Decentralized Load Balancing
Use Cases

Decentralized architectures
require decentralized LBs

Introducing ZeroLB, a better way to build load balancing via service mesh that is more perfomant and reliable than legacy load balancers.

    Decentralized Load Balancing

    Remove the biggest bottleneck

    Load Balancing Diagram

    Centralized load balancers have no place in decentralized architectures, with ZeroLB we can provide intelligent and portable load balancing on K8s and VMs.

    Increase performance by 2-4x

    Load Balancing Diagram

    By removing an extra hop in the network we can decrease network latency by 2x on request lifecycles and 4x on full roundtrips. Being slow is not an option.

    Reduce annual TCO by 6-7 figures

    Load Balancing Diagram

    Without a centralized load balancer in front of every service in our organization, we can reduce costs up to 7 figures per year while decreasing complexity.

    MS3 Logo
    Yahoo Japan Logo
    Verifone Logo
    Papa John's Case Study

    Deploy ZeroLB architectures

    Kong Mesh can automatically inject a modern service mesh across every service, API, and application to provide smart decentralized load balancing.

    • Round Robin, Least Connect, Ring Hash, Maglev, Random
    • Traffic splitting, weights, canary, blue/green, feature flagging.
    • Support for multi-clusters and multi-cloud.
    • Injects on Kubernetes, VMs and bare metal.
    • Zone aware load balancing to reduce egress costs.
    • Portable across every cloud and datacenter.

    Self-healing and intelligent

    With ZeroLB, load balancing not only gets faster, but it also provides automatic self-healing and zone failover to maximize uptime.

    • Dynamic health checks and circuit breakers
    • Zone failover across both containers and VMs
    • Cross zone load balancing to lift and shift workloads
    • Dyamically configured via Envoy proxy

    Native load balancing telemetry

    Load balancing our workloads is not enough, being able to determine where and when error occurs is critical to determine the health of our applications.

    • 70+ observability charts out of the box
    • Service topology map
    • Native tracing via Zipkin, Jaeger or Datadog
    • Logging and metrics collection to any 3rd party

    Questions about breaking your monolith down into smaller parts?

    Contact us today and tell us more about your configuration and we can offer details about features, support, plans and consulting.