The term network device covers a lot of ground. Anything from a simple unmanaged switch to a dial-up modem to a UTM (unified threat management) appliance can fit the bill. Additionally, the lines between what category different networking devices fall into can get fuzzy because in many cases a single device performs multiple functions.
Here, to help you sort through that fuzziness, we tackle the concept of network devices from the ground up. We’ll start with a definition, move into our network devices list, and then discuss some of the important points to consider when managing network devices.
What are network devices?
Network devices are building blocks that facilitate communication between services and the endpoints that consume those services. In other words, they’re connectors that enable communications between devices on a network.
You are watching: Which of the following hardware devices links multiple networks
Enabling communication means anything that helps data get from source to destination— a network switch transmitting frames between two devices on a LAN, for example.
Restricting communication means stopping that communication from occurring. Examples here include a firewall rule that stops traffic from going to TCP port 23 or a VLAN that breaks up broadcast domains.
You’ll notice the word “hardware” is missing from our network devices definition. This is because many network functions and appliances are virtualized or software-based. For example, you can deploy a virtual appliance that acts as a router, network switch, or firewall.
OSI model and data types crash course
One of the best ways to understand the purpose of different network devices is to understand a bit about the layers of the OSI (Open Systems Interconnection) model. The OSI model provides a conceptual model that explains how data flows across and within networks.
For understanding how the different network devices work, these two aspects of the OSI model will help:
The layer: The OSI model has seven layers starting at the physical layer (Layer 1) and going up to the application layer (Layer 7). Often, you’ll see network hardware described by the OSI layer it operates at.
The protocol data unit (PDU): As data is processed at different layers of the OSI model, headers and footers are stripped away, changing the type of PDU being transmitted. For example, when data transmitted at the network layer (Layer 3) is processed by the transport layer (Layer 4), the IP header is no longer needed. The PDU with the IP header at Layer 3 was a packet. The PDU at Layer 4 without the IP header is a segment. Understanding the different PDUs will help you make sense of the terminology associated with different networking devices.
Here’s a visual representation of the OSI model and the PDUs (data) associated with each layer.
11 types of network device types
A firewall is a network security device that monitors and either blocks or allows traffic based on a set of rules. Firewalls can be software, hardware, or a combination of both. Additionally, the rules firewalls use can be based on something straightforward like ports and IP addresses or use heuristics to identify malicious behavior.
Common examples of network firewalls include:
Packet-filtering firewalls: These firewalls operate at Layer 3 and Layer 4 and determine whether to drop or forward traffic using rules based on basic information like IP addresses, port numbers, and packet type. One of the upsides of this simple approach to inspecting traffic is that it has little impact on network speed.
Stateful-inspection firewalls: Unlike packet-filtering firewalls, stateful-inspection firewalls can track and “understand” when a TCP connection has occurred. This allows for reply traffic to be allowed through the firewall without the need for explicit rules. As a result, stateful-inspection firewalls make configuration easier, but can add a bit more inspection overhead and slow down traffic slightly more than packet-filtering firewalls.
Application firewalls: These firewalls operate at Layer 7 and can contextualize what’s normal vs malicious behavior for application-layer protocols like HTTPS, SSH, and SMTP. Proxy firewalls that provide URL filtering and WAFs (web application firewalls) that protect against cross-site scripting (XSS) and SQL injection attacks are two common types of application firewalls.
The textbook definition of a network switch is a Layer 2 device that sends and receives frames. These switches are the basic building block of Ethernet networks.
Here’s a basic example of how a Layer 2 switch works:
Devices are connected to the switch using Ethernet cables (e.g., a Cat5e or Cat6 cable) creating a small LAN.The switch learns the MAC addresses of the connected devices.When traffic needs to go to a specific device, the switch sees the MAC address in the packet and sends it to only that device.
By sending the data to a specific device, the switch is breaking up collision domains and greatly reducing network congestion when compared to network hubs. That breaking up of collision domains is the basic benefit of a Layer 2 switch.
