2018-12-03

Solving Integrated Weblogic Coherence Issue in JDeveloper 12.2.1.3

Recently after a java update (Oracle JDK 1.8.0_144) and reinstallation of a oracle virtual box version (5.2.18 r), I stumbled across as severe issue regarding my integrated weblogic server as it would not start anymore. Instead it would repeatedly print the following lines:

<03.12.2018 14:51 Uhr MEZ> <Warning> <com.oracle.coherence> <BEA-000000> <2018-12-03 14:51:03.621/41.438 Oracle Coherence GE 12.2.1.3.0 <Warning> (thread=Cluster, member=n/a): This Member(Id=0, Timestamp=2018-12-03 14:50:32.801, Address=xxx.xxx.xxx.xxx:61846, MachineId=27972, Location=site:team-pb.de,machine:pcteam114,process:7564,member:DefaultServer, Role=WeblogicServer) has been attempting to join the cluster using WKA list [/xxx.xxx.xxx.xxx:7574] for 30 seconds without success; this could indicate a mis-configured WKA, or it may simply be the result of a busy cluster or active failover.>

<03.12.2018 14:53 Uhr MEZ> <Warning> <com.oracle.coherence> <BEA-000000> <2018-12-03 14:53:26.720/184.537 Oracle Coherence GE 12.2.1.3.0 <Warning> (thread=Cluster, member=n/a): Delaying formation of a new cluster; unicast networking appears to be inoperable on interface xxx.xxx.xxx.xxx as this process isn't receiving even its own transmissions; consider switching network interfaces>


which stated that the technologically mapped coherence cluster (which consists of only the DefaultServer) was not getting any responses (and thus could not connect to the cluster itself). As there is only one server in our coherence cluster, it will be always the unicast-master and thus needs to check initial pings on itself (as others would ping it in this scenario).

To fix this issue, you could do the following:
  1. Remove the coherence cluster as a whole (as this is the integrated WLS, it seems irrelevant for most cases anyway)
  2. Set the coherence cluster to be in multicast mode. This would lead to n:m pinging in the cluster, but as there is only one server, it does not bring performance issues as you might get in a real Worldwide clustered environment

Solution 1:

To remove the coherence cluster awareness of the default server, you just have to remove the coherence-cluster-system-resource tag from the domains config.xml file (which resides in your $DOMAIN_HOME/config folder).

<server>   
  <name>DefaultServer</name>
     …  
  <coherence-cluster-system-resource>defaultCoherenceCluster</coherence-cluster-system-resource>
</server>

Just remove the highlighted entry and the server should start (without any awareness of coherence).

Solution 2:

To change the type of coherence the cluster is using, you have to add a property inside the CoherenceSystemResource config file (which is located in the $DOMAIN_HOME/config/coherence folder). The correct *-coherence.xml is the one with the same name as the coherence systemResource as seen in Solution 1 (by default: defaultCoherenceCluster-coherence.xml).

Just add the following highlighted line in the coherence-cluster-params element-tag:

  <coherence-cluster-params>
    <cluster-listen-port>7574</cluster-listen-port>
    <clustering-mode>multicast</clustering-mode>
  </coherence-cluster-params>

This also should lead to the server starting again.

Cheers!

Keine Kommentare:

Kommentar veröffentlichen