Network Overlays

October 25, 2014

by Matt Conran Blog

Network Overlays

In the world of networking, there is a hidden gem that has been revolutionizing the way we connect and communicate. Network overlays, the mystical layer that enhances our networks, are here to unlock new possibilities and transform the way we experience connectivity. In this blog post, we will delve into the enchanting world of network overlays, exploring their benefits, functionality, and potential applications.

Network overlays, at their core, are virtual networks created on top of physical networks. They act as an additional layer, abstracting the underlying infrastructure and providing a flexible and scalable network environment. By decoupling the logical and physical aspects of networking, overlays enable simplified management, efficient resource utilization, and dynamic adaptation to changing requirements.

One of the key elements that make network overlays so powerful is their ability to encapsulate and transport network traffic. By encapsulating data packets within packets of a different protocol, overlays create virtual tunnels that can traverse different networks, regardless of their underlying infrastructure. This magic enables seamless communication between geographically dispersed devices and networks, bringing about a new level of connectivity.

The versatility of network overlays opens up a world of possibilities. From enhancing security through encrypted tunnels to enabling network virtualization and multi-tenancy, overlays empower organizations to build complex and dynamic network architectures. They facilitate the deployment of services, applications, and virtual machines across different environments, allowing for efficient resource utilization and improved scalability.

Network overlays have found their place in various domains. In data centers, overlays enable the creation of virtual networks for different tenants, isolating their traffic and providing enhanced security. In cloud computing, overlays play a crucial role in enabling seamless communication between different cloud providers and environments. Additionally, overlays have been leveraged in Software-Defined Networking (SDN) to enable network programmability and agility.

Matt Conran

Highlights: Network Overlays

Overlay Networking

Logical networks

Overlay networks are computer networks that are layered on top of other networks (logical instead of physical). They differ from the traditional OSI layered network model and almost always assume that the underlay network is an IP network. These technologies include VXLAN, BGP VPNs, Layer 2 and Layer 3, and IP over IP, such as GRE or IPsec tunnels. Overlay networks, such as SD-WAN, use IP over IP technologies.

The overlay network (SDN overlay) allows multiple network layers to be run on top of each other, adding new applications and improving security. Multiple secure overlays can be created using software over existing networking hardware infrastructure by making virtual connections between two endpoints. In the cloud, endpoints can be physical locations, such as network ports, or logical locations, such as software addresses.

Software tags, labels, and encryption create a virtual tunnel between two network endpoints. End users must be authenticated to use the connection if encryption is used. Like a phone system, the technology can be considered endpoints with identification tags. An identification tag or number can be used to locate a device in a network, creating virtual connections.

Benefits of Network Overlays:

1. Simplified Network Management: With network overlays, organizations can manage their networks centrally, using software-defined networking (SDN) controllers. This centralized approach eliminates the need for manual configuration and reduces the complexity associated with traditional network management.

2. Enhanced Scalability: Network overlays enable businesses to scale their networks easily by provisioning virtual networks on demand. This flexibility allows rapid deployment of new services and applications without physical network reconfiguration.

3. Improved Security: Network overlays provide an additional layer of security by encapsulating traffic within virtual tunnels. This isolation helps prevent unauthorized access and reduces the risk of potential security breaches, especially in multi-tenant environments.

4. Interoperability: Network overlays can be deployed across heterogeneous environments, enabling seamless connectivity between different network types, such as private and public clouds. This interoperability extends the network across multiple locations and integrates various cloud services effortlessly.

Scalability and Elasticity

One critical advantage of network overlays is their ability to scale and adapt to changing network requirements. Overlays can dynamically allocate resources based on demand, allowing network administrators to rapidly deploy new services or expand existing ones. This elasticity enables organizations to meet the evolving needs of their users and applications without the constraints imposed by the underlying physical infrastructure.

Simplified Network Management

Network overlays simplify network management by providing a centralized control plane. This control plane abstracts the complexity of the underlying physical infrastructure, allowing administrators to configure and manage the virtual networks through a single interface. This simplification reduces operational overhead, minimizes human errors, and enhances network security.

Google Cloud Data Centers

**How Cloud Service Mesh Works**

At its core, a cloud service mesh consists of a data plane and a control plane. The data plane is responsible for handling the actual data transfer between services, while the control plane manages the policies and configurations that govern this communication. By deploying sidecar proxies alongside each microservice, the service mesh can intercept and manage all network traffic, ensuring that communication is secure, reliable, and observable.

**Key Benefits of Cloud Service Mesh**

1. **Enhanced Security:** One of the primary advantages of a cloud service mesh is the ability to enforce security policies consistently across all microservices. With features like mutual TLS (mTLS) for encryption, service mesh ensures that data in transit is secure, reducing the risk of breaches.

2. **Observability and Monitoring:** Service mesh provides comprehensive visibility into service-to-service communication. By collecting metrics, logs, and traces, it allows for detailed monitoring and troubleshooting, enabling teams to quickly identify and resolve issues.

3. **Traffic Management:** Advanced traffic management capabilities, such as load balancing, traffic splitting, and circuit breaking, are built into the service mesh. These features ensure that services can handle variable loads and maintain high availability and performance.

**Implementing a Cloud Service Mesh**

Adopting a cloud service mesh requires careful planning and consideration. Organizations should start by evaluating their current architecture and identifying key areas where a service mesh can bring immediate benefits. It’s also essential to choose the right service mesh solution, such as Istio, Linkerd, or Consul, based on specific needs and compatibility with the existing environment.

**Challenges and Considerations**

While the advantages of a cloud service mesh are clear, there are also challenges to consider. Implementing a service mesh introduces additional complexity and overhead, which can impact performance if not managed properly. It’s crucial to have a thorough understanding of the service mesh architecture and to invest in proper training and support for the team.

Example Technology: EIGRP and GRE

In simple terms, GRE is a tunneling protocol that encapsulates various network layer protocols within IP packets. It establishes a virtual point-to-point connection between different networks, facilitating data transmission across disparate networks. Encapsulating packets within GRE headers allows for secure and efficient communication.

To establish a GRE tunnel, two endpoints are required: a source and a destination. The source endpoint encapsulates the original packet by adding a GRE header, while the destination endpoint decapsulates the packet and forwards it to the appropriate destination. This encapsulation process involves adding an extra IP header, which allows the packet to traverse the network as if it were a regular IP packet.

Networking approach based on overlays

Different overlay networking approaches are often debated in the SDN community. Depending on the technology, some software-only solutions may not be able to integrate at the chip level. The layering of software and processing in overlay networking has been criticized for creating performance overhead. Network overlays are controlled by SDN controllers using the OpenFlow protocol, which requires specific software code or “agents” to be installed.

Change in Traffic Patterns

Thanks to the paradigm shift toward cloud computing, a host of physical servers and I/O devices can host multiple virtual servers that share the same logical network despite being in remote locations. In contrast to the traditional north-south direction of data traffic within data centers, virtualization has facilitated more significant east-west data traffic. Communication between servers and applications within a data center is known as east-west traffic.

In corporate networks or on the Internet, much of the data required by the end user involves more complex data that requires preprocessing. Using a web server (via an app server) to access a database as an example of east-west traffic, we can demonstrate the need for preprocessing.

