AWS Elastic Load Balancing

Overview

AWS Elastic Load Balancing (ELB) is a managed service that distributes application traffic across targets and virtual appliances in one or more Availability Zones. It suits teams running workloads on AWS that need routing and capacity to adjust with demand. Its range of load balancer types is a strength, but the right fit depends on traffic type and the full usage-based bill.

Key features

ELB offers Application, Network, Gateway, and Classic Load Balancers, matching different workloads instead of forcing them into a single routing model. Capacity scales with application traffic, and configurable health checks help route traffic to appropriate targets. AWS gives Classic, Application, and Network Load Balancers a monthly availability guarantee of at least 99.99%.

Application Load Balancer is the choice for HTTP requests: it supports header-based routing, redirects, fixed responses, HTTP/2, and gRPC. Network Load Balancer targets transport protocols and workloads where extreme performance or low latency matters; static Elastic IP addresses and AWS PrivateLink are supported for TCP and TLS listeners. Gateway Load Balancer is designed for third-party virtual appliances, while Classic is intended for applications built in EC2-Classic.

SSL/TLS termination, certificate management, and client certificate authentication address common connection-security needs. CloudWatch metrics and logging cover all four types. Kubernetes Controller support includes direct-to-pod routing for Application Load Balancers and fully private EKS clusters for Application and Network Load Balancers. A single load balancer can also span AWS and on-premises resources, which is useful for hybrid environments.

One migration caveat: Application Load Balancers require a different API set from Classic Load Balancers, so existing Classic integrations may need changes.

Pricing

ELB is paid, with a free tier for new AWS customers. The AWS Free Tier costs nothing for six months after account creation, with up to $200 in credits. It includes 750 hours per month shared between Classic and Application Load Balancers, plus either 15 GB of Classic data processing or 15 LCUs for Application Load Balancers. The shared hours and type-specific allowances make this useful for a limited evaluation, not a dependable estimate of ongoing production costs.

Paid plans use regional, usage-based rates, and standard AWS data-transfer and public IPv4 address charges may apply separately:

• Application Load Balancer: custom pricing; billed per hour or partial hour, plus Load Balancer Capacity Units used per minute. Fits HTTP routing workloads.
• Network Load Balancer: custom pricing; billed per hour or partial hour, plus Network Load Balancer Capacity Units used per minute. Fits transport protocols and low-latency or extreme-performance needs.
• Gateway Load Balancer: custom pricing; billed per hour or partial hour, plus Gateway Load Balancer Capacity Units used per minute. Gateway Load Balancer Endpoints are billed separately.
• Classic Load Balancer: custom pricing; billed per hour or partial hour and per GB transferred. It is for applications built in EC2-Classic.

Because rates vary by region and charges can extend beyond the load balancer itself, compare the expected workload and related transfer and address costs before choosing a configuration.

Platforms

ELB is available through API and web platforms. It is a managed service, with Layer 4 and Layer 7 routing and global traffic management.

Who it's for

ELB is a strong fit for AWS teams that want managed traffic distribution, automatic capacity scaling, health checks, and load-balancer options for HTTP, transport, or virtual-appliance workloads. Its Kubernetes and hybrid capabilities also suit teams routing to EKS workloads or spanning AWS and on-premises resources. It is less suitable when a fixed, predictable price is essential or when a Classic-based integration cannot accommodate a different Application Load Balancer API.

Pros and cons

• Pros: Four load balancer types let teams match routing to workload; capacity scaling and health checks reduce the need to manage those functions directly; CloudWatch metrics and logging cover all four types.
• Pros: The stated 99.99% monthly availability guarantee for Classic, Application, and Network Load Balancers gives teams a clear availability commitment for those options.
• Cons: Usage-based regional billing, capacity-unit charges, and possible data-transfer and public IPv4 charges make total costs harder to predict.
• Cons: Application and Classic Load Balancers use different API sets, adding potential work for Classic integrations.

Alternatives

HAProxy Community Edition is a free Linux, self-hosted option for TCP and HTTP proxying when managing the load balancer yourself is acceptable. Google Cloud Load Balancing is a paid API and web alternative with usage-based pricing tied to forwarding rules, traffic processed, and other services. Alibaba Cloud Server Load Balancer offers a web-based alternative with an ALB Basic reference price of $0.01 a month, billed hourly, and a 200-backend-server-per-instance limit.

Progress Kemp LoadMaster has a 30-day temporary LoadMaster trial license for the VLM-MAX model with Enterprise Plus Support, including ESP, WAF, and GEO. Cloudflare Load Balancing is a web-based paid option with a Basic plan at $5 a month for the first 500K DNS queries. HAProxy Enterprise is a paid Linux and self-hosted alternative with customized pricing. Citrix ADC (now NetScaler) is a paid self-hosted alternative with a platform license that includes unlimited Gbps throughput and unlimited NetScaler software instances. Radware Alteon is another paid alternative.

Compare more options in Load Balancing Software.

Verdict

Choose AWS Elastic Load Balancing if you run on AWS and want managed, traffic-responsive load balancing with distinct options for HTTP, transport, and virtual-appliance workloads. Its breadth and scaling are the main reasons to choose it; the usage-based bill and extra possible charges are the main reasons to look elsewhere when cost certainty matters.