TN Clustering errors in a Non-clustered environment

, ,
1

Hi there Experts,
We have a 7.1.1 IS / TN with 2 nodes and they are not clustered.
But we keep seeing this error:

[WmSharedCacheSC.impl.0200E] Failed to create/get distributed cache ‘TNQueryResults@systemsvu.com:9871’; reason: java.lang.RuntimeException: Failed to start Service “Cluster” (ServiceState=SERVICE_STOPPED, STATE_ANNOUNCE); – check configuration

Ive tried to change the following property in tangosol-coherence.xml as per the solution in advantage

stderr
to
logs\tangosol.log

And I have tried to enabled and disable the cluster immediately and restart the IS( also from advantage) …but the error keeps coming back.
This is only on one node. Rest are fine.

Any ideas?
Your help is monumentally appreciated.
-Scooby !

2

the cluster configs are:

tn.cluster.notifyCacheChange=[true|false]
tn.cluster.notifyProfileAddUser=[true|false]
tn.cluster.notifyProfileChange=[true|false]
tn.cluster.sync.remoteAliases=[true|false]
tn.query.threshold=number

true value, for cluster environment

3

We recently experienced a similar situation in our webMethods 8.0 environment.
Thanks to the tip below at softwareAG, we were able to solve the issue by setting following parameter in MyWebmethods & restarting the IS:
tn.cluster.notifyCacheChange=true

[INDENT]1-YSVUF (TNS_6-5_Fix28, TNS_6-1_SP1_Fix3, TNS_6-1_SP1_Fix4, TNS_6-1_Fix26)
Issues with synchronizing external ID types when using multiple Trading Networks servers that share the same database.
When using multiple Trading Networks servers that share the same database, changes made to external ID types are not properly sychronized across all Trading Networks servers.
This issue is resolved. Now all Trading Networks servers are properly synchronized whenever external ID types are added, changed, or deleted. Ensure synchronization is enabled by setting the Trading Networks property, tn.cluster.notifyCacheChange, to “true”.
[/INDENT]

4

Thanks for the update! back in the forum…

5

Errors when we try to get list of Trading Partners or get list of transactions; MWS transaction list times out; Just started happening in Dev and Production; has happened once before in production but bad behaviour “went away” after 2 days. Have restarted IS, MWS and rebooted servers. Have applied all TN patches to dev. Have played with the xxx.clusterxxx settings = false in …WmTN\config\properties.cnf (** where does wm.tn.mime:init get it’s values from on IS restart?)

TN errors:

1203 [WmSharedCacheSC.config.0595] 2011-02-09 19:20:06 PST INFO: Reading cache configuration: distributedCache

followed by

1204 [WmSharedCacheSC.impl.0200] 2011-02-09 19:21:23 PST ERROR: Failed to create/get distributed cache ‘TNQueryResults@bby1b2bap01d:5555’; reason: java.lang.RuntimeException: Failed to start Service “Cluster” (ServiceState=SERVICE_STOPPED, STATE_ANNOUNCE); – check configuration

Thoughts and feedback would be greatly appreciated

6

Scooby - how did you resolve your problem? See my post above.

7

As per SAG Support, enabled cluster, restart IS & TN works. “Toggling” cluster feature apparently rebuilds what must have been corrupt TN (dB) cache.

8

The Coherence-enabled application ( TN uses Coherence for caching query results) on a local machine or in LAN can accidentally form a cluster with the application running on other developers’ machines or other systems

Resolution:
Override the multicast address in tangosol-coherence-override file.
You can either go and edit your tangosol-coherence-override file and fix the multicast address.
OR
Backup the working configs and when you start seeing these errors , restore from backup and you should be good.

9

Fredsmith ,
Ive tried that “enable-clustering and restart TN/IS” but that didn’t work.