Cello CNI vs Liqo
Cello CNI
github.comLiqo
liqo.ioSide by side
| Figure | Cello CNI | Liqo |
|---|---|---|
| Price from | Free | — |
| Free plan | Yes | — |
| Free trial | — | — |
| Open source | — | — |
| Windows | — | Yes |
| macOS | — | Yes |
| Linux | Yes | Yes |
| Self-hosted | Yes | Yes |
| Plans published | 1 | 0 |
| Spec score · Container Networking Software | 7.9 (#6) | 7.1 (#18) |
| License | Cello is distributed under the Apache 2.0 license. | The official site says Liqo is licensed under Apache-2.0. |
| Purpose | Cello is a Kubernetes CNI plugin based on the Volcengine VPC network that lets cluster workloads communicate within the VPC. | Liqo is an open-source project for creating dynamic Kubernetes multi-cluster topologies across on-premise, cloud, and edge infrastructures. |
| Support | The project directs questions and contributions to GitHub issues or pull requests and also mentions maintainer email and a Feishu group. | The official site links to Liqo’s documentation, quick-start guide, and community Slack. |
Plans
In their own words
Liqo
Liqo connects Kubernetes clusters into multi-cluster topologies, allowing them to share namespaces, resources, and workload capacity. It can extend consumer namespaces and replicate configurations, secrets, services, and other resources. A remote cluster appears as a virtual node, so the consumer cluster can schedule workloads there. Liqo’s network fabric connects pods across peered clusters using WireGuard VPN tunnels by default, and it describes cross-cluster storage support for stateful workloads. Peering can be declared with Kubernetes resources and managed through GitOps tools such as ArgoCD. The liqoctl CLI installs and uninstalls Liqo, manages peering, configures namespace-level offloading, and reports status. Liqo targets heterogeneous on-premises, public cloud, edge, and multi-cloud environments, with support listed for Kubernetes, Kubeadm, K3s, K0s, AKS, EKS, GKE, Aruba Cloud KaaS, and OpenShift. It is free and open source under Apache-2.0. Liqo requires Linux kernel 5.10 or later and recommends at least 2 CPUs and 2 GB of RAM for a cluster.