Tibcohost unable to start managed nodes and throws the error: "TIBCO-AMX-HPA-012116: Failed to start node process. The error may be thrown by wrapper and is not available in the log files."
Article ID: KB0094123
Updated On:
| Products | Versions |
|---|---|
| TIBCO ActiveMatrix Service Grid | - |
| Not Applicable | - |
Description
Description:
If you set the node startup mode to "auto", then tibcohost starts all of its managed nodes with mode "auto" but in some cases the node startup mode changes it to "disabled", and tibcohost is unable to start it. You may get an error message like the following which shows the possible causes.
###############
11 Dec 2015 15:32:52,064 [TibcoHostInstance-tibcohost_DEVB04AMXD02] [ERROR] com.tibco.amx.hpa.NodeManager - TIBCO-AMX-HPA-000265: fatal error starting node SitNode2_2, node has been disabled and will not be startable until you change its start mode from "disabled" to either "manual" or "auto": TIBCO-AMX-HPA-012116: Failed to start node process. The error may be thrown by the wrapper and is not available in the log files.Try starting the process directly or with the --debug option to determine the cause.
###############
The tibamx_<nodename>.tra file points to a node classpath tra that does not exist.
The path to java.library in the node classpath .tra file is incorrect and does not point to a valid JVM library.
On AIX an "argument too long" error is thrown. Increase the maximum length for arguments by calling chdev -l sys0 -a ncargs=16 .
The Java debug port is conflicting with another process. See -Xrunjdwp:transport=dt_socket,server=y,address=3435,suspend=n argument in the .tra file: TIBCO-AMX-HPA-012116: Failed to start node process. The error may be thrown by the wrapper and is not available in the log files.Try starting the process directly or with the the --debug option to determine the cause.
Resolution
- A user may kill the node process in question if the node is running into OutOfMemory exceptions.
- In some cases, the JVM arguments are not properly set (i.e., missing parameter or its value) that may create this issue.
- If you enable the Java debugger on the problematic node and that port is already used by another process, then the node goes into "disabled" mode.
When the node runs into such a problem the "tibcohost" loses the handle to the node and marks the "nodeStartMode" as disabled. First check these options to see if anything went wrong. After fixing the issue, you can check your startup mode by using the tibcohost command ( ./tibcohost describeNodes) and change it to "auto". Use the Tibco host command (./tibcohost setNodeStartMode -nodeName foo -startMode auto). Restart the host that internally will start your node.(./tibcohost start -clearCache).
Issue/Introduction
Additional Information
Feedback
