How can you choose a routing protocol that balances scalability and performance?

BGP is definitely my favorite, but its main focus is to maintain AS reachability and not to provide fast convergence. It also provides routing for multi protocols with mpbgp (mpls VPNs and traffic engineering depend a lot on BGP, not to mention other features). On the other hand IS-IS has been growing exponentially with the advance of software defined networks, since it can also carry non L3 info and tunnel traffic even between a switch and an access-point for example (observed with SDA for LAN) and ACI for data centers, with the usage of VTEPs on the spine/leaf model. It all depends on what is the goal. Each network is its own unique universe and multiple routing protocols can be used for convergence, security, scalability, etc.

If fast convergence is not a must, I would choose BGP. It's a incredibly scalable protocol, has powerful route (path) control, easy to maintain. It can operate as IGP also.

Choosing the right protocols for a network depends on the requirements. A scalable routing protocol efficiently handles network growth, so it should perform well as the network expands. You must know the current number of nodes and a plan for future expansion. Performance is related to how well the protocol transports network traffic between nodes while maintaining adequate quality of service. Higher performance often comes with higher router costs. Link-state protocols (e.g., OSPF, IS-IS) offer a good balance of scalability and performance for Interior Gateway Protocols (IGP) within a single network domain. For Exterior Gateway Protocols (EGP) connecting different domains, Border Gateway Protocol (BGP) is the de facto standard.

Choosing a routing protocol balancing scalability and performance is critical. For small networks, IGPs like OSPF or EIGRP offer efficiency. Larger networks, spanning ASes, benefit from EGPs like BGP, designed for scalability. IGPs excel in quick convergence and resource efficiency, ideal for smaller environments. BGP's policy control suits larger networks with global needs. Evaluating size, convergence speed, administrative overhead, and resource utilization helps administrators make a balanced choice for optimal network performance.

A lot of people are doing BGP end to end, it keeps it simple, allows a lot control and is standards based. There are always new features/functionalities being added for instance AFI/SAFI. It is extremely scalable and flexible

Little is known and being a great lover of the protocol, but ISIS becomes attractive for Service Providers mainly because it has delay and reliability metrics in addition to the link bandwidth. However, it must be implemented on each device and interface. But for each topology and network, the scenario needs to be analyzed not only thinking about the links. But in terms of transport capacity and number of routes as well.

parfait ??. pour tout dire le choix d'un protocole de routage depend deja de ce que l'on veut faire ,c'est a dire circonstancielle (projet sous la main) et se rassurer que la métrique qui va avec est un chemin plus court pour rendre plus efficace et performant votre réseau

This is often underrated by many companies, although, regularly updating and maintaining routing protocol metrics is very important. Yes initial configuration is important too, but ongoing adjustments can significantly improve network performance and on the long run.

Common Metrics: Hop Count: Think of it as a city map, counting the number of intersections (routers) a packet must cross to reach its destination. The shortest path, in terms of hops, is often favored. Bandwidth: This metric measures available road width (bandwidth) to accommodate data traffic. Delay: Picture a stopwatch measuring time spent at each intersection. Cost: Imagine toll roads with varying fees. Cost metrics assign financial values to routes, often based on bandwidth or distance, to optimize budget-friendly paths. Reliability: Picture a road with potholes and detours. Reliability metrics favor routes with stable connections and minimal disruptions, ensuring smooth data journeys. Load: Think of rush hour congestion.

When it comes to metrics in routing protocols, nothing so easy and intuitive like IS-IS. It becomes very easy to take the traffic out of a link. But IS-IS is also feared by network engineers, but having a deep dive into it, you will love it. Very popular for Service Providers.

My 2 cents to consider regarding to Routing protocol convergence, talking about the protocol defaults convergence timers: - Use Equal-Cost Multipath (ECMP) - Have a Loop-Free Alternate Path (LFA) and - Use BFD in everywhere you can! Regardless of the protocol you’ve chosen, the defaults timers to choose and forward the packets across the alternative path is so long and using all 3 approaches above, you could do a “fail-over” in up to 50 milliseconds. It’s better, isn’t it? :D

One routing protocol known for scalability and performance is OSPF (Open Shortest Path First). It's widely used in large networks due to its efficient routing algorithm and ability to scale well as network size grows.

