Loading

On-Premises / RTF Cluster Stuck in Yellow Status — Open Port 5701 for Inter-Node Hazelcast Communication

Publiseringsdato: Jun 24, 2026
Beskrivelse

PROBLEM

A Runtime Fabric (RTF) or on-premises cluster reports a Yellow health status. One or more nodes appear degraded and the cluster does not recover on its own.

 

ROOT CAUSE

Hazelcast, the in-cluster data grid used for inter-node communication, requires TCP port 5701 to be reachable between all nodes. When this port is blocked, at either the cloud security group layer or the host-level firewall and nodes cannot form a complete cluster ring, resulting in a Yellow (degraded) status.

Løsning

SOLUTION

Open TCP port 5701 for inbound traffic at both levels:

 

1. AWS Security Group

  Add an ingress rule allowing TCP 5701 between all nodes in the cluster's security group. If the AWS IAM user associated with the environment lacks the necessary permissions to make this change directly, the rule must be applied via an infrastructure pipeline or by an administrator with sufficient privileges.

 

2. Host Firewall (firewalld)

  On each node, allow the port at the OS level:

  firewall-cmd --permanent --add-port=5701/tcp

  firewall-cmd --reload

 

After both changes are applied, confirm that the cluster status transitions from Yellow to Green. No application restart should be required — Hazelcast will re-establish inter-node connections automatically once the port is reachable.

Knowledge-artikkelnummer

005387638

 
Laster
Salesforce Help | Article