Troubleshooting
Problem
When starting a cluster, if there is a port conflict between two different Java Virtual Machine (JVM) processes, the following informational DCSV1031I message will be logged.
Symptom
9/4/07 15:12:22:261 EDT] 00000016 DiscoveryServ I DCSV1031I: DCS Stack
DefaultCoreGroup at Member
dmgrTestCell01\SIT2Node01\WebSphere_Portal_SIT2: Received a connection
from an undefined member
dmgrTestCell01\SIT2Node01\WebSphere_Portal_SIT2. Source address is
/9.37.0.137.
Cause
The port used for the DCS_UNICAST_ADDRESS is already in use by another JVM on the same node and this causes the error. In this particular case, two JVMs on the same node (nodeagent and one Application Server) were assigned the same address.
Diagnosing The Problem
Review the configuration data in the server.xml files to see if there is such a port conflict between two JVM processes running on a single physical server.
For example, if this is the DCS_UNICAST_ADDRESS port for Nodeagent.:
<specialEndpoints xmi:id="NamedEndPoint_1188503749604"
endPointName="DCS_UNICAST_ADDRESS">
<endPoint xmi:id="EndPoint_1188503749600" host="*" port="9355" />
Here is the DCS_UNICAST_ADDRESS for the Application Server named SIT2:
<specialEndpoints xmi:id="NamedEndPoint_1188561623176"
endPointName="DCS_UNICAST_ADDRESS">
<endPoint xmi:id="EndPoint_1188561623185" host="SIT2.ibm.com" port="9355" />
Resolving The Problem
Change the configuration so that the DCS_UNICAST_ADDRESS settings are unique for each JVM process on a system. Make the changes, then stop and restart all of the IBM WebSphere JVM processes on that system.
Was this topic helpful?
Document Information
Modified date:
15 June 2018
UID
swg21269678