This is when I say that considering the number of routes being sent and received plus link flaps, topology changes and other things, the protocol that will be used in the network mesh must be taken into account.

OSPF and EIGRP are two popular routing protocols that are widely used in enterprise networks. In terms of convergence time, EIGRP is faster than OSPF because it uses a Diffusing Update Algorithm (DUAL) to calculate the shortest path to a destination, avoiding network loops. On the other hand, OSPF uses the Shortest Path First (SPF) algorithm, which takes longer to converge.

Fast convergence is pretty important for network stability. Choosing a protocol that will quickly adapt to changes can really minimize downtime.

Number of areas supported by any one router: A router must run the link-state algorithm for each link-state change that occurs for every area in which the router resides. Every ABR is in at least two areas (the backbone and one adjacent area). In general, to maximize stability, one router should not be in more than three areas.

The biggest factor when it comes to scalability is resource utilization. Note that with longer routes and a lot of circuit switching, this can have a major impact on links and hardware. As well as the perception of failures. At some point, any routing protocol may not support so many routes in the backbone and the use of new scenarios must be used, such as Seamless MPLS.

Factors Influencing OSPF Scalability Scaling is determined by the utilization of three router resources: memory, CPU, and interface bandwidth.

Number of adjacent neighbors for any one router: OSPF floods all link-state changes to all routers in an area. Routers with many neighbors have the most work to do when link-state changes occur. In general, any one router should have no more than 60 neighbors.

Number of adjacent routers in an area: OSPF uses a CPU-intensive algorithm. The number of calculations that must be performed given n link-state packets is proportional to n log n. As a result, the larger and more unstable the area, the greater the likelihood for performance problems associated with routing protocol recalculation. Generally, an area should have no more than 50 routers. Areas that suffer with unstable links should be smaller.

Dans le domaine informatique, la sécurité du protocole de routage revêt une importance cruciale pour garantir l'intégrité, la confidentialité et la fiabilité des informations de routage. Face aux menaces externes, la robustesse des protocoles de routage devient essentielle, nécessitant une protection efficace contre les accès non autorisés et les attaques malveillantes. La détection proactive des risques potentiels et la mise en place de mesures d'atténuation renforcent la résilience du réseau, contribuant ainsi à maintenir la stabilité et la confiance dans le fonctionnement des réseaux informatiques. En somme, la sécurisation des protocoles de routage demeure un élément clé pour assurer un environnement réseau s?r et fiable.

The selection for routing protocols depends on the size and the complexity of the network. Traditionally, the common key factors for choosing the protocol, such as whether the network devices are homo- or heterogeneous, how large is the network, the security context of the routing protocol, and others. Nowadays, I think that the modern large enterprise network architecture start using virtual networking technologies, so the new types of addressing scheme for the fabric network needs a dynamic routing protocol that is capable to carry new addressing types relevant to the modern technology like LISP, VRF, SGT.

La seguridad en los protocolos de enrutamiento es un aspecto crucial, caracteristicas en los protocolos: 1. Border Gateway Protocol (BGP) * Hijacking y Spoofing de Rutas * BGPsec * Filtrado de Rutas * Autenticación de Peers 2. Open Shortest Path First (OSPF) * áreas y Dise?o de Red * Autenticación * OSPFv3 y Seguridad de IPv6 3. Intermediate System to Intermediate System (IS-IS) * Autenticación de Mensajes * Seguridad a Nivel de Red 4. Enhanced Interior Gateway Routing Protocol (EIGRP) * Autenticación de Rutas * Encriptación Prácticas Comunes de Seguridad en el Enrutamiento * Políticas de Seguridad y Monitoreo Continuo * Autenticación y Encriptación * Control de Acceso y Filtrado * Actualizaciones y Parches Regulares

Our traditional packet forwarding process is based on the destination IP/Prefix only, I mean, the router checks only that field. In a security perspective, if you DON’T VALIDATE the SOURCE, it doesn’t matter which protocol you’ve chosen.?? If you don’t validate the ORIGIN of the packet, your network probably will be member of a DDoS attack – for example, one user has spoofed a source-IP, make a simple request for a NTP update and wait for the answer from the server. Then, server will reply to the spoofed IP causing an overload in case requested many times from many clients. ?? Then, I strongly recommend: Always validate the ORIGIN to enforce your routing-protocol – there are since a simple Route-filters (ACL) until RPF-Check and RPKI. ??