The birth of network overlays

Network virtualization overlays have become the de facto solution for addressing the problems just described regarding data center expansion. Overlays allow existing network technologies to be abstracted, extending the capabilities of classic networks.

Networking has been using overlays for quite some time. As their name implies, overlays were developed to overcome the disadvantages of conventional networks. An overlay is a tunnel that runs on a physical network infrastructure.

Following MPLS- and GRE-based encapsulations in the 1990s, other tunneling technologies, such as IPsec,8 6in4,9, and L2TPv3,10, also gained popularity. For example, 6in4 tunnels were used to carry payloads over a transport network that could not support the payload type. These tunnels were utilized for security purposes, simplifying routing lookups, or carrying payloads over unsupported transport networks.

Understanding MPLS Forwarding

MPLS (Multi-Protocol Label Switching) forwarding is used in modern computer networks to route data packets efficiently. Unlike traditional IP routing, which relies on complex table lookups, MPLS forwarding utilizes labels to simplify and expedite packet forwarding. These labels act as virtual shortcuts, enabling faster and more streamlined transmission. To comprehend MPLS forwarding in action, let’s consider a hypothetical scenario of a multinational corporation with branch offices in different countries.

The organization can establish a private network that connects all its branches securely and efficiently by implementing MPLS forwarding. MPLS labels are assigned to packets at the ingress router and guide them through the network, ensuring reliable and optimized data transmission. This enables seamless communication, data sharing, and collaborative workflows across geographically dispersed locations.

What is the LDP Protocol?

The LDP protocol, short for Label Distribution Protocol, is a signaling protocol used in Multiprotocol Label Switching (MPLS) networks. It facilitates the exchange of label mapping information between Label Switching Routers (LSRs), allowing them to establish forwarding equivalence classes (FECs) and efficiently forward data packets.

Label Distribution: The core functionality of the LDP protocol lies in its ability to distribute and assign labels to network paths. These labels help LSRs establish predetermined forwarding paths, enabling faster and more efficient packet forwarding.

Traffic Engineering: Through its Traffic Engineering (TE) extensions, the LDP protocol allows network administrators to optimize the utilization of network resources. Dynamically adjusting label assignments and traffic flows enables better load balancing and network performance.

Network Overlays and Virtual Networks

Network overlays have emerged as a powerful solution to address the challenges of modern networks’ increasing complexity. This blog post will explore network overlays, their benefits, and how they improve connectivity and scalability in today’s digital landscape.

Network overlays are virtual networks that run on physical networks, providing an additional abstraction layer. They allow organizations to create logical networks independent of the underlying physical infrastructure. This decoupling enables flexibility, scalability, and simplified management of complex network architectures.

Overlay networking — Diagram: Overlay Networking with VXLAN

Virtual Network Services

Network overlays refer to virtualizing network services and infrastructure over existing physical networks. By decoupling the network control plane from the underlying hardware, network overlays provide a layer of abstraction that simplifies network management while offering enhanced flexibility and scalability. This approach allows organizations to create virtual networks tailored to their specific needs without the constraints imposed by physical infrastructure limitations.

Creating an overlay tunnel

A network overlay is an architecture that creates a virtualized network on top of an existing physical network. It allows multiple virtual networks to run independently and securely on the same physical infrastructure. Network overlays are a great way to create a more secure and flexible network environment without investing in new infrastructure.

Network overlays can be used for various applications, such as creating virtual LANs (VLANs), virtual private networks (VPNs), and multicast networks. For example, we have DMVPN (Dynamic Multipoint VPN), with several DMVPN phases providing a secure network technology that allows for multiple sites’ efficient and secure connection.

In addition, they can segment traffic and provide secure communication between two or more networks. As a result, network overlays allow for more efficient resource use and provide better performance, scalability, and security.

Securing and overlay: GRE and IPSec

When combined, GRE and IPSec create a robust security infrastructure that addresses tunneling and encryption requirements. GRE tunnels establish secure connections between networks, enabling the transmission of encapsulated packets. IPSec then encrypts these packets, ensuring that data remains confidential and protected from interception. This powerful combination allows organizations to establish secure and private communication channels over untrusted networks like the Internet.

The utilization of GRE and IPSec brings numerous benefits to network security. Firstly, organizations can establish secure and scalable virtual private networks (VPNs) using GRE tunnels, allowing remote employees to access internal resources securely. Secondly, IPSec encryption protects data during transmission, safeguarding against eavesdropping and tampering. Additionally, the combination of GRE and IPSec facilitates secure communication between branch offices, enabling seamless collaboration and data sharing.

Enhanced Connectivity:

Network overlays improve connectivity by enabling seamless communication between different network domains. By abstracting the underlying physical infrastructure, overlays facilitate the creation of virtual network segments that can span geographical locations, data centers, and cloud environments. This enhanced connectivity promotes better collaboration, data sharing, and application access within and across organizations.

Example: VXLAN Flood and Learn

Unveiling Flood and Learn

Flood and learn is a fundamental concept in VXLAN that allows hosts to efficiently learn the MAC addresses of virtual machines within the same VXLAN segment. When a host receives an unknown unicast, it floods the packet to all other hosts within the VXLAN segment. The destination host then learns the MAC address and updates its forwarding table accordingly. This process ensures that subsequent packets are directly forwarded to the destination host, minimizing unnecessary flooding.

In traditional flood and learn implementations, flooding occurs using broadcast or unicast methods, which can lead to significant network congestion. However, by leveraging multicast, VXLAN flood and learn achieve enhanced efficiency and scalability. Multicast groups are established for each VXLAN segment, ensuring unknown unicast packets are only flooded to hosts subscribed to the respective multicast group. This targeted flooding reduces network overhead and optimizes bandwidth utilization.

VXLAN flood and learn with multicast opens up a world of possibilities in various networking scenarios. From data center virtualization to cloud computing environments, this approach offers benefits such as improved performance, reduced network latency, and simplified network management. It enables seamless communication between virtual machines, even across different physical hosts or clusters, enhancing the flexibility and scalability of virtualized networks.

Understanding IPv6 Tunneling

IPv6 tunneling is a mechanism that encapsulates IPv6 packets within IPv4 packets, allowing them to traverse an IPv4 network. This enables communication between IPv6-enabled devices across IPv4-only networks. By encapsulating IPv6 packets within IPv4 packets, tunneling provides a practical solution for the coexistence of both protocols.

Types of IPv6 Tunneling

There are several methods for implementing IPv6 tunneling over IPv4, each with advantages and considerations. Let’s explore some popular types:

Manual Tunneling: Manual tunneling involves configuring tunnels between IPv6 and IPv4 endpoints. This method requires configuring tunnel endpoints and addressing them, making it suitable for smaller networks or specific scenarios.

Automatic tunneling, also known as 6to4 tunneling, allows for automatically creating IPv6 tunnels over IPv4 networks. It utilizes a 6to4 relay router to facilitate communication between IPv6 and IPv4 networks. Automatic tunneling is relatively easy to set up and does not require manual configuration.

