There are two machines, one of which has the admin server and a managed server, and the other just the one managed server. Both managed servers are in a cluster.
In the managed server log of the failed server, the following error is seen:
<Nov 21, 2012 11:18:16 AM CET> <Notice> <WebLogicServer> <BEA-000358> <Started WebLogic Independent Managed Server "Sefira_nadace_cez1" for domain "appl_domain" running in Production Mode>
<Nov 21, 2012 11:20:16 AM CET> <Warning> <JMX> <BEA-149509> <Unable to establish JMX Connectivity with the Adminstration Server AdminServer at service:jmx:t3://aaa.bb.c.ddd:8001/jndi/weblogic.management.mbeanservers.domainruntime.
java.io.IOException
at weblogic.management.remote.common.ClientProviderBase.makeConnection(ClientProviderBase.java:196)
at weblogic.management.remote.common.ClientProviderBase.newJMXConnector(ClientProviderBase.java:84)
at javax.management.remote.JMXConnectorFactory.newJMXConnector(JMXConnectorFactory.java:338)
at javax.management.remote.JMXConnectorFactory.connect(JMXConnectorFactory.java:247)
at weblogic.management.mbeanservers.runtime.internal.RegisterWithDomainRuntimeService.getDomainMBeanServerConnection(RegisterWithDomainRuntimeService.java:222)
Truncated. see log file for complete stacktrace
Caused By: javax.naming.CommunicationException [Root exception is java.net.ConnectException: t3://aaa.bb.c.ddd:8001: Destination unreachable; nested exception is:
java.net.SocketTimeoutException: Read timed out; No available router to destination]
Cause
There are two known causes for this:
1) This can be caused by a combination of :
a) The managed servers were set to listen on *all* addresses
b) The physical machines on which the managed servers were running had multiple active network cards.
In the online help for WLS 10.3.4 through 10.3.6, it states:
"Do not leave the listen address undefined on a computer that uses multiple IP address (a multihomed computer). On such a computer, the server will bind to all available IP addresses."
No consequences are given in the online help doc, but this setup clearly is not recommended.
2) Second known possible cause :
The warning message can be seen when the admin server tries to connect to a managed server unsuccessfully, for example, that for some reason it is not responding. (In the case where this was observed, the managed server was a unix Zombie process.)
Solution
Go into the admin server console, navigate to servers and for each managed server, modify the configuration so that it has a specific listen IP address - that of the Load Balancer.
Note that this problem was tested and not seen on a computer with a single network card.
- or - in the case of an unresponsive managed server, investigate this first; ensure running process if seen has not entered zombie state.
THANKS,
Sandeep
In the managed server log of the failed server, the following error is seen:
<Nov 21, 2012 11:18:16 AM CET> <Notice> <WebLogicServer> <BEA-000358> <Started WebLogic Independent Managed Server "Sefira_nadace_cez1" for domain "appl_domain" running in Production Mode>
<Nov 21, 2012 11:20:16 AM CET> <Warning> <JMX> <BEA-149509> <Unable to establish JMX Connectivity with the Adminstration Server AdminServer at service:jmx:t3://aaa.bb.c.ddd:8001/jndi/weblogic.management.mbeanservers.domainruntime.
java.io.IOException
at weblogic.management.remote.common.ClientProviderBase.makeConnection(ClientProviderBase.java:196)
at weblogic.management.remote.common.ClientProviderBase.newJMXConnector(ClientProviderBase.java:84)
at javax.management.remote.JMXConnectorFactory.newJMXConnector(JMXConnectorFactory.java:338)
at javax.management.remote.JMXConnectorFactory.connect(JMXConnectorFactory.java:247)
at weblogic.management.mbeanservers.runtime.internal.RegisterWithDomainRuntimeService.getDomainMBeanServerConnection(RegisterWithDomainRuntimeService.java:222)
Truncated. see log file for complete stacktrace
Caused By: javax.naming.CommunicationException [Root exception is java.net.ConnectException: t3://aaa.bb.c.ddd:8001: Destination unreachable; nested exception is:
java.net.SocketTimeoutException: Read timed out; No available router to destination]
Cause
There are two known causes for this:
1) This can be caused by a combination of :
a) The managed servers were set to listen on *all* addresses
b) The physical machines on which the managed servers were running had multiple active network cards.
In the online help for WLS 10.3.4 through 10.3.6, it states:
"Do not leave the listen address undefined on a computer that uses multiple IP address (a multihomed computer). On such a computer, the server will bind to all available IP addresses."
No consequences are given in the online help doc, but this setup clearly is not recommended.
2) Second known possible cause :
The warning message can be seen when the admin server tries to connect to a managed server unsuccessfully, for example, that for some reason it is not responding. (In the case where this was observed, the managed server was a unix Zombie process.)
Solution
Go into the admin server console, navigate to servers and for each managed server, modify the configuration so that it has a specific listen IP address - that of the Load Balancer.
Note that this problem was tested and not seen on a computer with a single network card.
- or - in the case of an unresponsive managed server, investigate this first; ensure running process if seen has not entered zombie state.
THANKS,
Sandeep
0 comments:
Post a Comment