We need also to consider the stronger security features for our network protocols in order to have a stronger protection against the network attacks

Technical comparisons between protocols don't give you all the information you need. Even if RIP is able to provide you with all your needs, should you choose a protocol that is usually avoided, and thus less and less tested/supported? ISIS does not have the weakness of OSPF's inter-area handling, but has operational requirements that may be more cumbersome than simply observing OSPF's requirements. If I am free to choose, I'd pick ISIS for IGP and BGP (what else?) for EGP. I would go against EIGRP due to vendor lock-in. The network should not depend on specific features only a particular product or vendor can provide for the same reason companies should avoid rushing to move to clouds: If you want to change later, it's going to cost a lot.

EIGRP has the fastest convergence time among the popular routing protocols. It uses a Diffusing Update Algorithm (DUAL) to calculate the shortest path to a destination, avoiding network loops. This algorithm allows EIGRP to quickly recalculate routes if a link or router fails, resulting in a shorter convergence time. On the other hand, OSPF uses the Shortest Path First (SPF) algorithm, which takes longer to converge.

There is no one-size-fits-all solution when it comes to routing protocols. The best choice depends on the specific needs and characteristics of your network. Some of the most common routing protocols include RIP, IGRP, EIGRP, OSPF, IS-IS, and BGP. Each of these has its own strengths and weaknesses, and the choice between them depends on the specific requirements of your network.

RIP (Routing Information Protocol): - Simple, distance-vector protocol. - Limited scalability and convergence speed. OSPF (Open Shortest Path First): - Link-state protocol with fast convergence and efficient path calculation. - Suitable for medium to large enterprise networks. - More complex configuration and resource requirements. EIGRP (Enhanced Interior Gateway Routing Protocol): - Cisco proprietary protocol combining distance-vector and link-state features. - Fast convergence and low overhead. BGP (Border Gateway Protocol): - Path-vector protocol for inter-domain routing (Internet backbone). - Highly scalable and supports complex routing policies. - IS-IS (Intermediate System to Intermediate System): Link-state protocol similar to OSPF.

An often overlooked factor when comparing routing protocols is the experience and knowledge of the operating staff. If they know about an specific protocol and are comfortable operating it, this may greatly reduce the associated costs and downtimes during the operation.

Overall, how you configure your routing protocol is going to matter a lot more than which you choose. So be sure to focus on learning deeply, designing thoughtfully, and implementing carefully.

Protocol choice usually matters much less than how you "Day 2 Operate" it. Avoid the faux sense of accomplishment that picking a Routing Protocol will solve world hunger for your Network Design; you can get more mileage out of RIP or EIGRP than you think. Not OSPF though; there's only three Types of people in that world - Type 1 Incrementing; Type 2 Staying-the-same and Type 3...

Routing protocols are typically chosen for an organization and supported across the enterprise. OSPF Open Shortest Path First is the most popular for Interior use, where BGP Border Gateway Protocol is external typically to the Internet or large enterprise or carrier network. One of the biggest issues internally is the Spanning Tree Protocol that prevents loops on a multipath Ethernet Switched network that connects systems at layer two. Many issues thought to be caused by a routing protocol can be caused by layer two connectivity between router ports on the same network segment.

I agree with Kamran, that we should start looking network size, topology and traffic patterns, consider present situation and project it to one, two or five years. Look for protocol that best suits your network, for IGP is very common to use OSPF, IS-IS, for EGP consider BGP. In all these, take consideration on your routing/switching providers. Last but not least ,consider what your team know.

In my experience, the choice of unicast routing protocol has already been converged. For Ent IGP: OSPF For SP IGP: ISIS. For EGP or VPN signaling on both Ent/SP: BGP. Those inferior and/or proprietary protocols are losing their ground just like we no longer talk IPX and AppleTalk. The only doubt I still have is to choose the multicast routing protocol. Should I use PIM-DM/PIM-SM/PIM-SSM/DVMRP... Here really do not have one size fit all situation.