Teredo Tunneling: Teredo tunneling is a mechanism that enables IPv6 connectivity over IPv4 networks, even behind Network Address Translations (NATs). It provides a way for IPv6 traffic to traverse NAT devices by encapsulating IPv6 packets within UDP packets. Teredo tunneling is particularly useful for home networks and scenarios where IPv6 connectivity is limited.

Advanced Topic

DMVPM:

The underlay network forms the foundation of any DMVPN deployment. It consists of the physical infrastructure that connects the various endpoints. The underlay network ensures reliable and efficient data transmission from routers and switches to cables and network protocols. Key considerations in establishing a robust underlay include network design, redundancy, Quality of Service (QoS) policies, and security measures.

DMVPN truly shines in the overlay network, built on top of the underlay. It enables secure and efficient communication between remote sites, regardless of their physical locations. By leveraging multipoint GRE tunnels and dynamic routing protocols such as EIGRP or OSPF, DMVPN establishes a mesh network that seamlessly connects all endpoints. This overlay architecture eliminates the need for complex and static point-to-point VPN configurations, providing scalability and ease of management.

Benefits and Use Cases:

DMVPN offers a plethora of benefits and is extensively used across various industries. Its ability to provide secure and scalable connectivity makes it ideal for enterprises with multiple branch offices. By utilizing DMVPN, organizations can optimize their network infrastructure, reduce costs associated with traditional VPN solutions, and enhance overall network performance. Additionally, DMVPN enables seamless integration with cloud services and facilitates secure remote access for teleworkers.

Introducing Single Hub Dual Cloud

The single-hub dual cloud architecture takes DMVPN’s capabilities to the next level. In this setup, a central hub connects to two separate cloud service providers simultaneously, forming redundant paths for traffic. This redundancy ensures high availability and fault tolerance, making it an attractive option for businesses with critical network requirements.

Implementing DMVPN with a single hub dual cloud configuration offers several advantages. Firstly, it enhances network resilience by providing built-in failover capabilities. In a cloud service provider outage, traffic can seamlessly transition to the alternate cloud, minimizing downtime. Additionally, this architecture improves network performance through load balancing, distributing traffic across multiple paths for optimal utilization.

Related: Before you proceed, you may find the following helpful:

Highlights: Network Overlays

Supporting distributed application

There has been a significant paradigm shift in data center networks. This evolution has driven network overlays known as tunnel overlay, bringing several new requirements to data center designs. Distributed applications are transforming traffic profiles, and there is a rapid rise in intra-DC traffic ( East-West ).

We designers face several challenges in supporting this type of scale. First, we must implement network virtualization with the overlay tunnel for large cloud deployments.

Suppose a customer requires a logical segment per application, and each application requires load balancing or firewall services between segments. In that case, having an all-physical network using traditional VLANs is impossible. The limitations of 4000 VLANS and the requirement for stretched Layer 2 subnets have pushed designers to virtualize workloads over an underlying network.

Concepts of network Virtualization

Network virtualization involves dividing a single physical network into multiple virtual networks. Virtualizing a resource allows it to be shared by various users. Numerous virtual networks have emerged over the decades to satisfy different needs.

A primary distinction between these different types is their model for providing network connectivity. Networks can provide connectivity via bridging (L2) or routing (L3). Thus, virtual networks can be either virtual L2 networks or virtual L3 networks.

Virtual networks started with the Virtual Local Area Network (VLAN). First, the VLAN was invented to lessen the unnecessary chatter in a Layer 2 network by isolating applications from their noisy neighbors. Then VLAN was then pushed into the world of security.

Then, we had Virtual Routing and Forwarding (VRF). The virtual L3 network was invented along with the L3 Virtual Private Network (L3VPN) to solve the problem of interconnecting geographically disparate enterprise networks over a public network.

VXLAN vs VLAN

One of the first notable differences between VXLAN and VLAN was increased scalability. The VXLAN ID is 24 bits, enabling you to create up to 16 million isolated networks. This overcomes the limitation of VLANs having the 12-bit VLAN ID, which enables only a maximum of 4094 isolated networks.

Tunnel Overlay — Multiple segments per application and the need for a tunnel overlay,

What are the drawbacks of network overlays, and how does it affect network stability?

Control Plane Interaction

Tunneled network overlays

Virtualization adds a level of complexity to the network. Consider the example of a standard tunnel. We are essentially virtualizing workloads over an underlying network. From a control plane perspective, there must be more than one control plane.

This results in two views of the network’s forwarding and reachability information—a view from the tunnel endpoints and a view from the underlying network. The control plane may be static or dynamic and provides reachability through the virtual topology on top of it, which provides reachability to the tunnel endpoints.

The overlay tunnel and potential consequences.

Router A has two paths to reach 192.0.2.0/24. Already, we have the complexity of influencing and managing what traffic should and shouldn’t go down the tunnel. Modifying metrics for specific destinations will influence path selection, but this comes with additional configuration complexity and policies’ manual management.

The incorrect interaction configuration between two control planes may cause a routing loop or suboptimal routing through the tunnel interfaces. The “routers in the middle” and the “routers at tunnel edges” have different views of the network – increasing network complexity.

A key point: Not an independent topology

These two control planes may seem to act independently but are not independent topologies. The control plane of the virtual topology relies heavily on the control plane of the underlying network. These control planes should not be allowed to interplay freely, as both can react differently to inevitable failures. The timing of the convergence process and how quickly each control plane reacts may be the same or different.

The underlying network could converge faster or slower than the overlaying control plane, affecting application performance. Design best practice is to design the network overlays control plane so that it detects and reacts to network failures faster than the underlying control plane or have the underlying control plane detect and respond faster than the network overlays control plane.

Encapsulation overhead

Every VXLAN packet originating from the end host and sent toward the IP core will be stamped with a VXLAN header. This leads to an additional 50 bytes per packet from the source to the destination server. If the core cannot accommodate the greater MTU size or the Path MTU is broken, the packet may be fragmented into smaller pieces. Also, the VXLAN header must be encapsulated and de-encapsulated on the virtual switch, which takes up computing cycles. Both of these are problematic for network performance.

Security in a tunnel overlay

Tunnels and network overlays have many security flaws. The most notable is that they hide path information. A tunnel can pass one route on one day and take another path on a different day, and the change of path may be unknown to the network administrator. Traditional routing is hop-by-hop; every router decides where the traffic should be routed.

However, independent hop-by-hop decisions are not signaled or known by the tunnel endpoints. As a result, an attacker can direct the tunnel traffic via an unintended path where the rerouted traffic can be monitored and snooped.

Tunneled traffic hides from any policies or security checkpoints. Many firewalls have HTTP port 80 open to support web browsing. This can allow an attacker to tunnel traffic in an HTTP envelope, bypassing all the security checks. There are also several security implications if you are tunneling with GRE.

First, GRE does not perform encryption or authentication on any part of the data journey. The optional 32-bit tunnel key for identifying individual traffic flows can easily be brute-forced due to the restriction of 2×32 number combinations.

Finally, it implements the sequence number used to provide a method of in-order delivery poorly. These shortcomings have opened up to several MTU-based and GRE packet injection attacks.

