A data mesh architecture is a network design that allows data to be routed across multiple paths between network nodes. Whereas traditional network designs routed data along a single path, data mesh architecture enables data to be routed simultaneously across multiple paths between nodes.
This architecture provides several advantages over other architectures, such as a hub-and-spoke or partial mesh topology. Data mesh architectures are composed of three core elements: hubs, spokes, and links. The hubs are central devices that manage the routing paths for the spokes. The spokes are the network devices that connect the hubs and their neighbors. The links are the physical or logical connections between the spokes.
This article will focus on data mesh architecture and its benefits for enterprise networks. You’ll learn about the different components of data mesh, key considerations when designing one, sample architectures, and more about this topic.
The Benefits of Using a Data Mesh Architecture
The data mesh architecture is used to build data center wide area networks (WANs) that are scalable, simplified, and resilient. This architecture provides three core benefits – simplicity, scalability, and reliable remote connectivity – which are essential for IT organizations of any size seeking to provide high-quality services to their customers.
Additionally, a data mesh network topology provides more options for network traffic routing, which enables the network to scale with your company’s needs. It also provides more options for redundancy, which enables the network to stay up even if one of the links or nodes fails.
With a data mesh network, you can use a variety of redundant routing protocols and techniques to make the network more reliable. Also, the data ownership is shifted from the central data team to domain teams.
In addition to the above benefits, a data mesh network has a lower cost than hybrid or fully meshed solutions. A data mesh network provides better resiliency than a hub-and-spoke architecture since data can traverse multiple links between nodes. This means that if one link fails, data can still be routed through other links in the network. Hub-and-spoke architectures have only one path for data to travel between nodes, so if that path fails, the entire network goes down.
A data mesh network provides better resiliency than an end-to-end solution since it doesn’t require every pair of nodes to be directly connected. This means that the data mesh architecture needs fewer links than an end-to-end solution.
The data mesh removes the need for data transportation by offering data query at the source. The primary goal of data mesh is to facilitate easy access to data at scale. In a nutshell, data mesh makes data accessible, available, discoverable, secure, and interoperable.
The Core Components of a Data Mesh Architecture
Before designing a data mesh network, it’s important to understand the core components that make it up. The data mesh architecture includes hub nodes, spokes, links, and routing protocols.
- Hub nodes: These are the central devices that are responsible for managing the routing paths for all the spokes. Hub nodes can also be responsible for enforcing policies such as quality of service (QoS) or security. Hub nodes can be implemented in a variety of ways, such as with a hardware device or a software-based device.
- Spokes: The spokes are the network devices that are used to connect the hubs and their neighbors. They are responsible for routing network traffic and sending data to the hub nodes. Spokes can be implemented in a variety of ways, such as with a hardware device or a software-based device.
- Links: The links are the physical or logical connections between the spokes. These connections can be a combination of fiber-optic cables, copper wires, or a software connection.
- Routing protocols: Routing protocols exchange data between the hubs and the spokes. These protocols are responsible for ensuring network traffic is routed properly between the devices.
Designing and Implementing a Data Mesh Architecture
When designing a data mesh network, organizations must consider several factors. These include the topology of the network, the traffic flow between the devices, and the number of devices used. The topology of the network refers to the physical layout of the network: You need to determine where the hubs, spokes, and links should be physically located. Keep in mind that hubs and spokes should be as close as possible to one another to minimize the network distance between them.
The traffic flow between the devices refers to how data is routed between the hubs and the spokes. This can vary depending on the routing protocols used in the network. The number of devices used refers to the number of hubs, spokes, and links used in the network. This is typically determined by the network architecture and how many devices are in the network.
You can create a simple diagram of your network to help you design it:
- Implementing hub nodes: These are the central devices responsible for managing the routing paths for all the spokes. The number of hubs in your network will depend on the number of nodes in your network. You can determine the number of hubs based on the number of nodes and the distance between them.
- Implementing spokes: The number of spokes in your network will depend on the number of nodes in your network. The number of spokes can also depend on the number of hubs in your network. For example, if you have 10 nodes and five hubs, you need 10 spokes.
- Implementing links: The number of links in your network will depend on the number of nodes and the distance between them. For example, if you have 10 nodes, each with a distance of 100 kilometers to the nearest hub, you need a total of 10 links.
- Implementing routing protocols: You also need to select the routing protocols for your network. You can select any routing protocols that are available in your network.
Drawbacks of a Data Mesh Architecture
A data mesh network architecture has many advantages, but it also has some drawbacks. The biggest drawback of a data mesh network is that it requires more devices than an end-to-end solution. This means the network will be larger, more complex, and more expensive to implement than an end-to-end solution.
A data mesh network also provides lower performance than an end-to-end solution since data must traverse multiple hops before reaching its destination. This increased latency and reduced bandwidth are both disadvantages of a data mesh network.
As networking technologies continue to advance, the need for scalable and reliable data center WANs grows. To build these networks, enterprises need an architecture that can scale with their needs and provide simplified operations. The data mesh architecture is a network design that provides all of these benefits.
The data mesh architecture can be implemented in a variety of ways and provides better resiliency, simplified operations, and improved scalability than a hub-and-spoke architecture.
Image used under license from Shutterstock.com