows that are in the worst-performing group. You can then drill down to find the exact flows that are not performing and analyze the cause.

Other latency metrics that are also available, are latency metrics on virtual NICs (vNIC) and physical NICs (pNIC). These latency metrics are currently tracked inside the same vSphere host and the different paths that are supported, are vNIC to pNIC, pNIC to vNIC and vNIC to vNIC.

To get latency metrics, NSX Data Center for vSphere 6.4.5 or higher is required.

Stay tuned for more exciting developments on network latency, the future is bright for network administrators that need to troubleshoot application performance!

Application Discovery

Network Insight 4.1 introduced application discovery, which discovers application boundaries by only using the metadata of the infrastructure. No agents required. There are a few different methods; using workload tags, importing applications from the ServiceNow CMDB or using a naming convention.

The last one, the naming convention, uses a regular expression based discovery to pull the application name straight from the workload (VMs or AWS EC2 instances) name. This requires you to translate the naming convention to a regular expression, which can be hard. Not anymore!

The Pattern Builder simplifies building regular expressions. Click here if the video doesn’t show.

By using the new Pattern Builder (located next to the application and tier name field in the application discovery wizard), it’s possible to build these regular expressions without any knowledge about these expressions. Simply select a group that is between (automatically detected) separators or select specific positions of a name, if there are no separators.

Starting Network Insight 4.2, you can also store application discovery runs into templates. That way it is easier to run the application discovery multiple times, without having to fill out the options each time.

Application Dashboard Improvements

There is a spectacular dashboard in Network Insight that is focused on the application. Everything is there; the topology of the application itself, real-time network flows that flow between the tiers and the outside of the application, all infrastructure events, all workload metrics; everything you need to troubleshoot the application. The network topology is especially useful as it displays network flows, but this was limited to the tiers in 4.1. With Network Insight 4.2, every single flow is visualized when you zoom in. By zooming in, the tiers get expanded to include all workloads (VMs, containers) and the flow connections for each of those workloads:

Click here if the video doesn’t show

More? Yes, more!

As you might be used to with new Network Insight features, the above highlights are just the top new things; but there’s plenty more. Here’s a small list that’s also awesome:

• Application Discovery can now also save applications in bulk.
• Kubernetes objects can now be part of an application, including hybrid (VMs, Physical and Containers) applications.
• AWS EC2 instance types and subnets are now tracked.
• AWS Security Groups are tracked with an audit log.
• Modifying Pinboards is much more intuitive (move pins around, edit descriptions, etc.)

For the full release notes, click here. For the downloadable bits, click here. Upgrades to 4.2 are supported from version 4.0 and higher.