STP and Layer 2 attacks

VXLAN extends layer 2 domains across layer 3 boundaries, resulting in more extensive layer 2 flat networks. Regarding intrusion, the attack zones become much more significant as we connect up to two remote disjointed endpoints. This increases the attack zones over traditional VLANs where the Layer 2 broadcast domain was much smaller.

You are open to various STP attacks if you run STP over VXLAN. Tools such as BSD brconfig and Linux bridge-utilis allow you to generate STP frames into a Layer 2 network and can be used to insert a rogue root bridge to modify the traffic path.

Tunnel overlay with VXLAN inbuilt security?

The VXLAN standard has no built-in security, so if your core is not secure and becomes compromised, so will all your VXLAN tunneled traffic. Schemes such as 802.1x should be deployed for the admission control of VTEP ( tunnel endpoints ). 802.1x at the edges provides defense so that rogue endpoints may not inject traffic into the VXLAN cloud. The VXLAN payload can also be encrypted with IPsec.

Closing Points: Understanding Network Overlays

At its core, a network overlay is a virtual network created using software-defined networking (SDN) technologies. It enables the creation of logical network segments independent of the physical infrastructure. By decoupling the network’s control plane from its data plane, overlays provide network architectures flexibility, scalability, and agility.

Benefits of Network Overlays

Enhanced Security and Isolation

Network overlays provide strong isolation between virtual networks, ensuring that traffic remains separate and secure. This isolation helps protect sensitive data and prevents unauthorized access, making overlays an ideal solution for multi-tenant environments.

Simplified Network Management

With network overlays, administrators can manage and control the network centrally, regardless of the underlying physical infrastructure. This centralized management simplifies network provisioning, configuration, and troubleshooting, improving operational efficiency.

Summary: Network Overlays

Network overlays have revolutionized the way we connect and communicate in the digital realm. In this blog post, we will explore the fascinating world of network overlays, their purpose, benefits, and how they function. So, fasten your seatbelts as we embark on this exciting journey!

What are Network Overlays?

Network overlays are virtual networks that are built on top of an existing physical network infrastructure. They provide an additional layer of abstraction, allowing for enhanced flexibility, scalability, and security. By decoupling the logical network from the physical infrastructure, network overlays enable organizations to optimize their network resources and streamline operations.

Benefits of Network Overlays

Improved Scalability:

Network overlays allow for seamless scaling of network resources without disrupting the underlying infrastructure. This means that as your network demands grow, you can easily add or remove virtual network components without affecting the overall network performance.

Enhanced Security:

With network overlays, organizations can implement advanced security measures to protect their data and applications. By creating isolated virtual networks, sensitive information can be shielded from unauthorized access, reducing the risk of potential security breaches.

Simplified Network Management:

Network overlays provide a centralized management interface, allowing administrators to control and monitor the entire network from a single point of control. This simplifies network management tasks, improves troubleshooting capabilities, and enhances overall operational efficiency.

How Network Overlays Work

Overlay Protocols:

Network overlays utilize various overlay protocols such as VXLAN (Virtual Extensible LAN), NVGRE (Network Virtualization using Generic Routing Encapsulation), and GRE (Generic Routing Encapsulation) to encapsulate and transmit data packets across the physical network.

Control Plane and Data Plane Separation:

Network overlays separate the control plane from the data plane. The control plane handles the creation, configuration, and management of virtual networks, while the data plane deals with the actual forwarding of data packets.

Use Cases of Network Overlays

Multi-Tenancy Environments:

Network overlays are highly beneficial in multi-tenant environments, where multiple organizations or users share the same physical network infrastructure. By creating isolated virtual networks, each tenant can have their own dedicated resources while maintaining logical separation.

Data Center Interconnectivity:

Network overlays enable seamless connectivity between geographically dispersed data centers. By extending virtual networks across different locations, organizations can achieve efficient workload migration, disaster recovery, and improved application performance.

Hybrid Cloud Deployments:

Network overlays play a crucial role in hybrid cloud environments, where organizations combine public cloud services with on-premises infrastructure. They provide a unified network fabric that connects the different cloud environments, ensuring smooth data flow and consistent network policies.

Conclusion:

In conclusion, network overlays have revolutionized the networking landscape by providing virtualization and abstraction layers on top of physical networks. Their benefits, including improved scalability, enhanced security, and simplified management, make them an essential component in modern network architectures. As technology continues to evolve, network overlays will undoubtedly play a vital role in shaping the future of networking.

Concept set of mobile network, wireless Internet connection technology. Wifi illustration. People use device to connect Internet network Modern colorful flat vector illustration for poster, banner.

Routing Convergence

August 25, 2014

by Matt Conran Blog

Routing Convergence

Routing convergence, a critical aspect of network performance, refers to the process of network routers exchanging information to update their routing tables in the event of network changes. It ensures efficient and reliable data transmission, minimizing disruptions and optimizing network performance. In this blog post, we will delve into the intricacies of routing convergence, exploring its importance, challenges, and best practices.

Routing convergence refers to the process by which a network's routing tables reach a consistent and stable state after making changes. It ensures that all routers within a network have up-to-date information about the available paths and can make efficient routing decisions.

When a change occurs in a network, such as a link failure or the addition of a new router, routing convergence is necessary to update the routing tables and ensure that packets are delivered correctly. The goal is to minimize the time it takes for all routers in the network to converge and resume normal routing operations.

Several mechanisms and protocols contribute to routing convergence. One of the critical components is the exchange of routing information between routers. This can be done through protocols such as Routing Information Protocol (RIP), Open Shortest Path First (OSPF), or Border Gateway Protocol (BGP).

Matt Conran

Highlights: Routing Convergence

Understanding Convergence

What is router convergence?

Router convergence means they have the same topological information about the network they operate. To converge, a set of routers must have collected all topology information from each other using the routing protocol implemented. For this information to be accurate, it must reflect the current state of the network and not contradict other routers’ topology information. All routers agree upon the topology of a converged network. For dynamic routing to work, a set of routers must be able to communicate with each other. All Interior Gateway Protocols depend on convergence. An autonomous system in operation is usually converged or convergent. Exterior Gateway Routing Protocol BGP rarely converges due to the size of the Internet.

Convergence Process

Each router in a routing protocol attempts to exchange topology information about the network. The extent, method, and type of information exchanged between routing protocols, such as BGP4, OSPF, and RIP, differs. A routing protocol convergence occurs once all routing protocol-specific information has been distributed to all routers. In the event of a routing table change in the network, convergence will be temporarily broken until the change has been successfully communicated to all routers.

Example: The convergence process

During the convergence process, routers exchange information about the network’s topology. Based on this information, they update their routing tables and calculate the most efficient paths to reach destination networks. This process continues until all routers have consistent and accurate routing tables.

The convergence time can vary depending on the size and complexity of the network and the routing protocols used. Convergence can happen relatively quickly in smaller networks, while more extensive networks may take longer to achieve convergence.

Network administrators can employ various strategies to optimize routing convergence. These include implementing fast convergence protocols, such as OSPF’s Fast Hello and Bidirectional Forwarding Detection (BFD), which minimize the time it takes to detect and respond to network changes.

