Networking · in-mumbai-1 · in-noida-1

Load balancing at a flat rate you can memorise

AWS bills its ALB by the hour plus LCUs — a unit you can't predict until the invoice arrives. Ours is a flat monthly price: HAProxy-backed, SSL offload, health checks, sticky sessions, static VIP. From ₹999/mo.

Create a load balancer — ₹25,000 credit View capabilities ↓

Pricing flat, no LCUsSSl offload + SNIAlgorithms rr · leastconn · sourceChecks TCP / HTTPVIP static IPv4

Capabilities

The features you actually configure at 6 p.m. on launch day

Three balancing algorithms

Round-robin for uniform backends, least-connections for mixed workloads, source-IP for session-bound legacy apps. Switch per rule, live.

SSL offload with SNI

Terminate TLS at the balancer with multiple certificates per VIP. Backends stay plain HTTP on the private network; your CPUs stay on application work.

Health checks that eject fast

TCP or HTTP probes with tunable intervals and thresholds. Unhealthy backends leave rotation in seconds and rejoin automatically when they recover.

Sticky sessions, three ways

LB-issued cookie, application cookie, or source-IP persistence — pick per rule for apps that keep session state on the node.

Autoscaling hook

Attach an autoscale group: set min/max instances and a CPU threshold, and the platform adds or removes backends and updates the pool for you.

Static VIP + live metrics

One IPv4 that never changes under DDoS-filtered transit, with per-rule connection and traffic graphs in the console.

Under the hood

HAProxy on redundant virtual routers

Rules configure HAProxy on CloudStack virtual routers — Dedicated HA runs an active-passive pair with sub-second VIP failover. Stickiness maps to LbCookie, AppCookie and SourceBased policies; autoscaling uses CloudStack autoscale VM groups.

TCP and HTTP(S) modes, WebSocket-friendly timeouts
Certificate upload with full-chain validation, per-SNI routing
Backend pools managed by API, console or Terraform
Failover pair in Dedicated HA — maintenance without a maintenance window
VIP and backends both live in India

Frequently asked questions

TCP or HTTP mode — which do I want?+

HTTP(S) mode for web apps: you get SNI routing, cookie stickiness and header-aware checks. TCP mode for databases, MQTT, game servers or anything that isn't HTTP. Both are available on every plan; SSL offload needs Pro or above.

Can I bring my own certificate?+

Yes — upload certificate, chain and private key, and attach per rule. Pro and Dedicated support multiple certificates on one VIP via SNI. Automated renewal via API fits neatly into a certbot deploy hook.

How fast do failed backends leave rotation?+

Health checks run at your configured interval (default 5 s) with a failure threshold (default 2) — so a dead node stops receiving traffic in roughly 10 seconds and rejoins automatically once probes pass.

Does it autoscale my servers?+

Yes, via autoscale groups: define a template, min/max count and a CPU threshold. The platform provisions or destroys instances and keeps the LB pool in sync — no external scripts.

What happens when I outgrow a plan?+

Upgrade in place — the VIP stays, connections drain gracefully. If you're pushing past Dedicated HA's 200k concurrent, talk to us about a custom pair.

Pairs well with

Complete the stack

Cloud ServersThe D-series backends this balancer fronts→VPCKeep LB-to-backend traffic on a private tier→Managed Kubernetestype=LoadBalancer Services wired automatically→