However, this basic example of a Layer 2 switch is just one of the many types of network switches. Here’s a list of common types of network switches:
Unmanaged switches: Unmanaged switches simply provide Layer 2 switching of Ethernet frames. They don’t offer any additional management or configuration features.PoE switches: Switches that provide Power over Ethernet (PoE) functionality can provide both network connectivity and power to connected devices. For example, it’s common to power Voice over IP (VoIP) phones using PoE switches. PoE switches can be Layer 2 or Layer 3 switches and can be managed or unmanaged.
Layer 3 switches: These switches offer the same Layer 2 functionality as other switches, but add Layer 3 routing to the mix. Layer 3 switches are aware of IP addresses and can route packets between networks.
Stackable switches: Some network switches can be “stacked”. These stackable switches can be connected to one another to operate as a single logical switch. Stacking switches can be a useful way to increase the capacity of a network. For example, stacking two 24-port switches would create a single 48-port switch from a management and functionality perspective.
Operating at Layer 2, access points (APs)—also known as wireless access points (WAPs)—are the network switches of the wireless world. WAPs connect to a LAN through a wired connection and allow other Wi-Fi devices to communicate. The networks created by WAPs are WLANs (wireless local area networks).
At a high level, there are three main types of APs:
Fat APs: Fat access points are also known as autonomous APs. Fat APs directly handle the forwarding of traffic from their Wi-Fi radios to the wired network. These WAPs have interfaces where you can access and configure things like access control lists, Wi-Fi encryption, DHCP server functionality, and QoS directly. The upside of fat APs is that they’re an all-in-one solution in that they have everything you need to configure and manage the device baked in.Thin APs: These APs don’t offer the full functionality of an autonomous AP. Instead, they backhaul Wi-Fi traffic to a centralized WLAN controller that handles all the forwarding and functionality a fat AP would do on its own. While the limited functionality of thin APs has some tradeoffs, this centralized approach to WAP management can make large-scale deployments easier to manage.Fit APs: These WAPs offer a mix of fat and thin AP functionality. For example, most fat APs will handle Wi-Fi encryption at the AP level instead of waiting until the traffic gets back to the controller and may offer services like DHCP relay, but a centralized controller will be used for things like bridging forwarding traffic between wired and wireless networks.
Routers are the network devices that route packets between networks. These Layer 3 devices enable everything from communication between multiple subnets within the same WAN to the internet connection that allows you to read this article. A good way to think of routers is this: They are the network device that deals with IP addresses.
Of course, that definition might leave you wondering why we need routers if we have Layer 3 switches. For some insight on that topic, see Layer 2 vs Layer 3 Network Switches: What’s the Difference?
NAS (Network attached storage)
A NAS is a server dedicated to file storage. Within a LAN, a NAS provides a central storage point that can be used for things like shared access to files and storing backups of user data. NAS devices generally provide an affordable and simple way to provide network storage. In recent years, the lines between a NAS device and a general-purpose server are getting even more blurred as NASes begin to offer more advanced functionality suited to small and mid-sized environments.
Load balancers distribute connections from clients across multiple servers. As with firewalls, there are plenty of software and hardware implementations of load balancers. Load balancers usually operate at Layer 4 (filtering based on TCP or UDP traffic) or Layer 7 (filtering based on HTTP or DNS traffic). Common approaches to load balancing include:
Round-robin: Simply forwards requests to each server in a cycle.
Weighted round-robin: Like round-robin, but takes into account “weights” that help determine what amount of connections a given server should be sent.
Least connections: A form of load balancing that prioritizes sending connections to the server with the fewest connections.
Weighted least connections: Like weighted round-robin, weighted least connections allows the servers to be assigned “weights” that influence the load balancing algorithm.
A repeater is a simple Layer 1 device that rebroadcasts a signal. Repeaters are sometimes referred to as signal boosters. There are repeaters for Wi-Fi, Ethernet, and other network connections, but fundamentally they do the same thing: take a signal in and rebroadcast it.