Mechanisms for Achieving Routing Convergence:

1. Routing Protocols:

– Link-State Protocols: OSPF (Open Shortest Path First) and IS-IS (Intermediate System to Intermediate System) are examples of link-state protocols. They use flooding techniques to exchange information about network topology, allowing routers to calculate the shortest path to each destination.

– Distance-Vector Protocols: RIP (Routing Information Protocol) and EIGRP (Enhanced Interior Gateway Routing Protocol) are distance-vector protocols that use iterative algorithms to determine the best path based on distance metrics.

2. Fast Convergence Techniques:

– Triggered Updates: When a change occurs in network topology, routers immediately send updates to inform other routers about the change, reducing the convergence time.

– Route Flapping Detection: Route flapping occurs when a network route repeatedly becomes available and unavailable. By detecting and suppressing flapping routes, convergence time can be significantly improved.

– Convergence Optimization: Techniques like unequal-cost load balancing and route summarization help optimize routing convergence by distributing traffic across multiple paths and reducing the size of routing tables.

3. Redundancy and Resilience:

– Redundant Links: Multiple physical connections between routers increase network reliability and provide alternate paths in case of link failures.

– Virtual Router Redundancy Protocol (VRRP): VRRP allows multiple routers to act as a single virtual router, ensuring seamless failover in case of a primary router failure.

– Multi-Protocol Label Switching (MPLS): MPLS technology offers fast rerouting capabilities, enabling quick convergence in case of link or node failures.

Strategies for Achieving Optimal Routing Convergence

a. Enhanced Link-State Routing Protocol (EIGRP): EIGRP is a dynamic routing protocol that utilizes a Diffusing Update Algorithm (DUAL) to achieve fast convergence. By maintaining a backup route in case of link failures and employing triggered updates, EIGRP significantly reduces the time required for routing tables to converge.

b. Optimizing Routing Metrics: Carefully configuring routing metrics, such as bandwidth, delay, and reliability, can help achieve faster convergence. Assigning appropriate weights to these metrics ensures that routers quickly select the most efficient paths, leading to improved convergence times.

c. Implementing Route Summarization: Route summarization involves aggregating multiple network routes into a single summarized route. This technique reduces the size of routing tables and minimizes the complexity of route calculations, resulting in faster converge

BGP Next Hop Tracking

BGP next hop refers to the IP address of the next router in the path towards a destination network. It serves as crucial information for routers to make forwarding decisions. Typically, BGP relies on the reachability of the next hop to determine the best path. However, various factors can affect this reachability, including link failures, network congestion, or misconfigurations. This is where BGP next hop tracking comes into play.

By incorporating next hop tracking into BGP, network administrators gain valuable insights into the reachability status of next hops. This information enables more informed decision-making regarding routing policies and traffic engineering. With real-time tracking, administrators can dynamically adjust routing paths based on the availability and quality of next hops, leading to improved network performance and reliability.

Benefits of Efficient Routing Convergence:

1. Improved Network Performance: Efficient routing convergence reduces network congestion, latency, and packet loss, improving overall network performance.

2. Enhanced Reliability: Routing convergence ensures uninterrupted communication and minimizes downtime by quickly adapting to changes in network conditions.

3. Scalability: Proper routing convergence techniques facilitate network expansion and accommodate increased traffic demands without sacrificing performance or reliability.

Example Technology: BFD

Bidirectional Forwarding Detection (BFD) is a lightweight protocol designed to detect failures in communication paths between routers or switches. It operates independently of the routing protocols and detects rapid failure by utilizing fast packet exchanges. Unlike traditional methods like hello packets, BFD offers sub-second detection, allowing quicker convergence and network stability. BFD is pivotal in achieving fast routing convergence, providing real-time detection, and facilitating swift rerouting decisions.

-Enhanced Network Resilience: By swiftly detecting link failures, BFD enables routers to act immediately, rerouting traffic through alternate paths. This proactive approach ensures minimal disruption and enhances network resilience, especially in environments where redundancy is critical.

-Reduced Convergence Time: BFD’s ability to detect failures within milliseconds significantly reduces the time required for converging routing protocols. This translates into improved network responsiveness, reduced packet loss, and enhanced user experience.

-Scalability and Flexibility: BFD can be implemented across various network topologies and routing protocols, making it a versatile solution. Whether a small enterprise network or a large-scale service provider environment, BFD adapts seamlessly, providing consistent performance and stability.

Convergence Time

Convergence time measures the speed at which a group of routers converges. Fast and reliable convergent routers are a significant performance indicator for routing protocols. The size of the network is also essential. A more extensive network will converge more slowly than a smaller one.

When a few routers are connected to RIP, a routing protocol that converges slowly, it can take several minutes for the network to converge. A triggered update for a new route can speed up RIP’s convergence, but a hold-down timer will slow flushing an existing route. OSPF is an example of a fast-convergence routing protocol. It is impossible to limit the speed at which OSPF routers can converge.

Unless specific hardware and configuration conditions are met, networks can never converge. “Flapping” interfaces (ones that frequently change between “up” and “down”) propagate conflicting information throughout the network, so routers cannot agree on the current state. Route aggregation can deprive certain parts of a network of detailed routing information, resulting in faster convergence of topological information.

Topological information

A set of routers in a network share the same topological information during convergence or routing convergence. Routing protocols exchange topology information between routers in a network. Routers in a network receive routing information when convergence is reached. Therefore, all routers know the network topology and optimal route in a converged network. Any change in the network – for example, the failure of a device – affects convergence until all routers are informed of the change. The convergence time in a network is the time it takes for routers to achieve convergence after a topology change. In high-performance service provider networks, sensitive applications are run that require fast failover in case of failures. Several factors determine a network’s convergence rate:

Devices detect route failures. Finding a new forwarding path begins with identifying the failed device. The existence of virtual networks establishes device reachability through their longevity, as opposed to physical networks, in which events determine device availability. To achieve fast network convergence, the detection time – the time it takes to detect a failure – must be kept within acceptable limits.
In the event of a device failure on the primary route, traffic is diverted to the backup route. The failure or topology change has not yet affected all devices.
Routing protocols are said to achieve global repair or network convergence when they propagate a change in topology to all network devices.

Understanding UDLD

UDLD, at its core, is a layer 2 protocol designed to detect and mitigate unidirectional links in Ethernet connections. It actively monitors the link status, allowing network devices to promptly detect and address potential issues. By verifying the bidirectional communication between neighboring devices, UDLD acts as a guardian, preventing one-way communication breakdowns.

Implementing UDLD brings forth numerous advantages for network administrators and organizations alike. Firstly, it enhances network reliability by identifying and resolving unidirectional link failures that could otherwise lead to data loss and network disruptions. Secondly, UDLD helps troubleshoot by providing valuable insights into link quality and integrity. This proactive approach aids in reducing downtime and improving overall network performance.

Enhancing Routing Convergence

Network administrators can implement various strategies to improve routing convergence. One approach is to utilize route summarization, which reduces the number of routes advertised and processed by routers. This helps minimize the impact of changes in specific network segments on overall convergence time.

Furthermore, implementing fast link failure detection mechanisms, such as Bidirectional Forwarding Detection (BFD), can significantly reduce convergence time. BFD allows routers to quickly detect link failures and trigger immediate updates to routing tables, ensuring faster convergence.

