A load balancer does more than take turns between servers. It applies routing rules, chooses an eligible backend, and responds to health and availability signals—so it can determine where a request or new network flow goes. What it can inspect and control depends on its layer: an application-layer load balancer can route HTTP requests by content, while a Layer 4 load balancer generally distributes TCP or UDP flows without inspecting their payloads.
What does a load balancer actually decide?
Think of a load balancer as a traffic decision point between clients and a set of backend resources. It accepts traffic at a frontend address or listener, applies the configured rules, identifies the eligible backend pool, and selects a destination. Health checks and topology settings can affect that decision too.
As an Amazon Associate I earn from qualifying purchases.
Round robin describes only one possible way to select a destination after the relevant routing policy has identified where traffic belongs. It does not explain how the balancer chooses a service, whether a backend is considered healthy, or how traffic is distributed across availability zones.
Do these 3 things before closing this tab:
1Fix the driver behind crashes, sound loss and screen glitches2Clear out junk files and repair common Windows errors3Scan for outdated or missing drivers - takes under a minuteHow does a load balancer choose a destination?
First, rules identify the destination group
With an AWS Application Load Balancer (ALB), listeners accept connections on configured protocols and ports. Listener rules are evaluated in priority order; conditions can use URL paths, host headers, HTTP headers, methods, query parameters, or source IP addresses. A matching rule can send a request to a target group, allowing one listener to route different requests to different services. The target group contains the registered targets that can receive that traffic. AWS’s ALB overview describes this listener-and-rule model.
#1 Best Overall
- Professional 10Gbps Wired Routing – Route10 is a high-performance 10 Gigabit wired router designed for advanced home, business, and enterprise networks; it does not broadcast Wi-Fi, and wireless coverage requires pairing with one or multiple Wi-Fi access points such as ceiling, wall, or outdoor access points for full network coverage.
- Quad-Core Qualcomm Network Accelerator for High Throughput – Powered by a high-performance quad-core Qualcomm processor with hardware-accelerated networking, the Route10 delivers fast packet processing, low latency, and consistent multi-gigabit performance for routing, firewall rules, VPN traffic, VLAN segmentation, and high-bandwidth network workloads without bottlenecks.
- Integrated PoE+ Output to Power Network Devices – Select Ethernet ports provide Power over Ethernet Plus (PoE+) support, allowing the router to power compatible access points, network devices, or edge hardware directly through the Ethernet cable, reducing the need for additional power adapters or injectors.
- Enterprise-Grade Routing, Firewall, and Network Control – Supports advanced routing features including VLAN tagging, QoS traffic prioritization, NAT port forwarding, firewall rules, DHCP services, and professional network segmentation for secure, reliable, and scalable wired network deployments.
- Real-Time Network Monitoring and Traffic Visibility – Provides live network statistics and real-time monitoring of bandwidth usage, connected devices, WAN and LAN traffic, and system performance, allowing network administrators to quickly identify issues, optimize traffic flow, and maintain stable, high-performance wired networks.
Then, the product’s selection method chooses among targets
For AWS ALB, the documented default target-selection algorithm is round robin; least outstanding requests is also available. This is a request-level choice within the selected target group, not a universal rule for all load balancers. AWS documents the ALB algorithms.
Azure Load Balancer works differently. It operates at Layer 4 and, by default, hashes a five-tuple: source IP address, source port, destination IP address, destination port, and protocol. The resulting distribution is over TCP or UDP flows, not individual HTTP requests. Azure also offers two-tuple and three-tuple distribution modes that provide session affinity. Microsoft Learn explains Azure’s algorithm and modes.
Why does the network layer matter?
An application-layer balancer such as AWS ALB can use HTTP request content to make routing decisions. Azure Load Balancer is a Layer 4 service: it distributes TCP and UDP flows but does not inspect application payloads, rewrite HTTP headers, provide application-gateway behavior, or offload TLS. It also does not terminate or originate flows. Those differences determine which routing policies are possible and whether the balancer handles individual requests or network flows. AWS ALB documentation and Microsoft’s Azure documentation describe these respective scopes.
