Kbase P104193: Remote monitored resources terminate after fmconfig -update has been run on the fathom-host server
| Autor |
Progress Software Corporation - Progress |
| Acesso |
Público |
| Publicação |
03/11/2008 |
|
Status: Verified
SYMPTOM(s):
Remote monitored resources terminate after fmconfig -update has been run on the machine Fathom is running on
Remote monitoring enabled
Remote monitored database shuts down abnormally
Connection to Admin Server lost. (8848)
read() error in comsocket between Admin Server and PROGRESS broker: s=1124 retc=-1 errno=10053. (8859)
Registration ACK not received, msgcode=-3. (9148)
Error registering with Admin Server. (8845)
Unrecoverable error in connection to Admin Server, running unmanaged. (8850)
Admin Server is shutting down, connection terminated. (9151)
Admin Server rejected registration. (8844)
Database for <dbname> not found.
Unrecoverable error in connection to Admin Server. (9607)
Invalid shutdown request from <FMAGNT> on <tty>. (541)
Begin ABNORMAL shutdown code 2 (2249)
The remote AdminServer shuts down
The remote AdminServer fails to restart with the following admserv.log errors:
Boot failure
Failed to instantiate the management framework container.
Failed to load the management framework container.
The following information alert is received about the update:
Information Alert: Fathom project file upgraded for compatibility with current Fathom version. Old version 3.0A0100, New version 3.0A0200 (10128)
The Fathom Management Console shows that all remote monitored servers (containers) are not active:
Contain Offline: <container_name>
The container is offline. If the container is no longer to be monitored you can delete it from Fathom with the "delete Container" button below. Deleting the container will delete all resource-definitions in Fathom for the container.
FACT(s) (Environment):
All Supported Operating Systems
Fathom Management
OpenEdge Management
CAUSE:
When running "fmconfig -update", the container id get's incremented. Since it was only updated on the fathom-host server, the remote servers still had the previous container id.
FIX:
For each server being remote monitored by Fathom (container):
1.) Shut down all monitored resources and then shutdown the AdminServer
2.) Run the following cmd from the Progress AdminServer's working directory, iow: the directory where the admserv.log is normally written:
fmconfig -update -host <host_Fathom_is_running_on>
3.) MFConfigurationElements.xsd will have been created on sucessful completion
3.) Restart the AdminServer