Factors Influencing Routing Convergence

Several factors impact routing convergence in a network. Firstly, the efficiency of the routing protocols being used plays a crucial role. Protocols such as OSPF (Open Shortest Path First) and EIGRP (Enhanced Interior Gateway Routing Protocol) are designed to facilitate fast convergence by quickly adapting to network changes.

Additionally, network topology and scale can affect routing convergence. Large networks with complex topologies may require more time for routers to converge due to the increased number of routes and potential link failures. Network administrators must carefully design and optimize the network architecture to minimize convergence time.

Control and data plane

When considering routing convergence with forwarding routing protocols, we must first highlight that a networking device is tasked with two planes of operation—the control plane and the data plane. The job of the data plane is to switch traffic across the router’s interfaces as fast as possible, i.e., move packets. The control plane has the more complex operation of putting together and creating the controls so the data plane can operate efficiently. How these two planes interact will affect network convergence time.

The network’s control plane finds the best path for routing convergence from any source to any network destination. For quick convergence routing, it must react quickly and dynamically to changes in the network, both on the LAN and on the WAN.

Monitoring and Troubleshooting Routing Convergence

Network administrators must monitor routing convergence to identify and promptly address potential issues. Network management tools, such as SNMP (Simple Network Management Protocol) and NetFlow analysis, can provide valuable insights into routing convergence performance, including convergence time, route flapping, and stability.

When troubleshooting routing convergence problems, administrators should carefully analyze routing table updates, link state information, and routing protocol logs. This information can help pinpoint the root cause of convergence delays or inconsistencies, allowing for targeted remediation.

Related: For pre-information, you may find the following posts helpful:

Routing Convergence

Convergence Time Definition.

I found two similar definitions of convergence time:

“Convergence is the amount of time ( and thus packet loss ) after a failure in the network and before the network settles into a steady state.” Also, ” Convergence is the amount of time ( and thus packet loss) after a failure in the network and before the network responds to the failure.”

The difference between the two convergence time definitions is subtle but essential – steady-state vs. just responding. The control plane and its reaction to topology changes can be separated into four parts below. Each area must be addressed individually, as leaving one area out results in slow network convergence time and application time-out.

Back to Basics: IP routing

Moving IP Packets

A router’s primary role is moving an IP packet from one network to another. Routers select the best loop-free path in a network to forward a packet to its destination IP address. A router learns about nonattached networks through static configuration or dynamic IP routing protocols. Both static and dynamic are examples of routing protocols.

With dynamic IP routing protocols, we can handle network topology changes dynamically. Here, we can distribute network topology information between routers in the network. When there is a change in the network topology, the dynamic routing protocol provides updates without intervention when a topology change occurs.

On the other hand, we have IP routing to static routes, which do not accommodate topology changes very well and can be a burden depending on the network size. However, static routing is a viable solution for minimal networks with no modifications.

Diagram: Dynamic Routing Protocols. Source Cisco Press.

Knowledge Check: Bidirectional Forwarding Detection

Understanding BFD

BFD is a lightweight protocol designed to detect faults in the forwarding path between network devices. It operates at a low level, constantly monitoring the connectivity and responsiveness of neighboring devices. BFD can quickly detect failures by exchanging control packets and taking appropriate action to maintain network stability.

The Benefits of BFD

The implementation of BFD brings numerous advantages to network administrators and operators. Firstly, it provides rapid fault detection, reducing downtime and minimizing the impact of network failures. Additionally, BFD offers scalable and efficient operation, as it consumes minimal network resources. This makes it an ideal choice for large-scale networks where resource optimization is crucial.

BFD runs independently from other (routing) protocols. Once it’s up and running, you can configure protocols like OSPF, EIGRP, BGP, HSRP, MPLS LDP, etc., to use BFD for link failure detection instead of their mechanisms. When the link fails, BFD informs the protocol. When BFD no longer receives its control packets, it realizes we have a link failure and reports this to OSPF. OSPF will then tear down the neighbor adjacency.

Use Cases of BFD

BFD finds its applications in various networking scenarios. One prominent use case is link aggregation, where BFD helps detect link failures and ensures seamless failover to alternate links. BFD is also widely utilized in Virtual Private Networks (VPNs) to monitor the connectivity of tunnel endpoints, enabling quick detection of connectivity issues and swift rerouting.

Implementing BFD in Practice

Implementing BFD requires careful consideration and configuration. Network devices must be appropriately configured to enable BFD sessions and define appropriate parameters such as timers and thresholds. Additionally, network administrators must ensure proper integration with underlying routing protocols to maximize BFD’s efficiency.

Convergence Routing and Network Convergence Time

Network convergence connects multiple computer systems, networks, or components to establish communication and efficient data transfer. However, it can be a slow process, depending on the size and complexity of the network, the amount of data that needs to be transferred, and the speed of the underlying technologies.

For networks to converge, all of the components must interact with each other and establish rules for data transfer. This process requires that the various components communicate with each other and usually involves exchanging configuration data to ensure that all components use the same protocols.
Network convergence is also dependent on the speed of the underlying technologies.

To speed up convergence, administrators should use the latest technologies, minimize the amount of data that needs to be transferred, and ensure that all components are correctly configured to be compatible. By following these steps, network convergence can be made faster and more efficient.

Example: OSPF

To put it simply, convergence or routing convergence is a state in which a set of routers in a network share the same topological information. For example, we have ten routers in one OSFP area. OSPF is an example of a fast-converging routing protocol. A network of a few OSPF routers can converge in seconds.

The routers within the OSPF area in the network collect the topology information from one another through the routing protocol. Depending on the routing protocol used to collect the data, the routers in the same network should have identical copies of routing information.

Different routing protocols will have additional convergence time. The time the routers take to reach convergence after a change in topology is termed convergence time. Fast network convergence and fast failover are critical factors in network performance. Before we get into the details of routing convergence, let us recap how networking works.

network convergence time — Diagram: Network convergence time.

Unlike IS-IS, OSPF has fewer “knobs” for optimizing convergence. This is probably because IS-IS is being developed and supported by a separate team geared towards ISPs, where fast convergence is a competitive advantage.

Diagram: Example Convergence Time with OSPF. Source INE.

OSPF: Incremental SPF

OSPF calculates the SPT (Shortest Path Tree) using the SPF (Shortest Path First) algorithm. SPTs are built by OSPF routers within the same area with the same LSAs, LSDBs, and LSAs. OSPF routers will rerun a full SPF calculation even when there is just a single change in the network topology (change to an LSA type 1 and LSA type 2).

If a topology change occurs, we should run a full SPT calculation to find the shortest paths to all destinations. Unfortunately, we also calculate paths that have not changed since the last SPF.

In incremental SPF, OSPF only recalculates the parts of the SPT that have changed.

Because you don’t run a full SPF all the time, the router’s CPU load decreases,, and convergence times improve—additionally, your router stores the previous SPT copy, which requires more memory.

In three scenarios, incremental SPF is beneficial:

Adding (or removing) a leaf node to a branch
Link failure in non-SPT
Link failure in a branch of SPT

When many routers are in a single area, and the CPU load is high because of OSPF, incremental SPF can be enabled per router.

