How to Monitor MPLS Networks

• Why Monitoring MPLS Networks is Important
• Common MPLS Issues
• The Challenges of Monitoring MPLS Networks
• 1. Understand the MPLS Network Design
• 2. Deploy MPLS Network Monitoring
• 3. Monitor MPLS Network Locations
• 4. Activate QoS Monitoring
• 5. Monitor MPLS Performance Metrics
• 6. Leverage Device Monitoring

MPLS (Multiprotocol Label Switching) is a type of data-carrying technique for high-performance telecommunications networks. MPLS networks are generally used for a variety of applications, including:

• VPNs (Virtual Private Networks)
• Traffic Engineering
• Carrier Networks
• Data Center Networks
• Internet Service Provider Networks

In an MPLS network, QoS (Quality of Service) is the most important feature. The big advantage of an MPLS network is about having a guarantee of level of performance based on your SLA (Service Level Agreement).

Of course, if you’re paying for a certain level of service, you can’t just rely on your SLA. You need to ensure that you’re really getting what you paid for and that the service quality is matching the SLA.

Additionally, many businesses are actually migrating from MPLS networks to SD-WAN networks. So if your business is still using an MPLS network, it’s because you really need it. And if you need it, you need to make sure it performs as promised.

How can you do that?

Well, you monitor your MPLS network performance.

This way you can identify:

• If your MPLS network is performing as promised
• If it isn’t, what problems are affecting your MPLS network
• Are the problems happening in your Local Network or Service Provider Network
• Who is responsible for troubleshooting

Continuously monitoring MPLS network performance allows you to proactively detect and troubleshoot any MPLS issues to ensure that your MPLS network operates reliably and optimally.

Like any network technology, MPLS networks can encounter various problems that can impact network performance and user experience. Before we get to how to monitor, let’s briefly go over some common problems that can occur in MPLS networks:

1. Quality of Service (QoS) Problems: MPLS networks often include QoS mechanisms to prioritize and shape traffic, but problems can occur when traffic is misclassified or when bandwidth is not properly managed.

2. Scalability Issues: One of the huge disadvantages of MPLS is the price related to the speed of the service. As MPLS networks grow and traffic levels increase, there can be budgetary limitations when upgrading and scaling up the service. But if you don’t upgrade your service, the increased levels of traffic can cause network congestion issues.

3. Network Congestion: Continued from the last point MPLS networks can experience network congestion and, when they do, their performance will degrade horribly. When this happens, it’s important to ensure that QoS and prioritization is done correctly. When the lowest priority apps are congestion, they don’t affect the highest priority application.

4. Network Failures: MPLS networks can be impacted by various types of failures, including hardware failures, software failures, and network outages. MPLS networks are usually used for a large number of sites, so with a huge number of sites, it’s not about if you’re going to have an issue, but more about when and how you’re going to react when it does happen.

5. Configuration Errors: MPLS networks are highly configurable. The retaging of DSCP codes by an MPLS equipment or errors in the configuration of routers, switches, and other network devices can result in incorrect routing, degraded performance, and connectivity issues. Even if just 1 piece of network equipment is misconfigured, it can have an impact on the entirety of the service.

Obkio Network Monitoring graphs showing network congestion

Because of the complexity of MPLS networks, there are some challenges with monitoring them.

1. It’s a Managed Service: Firstly, MPLS Networks are usually hosted by Service Providers in their network, so it’s hard to monitor if you don’t have the tools available.

2. Monitoring isn’t included: Because MPLS is a managed service, many businesses think that monitoring services are included in that service. But, Service Providers don’t monitor the performance of MPLS. With MPLS services, there’s usually an SLA included in the contract. But Service Providers don’t perform SLA monitoring to ensure that SLA is upheld. It’s always up to the customer to prove that.

3. There’s no visibility: When network problems arise, Service Providers may have internal alerts to know if the service is up or down. But, if the performance degrades, there’s no trigger unless the customer alerts it themselves. Then when a problem is flagged, the Service Provider will be blind to what is actually happening, since they have no visibility and MPLS networks are so vast.

Because of this, it’s up to businesses to implement their own MPLS monitoring to get the visibility they need of their MPLS network to identify issues, troubleshoot internally, or get support from their Service Providers if the problem is on their end. Here’s how:

To accurately and proactively monitor your MPLS service, you need to continuously monitor every end of your network. To do so, you need to understand the design of a MPLS network to know what to monitor. When monitoring your MPLS network, you’re going to adapt your monitoring setup to the design of your MPLS.

There are several different types of MPLS network topologies, including:

• Full Mesh Topology: In a full mesh topology, every router is connected to every other router in the network. This type of topology is suitable for large networks and provides maximum reliability, as traffic can be routed through multiple paths in the case of a network failure.