Pro tip: Simple Wi-Fi repeaters can lead to significant network congestion. If you’re looking to optimize performance, a WAP is often a better choice.
From a hardware perspective, there’s no difference between a gateway and a router. Gateways are simply routers that serve a specific purpose. Gateways are routers that act as the default next hop. When there’s no other route to an IP address on a network, the packets get routed to the network’s default gateway. From there, the default gateway routes the packets on to their next “hop” and the process repeats until the destination is reached.
By the traditional definition, a modem is a device that modulates and demodulates a signal to and from analog and digital. In the days of dial-up internet, modems were used to connect local computer networks to analog telephone lines for internet access. Today, when people say “modem” they may be referring to any number of devices that allow you to connect to a carrier’s network.
Examples of popular types of modems include:
Analog modemsCable modemsCellular modemsFiber optic modemsxDSL modems
A hub is a simple type of Ethernet repeater that operates at Layer 1, enabling the connection of multiple devices to the same Ethernet network. Unlike a switch, a hub does NOT break up collision domains, meaning all ports on a network hub get sent the same traffic. As a result, hubs—similar to WiFi repeaters—can cause quite a bit of congestion.
Pro tip: By rebroadcasting traffic to all ports, hubs can create significant network congestion. If you have a use case for a hub, consider an unmanaged switch, which will break up collision domains, instead.
Traditionally, network bridges were Layer two devices that often had only two ports. Like switches, they broke up collision domains and could reduce network congestion when compared to hubs by separating the network into multiple collision domains.
However, today the term “bridge” can mean a lot more than that traditional definition depending on the context. Other interpretations of the term network bridge:
A device that can connect VLANs at Layer 2A network switchA device that can connect two networks using different connection mediums, such as connecting a WLAN to a wired LAN
What about “brouters”?You may have heard the term brouter before. As the name implies, these hybrid devices deliver the Layer 2 bridging functionality of a bridge and the Layer 3 routing functionality of a router.
Considering needs for network device management
Modern networks include many different types of network devices. Monitoring and managing all of them efficiently can be a difficult task. This is true both when you’re designing a network from the ground up or when you inherit a network someone else designed.
Below are some of the key needs to consider when planning your network device management strategy.
The infrastructure, devices, services, and topology of your network fall under the umbrella of network architecture. Good network architecture will help you get better overall performance. For example, effective network segmentation can significantly reduce network congestion. Further, by designing for redundancy you can improve network resilience and reduce downtime.
Even if you nail your network design and optimization on day one, managing that network long-term can be a real challenge. This is particularly true in large networks where multiple parties were involved in the network design and implementation. Up-to-date and easily accessible network documentation—like network maps, for example—can go a long way in enabling effective network troubleshooting and device management.
Of course, there’s a lot that goes into getting network architecture and documentation right and we’ve only scratched the surface here.
No network device management strategy can overlook network security. If an attacker gains control of your network devices, they’re not really yours anymore.
In addition to building security into your network architecture with firewalls, IPS (Intrusion Prevention Systems), and other security appliances, you should take steps to harden your network devices. For example:
Disable unused services: Many network devices support a variety of network protocols. By disabling unused protocols, you limit attack surface.
Rotate credentials regularly: Passwords can get compromised. In the event they do, a policy that forces users accessing your network devices to change their passwords regularly (every 90 days or less is a good rule of thumb) can limit the amount of time a hacker can use compromised credentials.
Use certificate-based authentication or SSH keys when possible: Many network devices allow you to use certificate-based authentication or authenticate using SSH keys. When possible, using these methods can be significantly more secure than username/password authentication alone.
Network monitoring tools
The right network monitoring tools can make a big difference when it comes to network device management. By using protocols like SNMP, syslog, IPMI, and the various flow protocols, network monitoring tools allow you to gain full visibility into your network, create detailed reports on network performance, automate responses to network outages, and monitor all your network devices in real-time.
—The most efficient way to manage network devices is using positiveeast.org’s cloud-based network management system. Start your free 14-day trial today.