Forwarding Paradigms

We have bridging routing and switching with data and the control plane. So, we need to get packets across a network, which is easy if we have a single cable. You need to find the node’s address, and small and non-IP protocols would use a broadcast. When devices in the middle break this path, we can use source routing, path-based forwarding, and hop-by-hop address-based forwarding based solely on the destination address.

When protocols like IP came into play, hop-by-hop destination-based forwarding became the most popular; this is how IP forwarding works. Everyone in the path makes independent forwarding decisions. Each device looks at the destination address, examines its lookup tables, and decides where to send the packet.

Finding paths across the network

How do we find a path across the network? We know there are three ways to get packets across the network – source routing, path-based forwarding, and hop-by-hop destination-based forwarding. So, we need some way to populate the forwarding tables. You need to know how your neighbors are and who your endpoints are. This can be static routing, but it is more likely to be a routing protocol. Routing protocols have to solve and describe the routing convergence on the network at a high level.

When we are up and running, events can happen to the topology that force or make the routing protocols react and perform a convergence routing state. For example, we have a link failure, and the topology has changed, impacting our forwarding information. So, we must propagate the information and adjust the path information after the topology change. We know these convergence routing states are to detect, describe, switch, and find.

Rouitng Convergence	Convergence
Detect
Describe
Switch
Find

To better understand routing convergence, I would like to share the network convergence time for each routing protocol before diving into each step. The times displayed below are from a Cisco Live session based on real-world case studies and field research. We are separating each of the convergence routing steps described above into the following fields: Detect, describe, find alternative, and total time.

Routing Protocol	RIP	OSPF	EIGRP
Detect	<1 second-best, 105 seconds average	<1 second-best, 20 seconds average	<1 second-best, 15 seconds average.30 seconds worst
Describe	15 seconds average, 30 seconds worst	1 second-best, 5 seconds average.	2 seconds
Find Alternative	15 seconds average, 30 seconds worst	1-second average.	*** <500ms per query hop average Assume a 2-second average
Total Time	Best Average Case: 31 seconds Average Case: 135 seconds Worse Case: 179 seconds	Best Average Case: 2 to 3 seconds Average Case: 25 seconds Worse Case: 45 seconds	Best Average Case: <1 second Average Case: 20 seconds Worse Case: 35 seconds

*** The alternate route is found before the described phase due to the feasible successor design with EIGRP path selection.

Convergence Routing

Convergence routing: EIGPR

EIGRP is the fastest but only fractional. EIGRP has a pre-built loop-free path known as a feasible successor. The FS route has a higher metric than the successor, making it a backup route to the successor route. The effect of a pre-computed backup route on convergence is that EIGRP can react locally to a change in the network topology; nowadays, this is usually done in the FIB. EIGRP would have to query for the alternative route without a feasible successor, increasing convergence time.

However, you can have a Loop Free Alternative ( LFA ) for OSPF, which can have a pre-computed alternate path. Still, LFAs can only work with specific typologies and don’t guarantee against micro-loops ( EIGRP guarantees against micro-loops).

Lab Guide: EIGRP LFA FRR

With Loop-Free Alternate (LFA) Fast Reroute (FRR), EIGRP can switch to a backup path in less than 50 milliseconds. Fast rerouting means switching to another next hop, and a loop-free alternate refers to a loop-free alternative path.

Perhaps this sounds familiar to you. After all, EIGRP has feasible successors. The alternate paths calculated by EIGRP are loop-free. As soon as the successor fails, EIGRP can use a feasible successor.

It’s true, but there’s one big catch. In the routing table, EIGRP feasible successors are not immediately installed. Only one route is installed, the successor route. EIGRP installs the feasible successor when the successor fails, which takes time. By installing both successor routes and feasible successor routes in the routing table, fast rerouting makes convergence even faster.

These four routers run EIGRP; there’s a loopback on R4 with network 4.4.4.4/32. R1 can go through R2 or R3 to get there. The delay on R1’s GigabitEthernet3 interface has increased, so R2 is our successor, and R3 is our feasible successor. The output below is interesting. We still see the successor route, but at the bottom, you can see the repair path…that’s our feasible successor.

TCP Congestion control

Ask yourself, is < 1-second convergence fast enough for today’s applications? Indeed, the answer would be yes for some non-critical applications that work on TCP. TCP has built-in backoff algorithms that can deal with packet loss by re-transmitting to recover lost segments. However, non-bulk data applications like video and VOIP have stricter rules and require fast convergence and minimal packet loss.

For example, a 5-second delay in routing protocol convergence could mean several hundred dropped voice calls. A 50-second delay in a Gigabit Ethernet link implies about 6.25 GB of lost information.

Adding Resilience

To add resilience to a network, you can aim to make the network redundant. When you add redundancy, you are betting that outages of the original path and the backup path will not co-occur and that the primary path does not fate share with the backup path ( they do not share common underlying infrastructure, i.e., physical conducts or power ).

There needs to be a limit on the number of links you add to make your network redundant, and adding 50 extra links does not make your network 50 times more redundant. It does the opposite! The control plane is tasked with finding the best path and must react to modifications in the network as quickly as possible.

However, every additional link you add slows down the convergence of the router’s control plane as there is additional information to compute, resulting in longer convergence times. The correct number of backup links is a trade-off between redundancy versus availability. The optimal level of redundancy between two points should be two or three links. The fourth link would make the network converge slower.

Convergence Routing — Diagram: Convergence routing and adding resilience.

Routing Convergence and Routing Protocol Algorithms

Routing protocol algorithms can be tweaked to exponentially back off and deal with bulk information. However, no matter how many timers you use, the more data in the routing databases, the longer the convergence times. The primary way to reduce network convergence is to reduce the size of your routing tables by accepting just a default route, creating a flooding boundary domain, or using some other configuration method.

For example, a common approach in OSPF to reduce the size of routing tables and flooding boundaries is to create OSPF stub areas. OSPF stub areas limit the amount of information in the area. For example, EIGRP limits the flooding query domain by creating EIGRP stub routers and intelligently designing aggregation points. Now let us revisit the components of routing convergence:

Routing Convergence Step	Routing Convergence Details
Step 1	Failure detection
Step 2	Failure propagation ( flooding, etc.) IGP Reaction
Step 3	Topology/Routing calculation. IGP Reaction.
Step 4	Update the routing and forwarding table ( RIB & FIB)

Stage 1: Failure Detection

The first and foremost problem facing the control plane is quickly detecting topology changes. Detecting the failure is the most critical and challenging part of network convergence. It can occur at different layers of the OSI stack – Physical Layers ( Layer 1), Data Link Layer ( Layer 2 ), Network Layer ( Layer 3 ), and Application layer ( Layer 7 ). There are many types of techniques used to detect link failures, but they all generally come down to two basic types:

Event-driven notification occurs when a carrier is lost or when one network element detects a failure and notifies the other network elements.
Polling-driven notification – generally HELLO protocols that test the path for reachability, such as Bidirectional Forwarding Detection ( BFD ). Event-driven notifications are always preferred over polling-driven ones as the latter have to wait for three polls before declaring a path down. However, there are some cases when you have multiple Layer devices in the path, and HELLO polling systems are the only method that can be used to detect a failure.

