White‑Screen Operations Platform for Multi‑Cloud Kubernetes Middleware Management

Project background: Traditional operations face fragmented management, high cost, and reliance on command‑line tools (kubectl) that increase risk.

Kubernetes and Operators provide a unified CR‑based approach, offering standardization, automation, and a visual UI.

Core objectives: standardize middleware operations, automate tasks, and provide a visual interface.

Construction phases: architecture overview, multi‑cloud cluster management, middleware services (Kafka, Elasticsearch), node management, PV management, CPU burst, and YAML management.

Key services:

Multi‑cloud management: unified hosting of clusters from different cloud providers with visual resource scheduling.

Middleware service: unified deployment and management of Kafka and Elasticsearch via UI.

K8s generic resource service: Node, PV, PVC, Service, Pod management, reducing black‑screen commands.

YAML management: version control, audit, visual diff and rollback.

Operation audit: detailed logs integrated with DCheck for compliance.

Pain points addressed:

Multiple kubeconfig files cause context confusion and high risk of accidental operations.

Manual scripts for Kafka scaling, node selection, and PV cleanup are error‑prone and time‑consuming.

Solution examples:

Kafka scaling script (original bash) is replaced by a one‑click UI that automates replica increase, pod readiness check, and data migration via Cruise‑Control.

#!/bin/bash
export KUBECONFIG=/path/to/kubeconfig
echo "Step 1: 查询 Kafka 集群状态"
kubectl get kafka -n kafka-namespace
# ... rest of script ...

Node selection Java example illustrates the complexity of manual scripting.

public class ESNodeSelectorTest {
    public static void main(String[] args) {
        int needCpu = 9;
        int needMemory = 33;
        // ... rest of code ...
    }
}

By moving these operations to the white‑screen platform, tasks that previously took hours are reduced to minutes (e.g., node labeling from >1 h to 3 min, PV release from >15 min to 1 min).

Results to date: over 30 Kubernetes clusters managed, 300+ middleware clusters, 430+ white‑screen operations, saving >15 k RMB per month by reclaiming 675+ idle cloud disks.

Benefits include standardized processes, increased efficiency, reduced human error, auditability, compliance, and scalability for future Kubernetes resources.

Future outlook: expand to more Kubernetes resources (Deployments, StatefulSets, Ingress, ConfigMaps, Secrets), add intelligent operations (self‑healing, dynamic scheduling), improve UI/UX, and enhance multi‑cloud support.