Cannot start Cache and Inference engine due to “waiting for discovery nodes, retry count”.
Article ID: KB0089354
Updated On:
| Products | Versions |
|---|---|
| TIBCO BusinessEvents Enterprise Edition | - |
| Not Applicable | - |
Description
Cannot start Cache and Inference engine due to “waiting for discovery nodes, retry count” error.
Symptoms:
When Cache server starts the following error is thrown:
[2014-03-10 20:14:01.622][30748][32140][WARNING][XXXXX.$gmp] [MemberManager.cpp, 1072: MemberManager::processMemberConnectTimeout: ]waiting for discovery nodes, retry count [54], discovery [tcp://127.0.0.1:7901]
[2014-03-10 20:14:16.623][30748][32140][WARNING][XXXXX.$gmp] [MemberManager.cpp, 1072: MemberManager::processMemberConnectTimeout: ]waiting for discovery nodes, retry count [57], discovery [tcp://127.0.0.1:7901]
[2014-03-10 20:14:31.625][30748][32140][WARNING][XXXXX.$gmp] [MemberManager.cpp, 1072: MemberManager::processMemberConnectTimeout: ]waiting for discovery nodes, retry count [60], discovery [tcp://127.0.0.1:7901]
Cause:
The error occurs when a discovery URL is on port xyz (example) but no listening URL has been configured on this port. As a result, no discovery node can be started which is why the exception occurs. The format of the property used for both listening and discovering a URL is:
Property=be.engine.cluster.as.listen.url
Value = tcp://IPaddress : Port
Should be set in the PU for the Cache
Property= be.engine.cluster.as.discover.url
Value = tcp://IPaddress : Port
Should be set in the Cluster Tab on CDD file.
Issue/Introduction
Resolution
To resolve the issue, refer to the CDD file and ensure that the property “be.engine.cluster.as.listen.url” matches the port that has been defined for the Discover URL.
Example:
be.engine.cluster.as.listen.url = tcp://127.0.0.1:1214
be.engine.cluster.as.discover.url = tcp://127.0.0.1:1214.
A loopback IP address (127.0.0.1) is a valid entry and can be used.
Additional Information
TIBCO BusinessEvents5.1.1 Developers Guide Chapter 24
Feedback
