Frequently Asked Questions

Yes. The core routing engine, observability dashboard, and gRPC/HTTP/2 multiplexer are released under Apache 2.0. We only gate advanced rate-limiting algorithms and multi-cluster failover behind the FluxGate Enterprise license.

Run `fluxgatectl migrate --source kong --config kong.yaml`. The CLI parses your declarative routes, transforms JWT/OIDC plugins into FluxGate's native policy engine, and outputs a ready-to-apply `fluxgate-stack.yaml`. Most teams complete zero-downtime cutover in under 4 hours.

Check your `upstream_keepalive` and `worker_connections` in `fluxgate.conf`. The default pool is 256 connections. For clusters handling >50k RPS, increase to 1024 and enable `tcp_nodelay`. Also verify that your service mesh isn't double-terminating TLS.

Absolutely. Use the `fluxgate-inject` mutating webhook with the `--proxy-mode sidecar` flag. It automatically discovers your pod's `CLUSTER_IP` and routes outbound traffic through local port 15001. Memory overhead averages 42MB per sidecar instance.

The OSS version is free indefinitely. You only pay when you enable the `enterprise` feature flag for distributed rate limiting, audit logging, or SSO integration. Pricing scales at $0.04 per million requests after the first 50M, with a $2,400 annual floor for support SLAs.

Yes. Universities and registered 501(c)(3) organizations receive a waived enterprise license and dedicated sandbox clusters. Submit a request through our portal with a valid `.edu` domain or IRS determination letter.

Use our GitHub issue tracker with the `bug` or `enhancement` labels. For production incidents affecting >100 nodes, open a case in the customer portal. Our SRE team responds within 15 minutes for Tier-1 contracts.

Join `#fluxgate-general` and `#architectures` on our community Slack. Over 12,000 developers share routing configs, discuss eBPF integration, and troubleshoot TLS termination edge cases daily.