Layer 1 failure detection

Layer 1: Ethernet mechanisms like auto-negotiation ( 1 GigE ) and link fault signaling ( 10 GigE 802.3ae/ 40 GigE 802.3ba ) can signal local failures to the remote end.

However, the challenge is getting the signal across an optical cloud, as relaying the fault information to the other end is impossible. When there is a “bump” in the Layer 1 link, it is not always possible for the remote end to detect the failure. In this case, the link fault signaling from Ethernet would get lost in the service provider’s network.

The actual link-down / interface-down event detection is hardware-dependent. Older platforms, such as the 6704 line cards for the Catalyst 6500, used a per-port polling mechanism, resulting in a 1 sec detect link failure period. More recent Nexus switches and the latest Catalyst 5600 line cards have an interrupt-driven notification mechanism resulting in high-speed and predictable link failure detection.

Layer 2 failure detection

Layer 2: The layer 2 detection mechanism will kick in if the Layer 1 mechanism does not. Unidirectional Link Detection ( UDLD ) is a Cisco proprietary lightweight Layer 2 failure detection protocol designed for detecting one-way connections due to physical or soft failure and miss-wirings.

A key point: UDLD is a slow protocol

UDLD is a reasonably slow protocol that uses an average of 15 seconds for message interval and 21 seconds for detection. Its period has raised questions about its use in today’s data centers. However, the chances of miswirings are minimal; Layer 1 mechanisms always communicate unidirectional physical failure, and STP Bridge Assurance takes care of soft failures in either direction.

STP Bridge assurance turns STP into a bidirectional protocol and ensures that the spanning tree never fails to open and only fails to close. Failing open means that if a switch does not hear from its neighbor, it immediately starts forwarding on initially blocked ports, causing network havoc.

Layer 3 failure detection

Layer 3: In some cases, failure detection has to reply to HELLO protocols at Layer 3. This is needed when there are intermediate Layer 2 hops over Layer links and concerns over uni-direction failures on point-to-point physical links arise.

All Layer 3 protocols use HELLOs to maintain neighbor adjacency and a DEAD time to declare a neighbor dead. These timers can be tuned for faster convergence. However, it is generally not recommended due to the increase in CPU utilization causing false positives and the challenges ISSU and SSO face. They enable bidirectional forwarding detection ( BFD ) as the layer 3 detection mechanism is strongly recommended over aggressive protocol times, and they use BFD for all protocols.

Bidirectional Forwarding Detection ( BFD ) is a lightweight hello protocol for sub-second Layer 3 failure detection. It can run over multiple transport protocols, such as MPLS, THRILL, IPv6, and IPv4, making it the preferred method.

Stage 2: Routing convergence and failure propagation

When a change occurs in the network topology, it must be registered with the local router and transmitted throughout the rest of the network. The transmission of the change information will be carried out differently for Link-State and Distance Vector protocols. Link state protocols must flood information to every device in the network, and the distance vector must process the topology change at every hop through the network.

The processing of information at every hop may lead you to conclude that link-state protocols always converge more quickly than path-vector protocols, but this is not the case. EIGRP, due to its pre-computed backup path, will converge more rapidly than any link-state protocol.

Diagram: Routing convergence and failure propagation

To propagate topology changes as quickly as possible, OSPF ( Link state ) can group changes into a few LSA while slowing down the rate at which information is flooded, i.e., do not flood on every change. This is accomplished with link-state flood timer tuning combined with exponential backoff systems, such as link-state advertisement delay / initial link-state advertisement throttle delay.

Unfortunately, Distance Vector Protocols do not have such timers. Therefore, reducing the routing table size is the only option for EIGRP. This can be done by aggregating and filtering reachability information ( summary route or Stub areas ).

Stage 3: Topology/Routing calculation

Similar to the second step, link-state protocols use exponential back-off timers in this step. These timers adjust the waiting time for OSPF and ISIS to wait after receiving new topology information before calculating the best path.

Stage 4: Update the routing and forwarding table ( RIB & FIB)

Finally, after the topology information has been flooding through the network and a new best path has been calculated, the new best path must be installed in the Forwarding Information Base ( FIB ). The FIB is a copy of the RIB in hardware, and the forwarding process finds it much easier to read than the RIB. However, again, this is usually done in hardware. Most vendors offer features that will install a pre-computed backup path on the line cards forwarding table so the fail-over from the primary path to the backup path can be done milliseconds without interrupting the router CPU.

Closing Points: Routing Convergence

Routing convergence refers to the process by which network routers exchange routing information and adapt to changes in network topology or routing policies. It involves the timely update and synchronization of routing tables across the network, allowing routers to determine the best paths for forwarding data packets.

Routing convergence is vital for maintaining network stability and minimizing disruptions. Network traffic may experience delays, bottlenecks, or even failures without proper convergence. Routing convergence enables efficient and reliable communication by ensuring all network routers have consistent routing information.

Summary: Routing Convergence

Routing convergence is crucial in network management, ensuring smooth and efficient communication between devices. In this blog post, we will explore the concept of routing convergence, its importance in network operations, common challenges faced, and strategies to achieve faster convergence times.

Section 1: Understanding Routing Convergence

Routing convergence refers to network protocols adapting to changes in network topology, such as link failures or changes in network configurations. It involves recalculating and updating routing tables to ensure the most optimal paths for data transmission. Network downtime can be minimized by achieving convergence quickly, and data can flow seamlessly.

Section 2: The Importance of Fast Convergence

Fast routing convergence is critical for maintaining network stability and minimizing disruptions. In today’s fast-paced digital landscape, where businesses rely heavily on uninterrupted connectivity, delays in convergence can result in significant financial losses, degraded user experience, and even security vulnerabilities. Therefore, network administrators must prioritize measures to enhance convergence speed.

Section 3: Challenges in Routing Convergence

While routing convergence is essential, it comes with its challenges. Network size, complex topologies, and diverse routing protocols can significantly impact convergence times. Additionally, suboptimal route selection, route flapping, and inefficient link failure detection can further hinder the convergence process. Understanding these challenges is crucial for devising practical solutions.

Section 4: Strategies for Achieving Faster Convergence

To optimize routing convergence, network administrators can implement various strategies. These include:

1. Implementing Fast Convergence Protocols: Utilizing protocols like Bidirectional Forwarding Detection (BFD) and Link State Tracking (LST) can expedite the detection of link failures and trigger faster convergence.

2. Load Balancing and Redundancy: Distributing traffic across multiple paths and employing redundancy mechanisms, such as Equal-Cost Multipath (ECMP) routing, can mitigate the impact of link failures and improve convergence times.

3. Optimizing Routing Protocol Parameters: Fine-tuning routing protocol timers, hello intervals, and dead intervals can contribute to faster convergence by reducing the time it takes to detect and react to network changes.

Section 5: Conclusion

In conclusion, routing convergence is fundamental to network management, ensuring efficient data transmission and minimizing disruptions. By understanding the concept, recognizing the importance of fast convergence, and implementing appropriate strategies, network administrators can enhance network stability, improve user experience, and safeguard against potential financial and security risks.

Posts tagged: control plane