在OpenShift上部署Redis后无法连接的问题求助

Error 10060 is a connection timeout, meaning your redis-cli request isn’t reaching the Redis instance at all. Let’s walk through the most likely issues and fixes step by step:

1. OpenShift Routes are HTTP/HTTPS-only by default (Redis uses TCP) #

Redis runs on a raw TCP protocol, but standard OpenShift Routes are built to handle HTTP/HTTPS traffic. Trying to use a regular Route for Redis will fail because the ingress controller doesn’t know how to route TCP traffic through it.

Fixes: #

• Option 1: Use a TCP Route (OCP 4.4+): If you’re on OpenShift Container Platform, configure your ingress controller to support TCP routes:
  1. Edit the ingress controller config to expose Redis’ default port (6379) as a TCP endpoint.
  2. Create a Route with spec.tls.termination: passthrough and spec.port.targetPort pointing to your Redis service’s port.
• Option 2: Switch to NodePort/LoadBalancer Service: For simpler external access, change your Redis Service type to NodePort (connect via <node-ip>:<node-port>) or LoadBalancer (if your cluster supports it, you’ll get an external IP to use).

2. Redis is bound only to localhost #

By default, Redis is configured to listen on 127.0.0.1, which means it won’t accept connections from outside its container.

Fix: #

• Update your Redis deployment to bind to 0.0.0.0:
  • Add a command argument in your Deployment/DeploymentConfig: --bind 0.0.0.0
  • Or modify redis.conf to set bind 0.0.0.0 (if using a custom config)
• Verify the bind address by exec-ing into the pod:

  oc rsh <redis-pod-name>
  redis-cli CONFIG GET bind
  

  It should return ["0.0.0.0"].

3. Network policies or firewalls are blocking traffic #

A connection timeout usually means traffic is being blocked between your local machine and the Redis pod.

Checks: #

• Cluster External Firewall: Ensure the port used by your Route/Service is open in your cloud provider’s firewall or on-premise network rules.
• OpenShift NetworkPolicy: Run oc get networkpolicies to list existing policies, and confirm none are restricting incoming traffic to the Redis Service.
• Pod Security Context: Ensure the pod has permissions to accept incoming connections (less likely for timeout errors, but worth checking).

4. Route-Service-Port misconfiguration #

Double-check that your Route is correctly linked to the right Service and port:

• Run oc get routes <your-route-name> -o yaml to confirm:
  • spec.to.name matches your Redis Service name
  • spec.port.targetPort matches the port exposed by your Service (should be 6379 for Redis)
• Run oc get svc <redis-service-name> to verify the Service targets the correct pod labels and exposes port 6379.

5. TLS mismatch (if using encrypted Routes) #

If your Route uses TLS termination (like passthrough), your redis-cli needs to use TLS to connect. Without it, the connection will time out because the ingress controller expects TLS traffic.

Fix: #

• Connect with the --tls flag:

  redis-cli -h xx.com --tls
  

• If using a self-signed cert, add --insecure to skip verification:

  redis-cli -h xx.com --tls --insecure
  

Start with checking the Route type and Redis bind address—those are the two most frequent causes for this exact error.

内容的提问来源于stack exchange，提问作者Spaceship222