Rank #2
- Compatible management via CloudKey, Official UniFi Hosting, or UniFi Network Server running version 8.3.32 or newer
- Ensures continuous connection through Shadow Mode High Availability featuring automatic failover (VRRP)
- Delivers 12.5 Gbps routing performance equipped with IDS/IPS capabilities
- Offers license-free, real-time decryption and inspection of encrypted traffic using NeXT AI Inspection*
- Features 25G SFP28, 10G SFP+, and 2.5 GbE RJ45 ports where two interfaces can be reconfigured as WAN connections
How do health checks change where traffic goes?
Health checks determine which backends a product treats as eligible for traffic, but their meaning depends on the configured probe and the product’s failure behavior. A check that only confirms a network port responds may not prove that the application is ready to serve useful work.
AWS ALB: target-group checks and a fail-open exception
AWS configures health checks per target group. ALB checks use HTTP or HTTPS GET requests, with settings including the path, interval, timeout, consecutive success and failure thresholds, and accepted response codes. Normally, ELB routes to healthy targets. However, AWS ALB has an important exception: if every registered target in a target group is unhealthy, it fails open and routes traffic to all of them regardless of health status. AWS documents ALB health-check configuration and this behavior; it should not be assumed to apply to every load balancer.
Azure: probe protocol and SKU affect failure outcomes
Azure Load Balancer uses probes to determine backend health and eligibility for new flows. Standard Load Balancer supports TCP, HTTP, and HTTPS probes. For HTTP(S) probes, a response other than HTTP 200 counts as failure. Probe behavior also varies by protocol and SKU: in some all-probes-down situations, Standard SKU can allow existing TCP flows to continue, while UDP flows behave differently. That means “a backend failed its probe” is not enough information to predict what happens to established connections. Microsoft Learn details probe behavior and SKU caveats.
Rank #3
- Hardwired Router
- Titan Networx
- High performance router
- managed switch
- integrated router
Microsoft’s current documentation lists a 5-second default probe interval in the Azure portal and a 30-second built-in timeout for HTTP/S probes; defaults can differ by deployment interface. These are service configuration values, not universal timing guarantees. See Microsoft’s probe documentation for the configuration context.
Windows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallCrashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minuteMake the probe reflect the service you need
Microsoft recommends probes that reflect the health of both the instance and its application service. If a probe checks only a listening port, it may report a backend as healthy even when the application cannot serve requests. Conversely, probing through a backend appliance to another instance can make a downstream problem appear to be the appliance’s failure and trigger cascading effects. A probe can also be deliberately used to remove an instance from new-flow eligibility during maintenance, so its operational meaning should be intentional. Microsoft’s probe guidance covers these design considerations.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.How do zones and affinity shape traffic?
Backend selection is also affected by topology. AWS requires an ALB to have at least two enabled Availability Zones and recommends using multiple zones. Cross-zone load balancing determines how each load-balancer node distributes its share of client traffic among targets across zones. AWS zonal shift is a separate recovery control for moving traffic away from an impaired zone; existing connections may take time to complete. AWS also describes a 60-second DNS entry TTL for Elastic Load Balancing—this is an AWS service-documentation value, not a general DNS rule. AWS explains zone behavior, zonal shift, and DNS in its ELB documentation.
On Azure, the default five-tuple hash helps distribute flows, while two-tuple or three-tuple modes can keep flows associated more closely with a client. Choosing an affinity mode changes distribution behavior; it is not simply a different name for round robin. Microsoft Learn describes the available hash modes.
What should you compare when choosing a load balancer?
| Decision area | Questions to ask |
|---|---|
| Layer and traffic unit | Does the service route application requests using content, or distribute TCP/UDP flows without inspecting payloads? |
| Routing inputs and selection | Can it match the host, path, headers, or other request attributes you need? Which target-selection algorithms or flow-hash modes are available? |
| Health model | Which protocol, path, response criteria, interval, and thresholds define health? Does the probe reflect application readiness? |
| Failure behavior | What happens when one backend or every backend fails a probe? What happens to established connections, and can the service fail open? |
| Topology and affinity | How does cross-zone distribution work? Is client affinity needed, and what trade-offs does it introduce? |
| Operational controls | How are targets registered or removed for maintenance? Which health metrics and logs help explain routing decisions? |
The answers are product- and configuration-specific. The AWS ALB request-routing model and Azure Load Balancer flow-distribution model are examples of different control surfaces, not interchangeable implementations of one universal algorithm.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