• Point-to-Point (P2P) Topology: In a P2P topology, each router has a direct connection to every other router in the network. This type of topology is suitable for small networks.

• Star Topology: In a star topology, all routers are connected to a central hub. Traffic is routed through the hub and then out to the other routers. This type of topology is suitable for large networks.

• Hierarchical Topology: In a hierarchical topology, the network is divided into several levels, with routers at each level responsible for routing traffic to a specific portion of the network. This type of topology is used in large, complex networks.

In general, MPLS networks can be more complex to monitor than traditional IP networks due to the additional layer of abstraction introduced by the MPLS label switching technology. However, with the right tools and techniques, you can get the level of visibility needed to monitor and troubleshoot MPLS networks.

To monitor MPLS, you need a tool like Obkio Network Performance Monitoring software, which is designed for MPLS network monitoring.

An MPLS Network Monitoring tool like Obkio, uses Network Monitoring Agents deployed at key network locations, which exchange synthetic traffic between them to monitor key network performance metrics, such as packet loss, latency, and jitter to identify performance issues and ensure that network resources are being used efficiently.

MPLS monitoring also monitors network utilization, traffic patterns, and QoS, and uses SNMP to monitor network devices and identify bandwidth congestion.

Get started with Obkio’s Free Trial!

To achieve the level depth required to monitor MPLS networks, you need to install Network Monitoring Agents in all the network locations that you want to monitor, including”

• Head office
• Remote offices
• Branch offices
• Data centers

To know exactly where to deploy your Agents, you need to understand your MPLS design (as stated earlier). Monitoring Templates will also help you identify where to deploy your Agents.

But essentially, for MPLS, you’ll want to install your Monitoring Agents in the customer LAN, behind the MPLS Customer Edge Router (CE), to monitor your MPLS network performance from the end-user point of view. This way you can proactively identify network problems before your actual users even experience them.

MPLS Networks and Private Networks alike are used to transport various types of applications like Voice over IP (VoIP), Video Conferencing (ex: GotoMeeting, Zoom, Webex), Unified Communications (ex: Skype for Business) and Collaboration (ex: Microsoft Teams).

As mentioned above, MPLS networks are prone to experience network congestion. And the applications listed above are very sensitive to issues like network congestion.

When monitoring MPLS, it’s important to implement QoS (Quality of Service) to prioritize some traffic on the network in order to reduce latency, jitter and packet loss. In case of network congestion, QoS features ensure that performance sensitive applications can function without any degradation and that only less critical applications (such as web browsing) are impacted.

QoS needs to be configured end-to-end, and you need to ensure it’s done correctly on every connection.

For Example: if your business is monitoring two offices, QoS needs to be configured in between those two. Additionally, A Network Monitoring tool, like Obkio, can check if the DSCP configurations are done correctly for all those connections. If there is retagging of the DSCP codes in between, Obkio can detect them.

The way to monitor and measure MPLS network performance is to measure key network metrics. These metrics will help you understand if your MPLS performance is degrading. Your Network Monitoring solution will continuously monitor performance between all your network locations to and measure metrics like:

• Latency
• Jitter
• Packet Loss

In addition to measuring the usual metrics mentioned, for an MPLS network you also need to monitor DSCP Propagation. That means that you need to monitor latency, jitter, packet loss per Class of Service.

Think of Class of Service like different lines to get into a bar: There’s the normal line, and then there’s the VIP line. It's important to calculate metrics for every Class of Service and every line, but there may be some metrics that are more important than others for certain applications.

For Example: For VoIP and UC applications there is a very low tolerance for packet loss, therefore the traffic through those applications should be prioritized and fast-tracked through the “VIP” line. Actions like emailing and file transfer have a high tolerance for performance issues, and therefore have a low priority and can go through the “regular” line.

Finally, when monitoring your MPLS network, you can leverage SNMP Device Monitoring to identify if network problems are happening in your local network, on your network devices. on Obkio’s solution will perform SNMP Polling from your end, and the Customer Edge Router and measure metrics like:

• CPU Usage
• Bandwidth Usage

Obkio Network Deivice Monitoring graphs

If the MPLS issue is in your business’ network, you can use the data from Device Monitoring to troubleshoot. If it’s in your Service Provider’s network, you can collect more data using Obkio Vision Visual Traceroutes and share the data with your Service Provider to give them the visibility they need to troubleshoot on their end.

Obkio Vision Visual Traceroutes

Once you’ve deployed your MPLS monitoring setup, you can really understand if your MPLS service is performing like your Service Provider promised. But remember, it’s up to your business to identify performance degradation and proactively identify network problems that can affect your business.

Install more Monitoring Agents in all your network locations, and keep Obkio as your permanent MPLS monitoring tool!

Get started with Obkio's free trial.