OpenDeploy 5.6 UNIX Recerver Error

dtheimer

ISSUE:
The UNIX server crashes, after it reboots and OpenDeploy is being started using the S80iwod script. The script fails and complains that the RMI registry is already running, so OpenDeploy never comes up... when you stop it, then start it back up, it works...


SERVER LOG
Utlb33# ./S80iwod start
S80iwod:20030305:1700: Starting OpenDeploy server and its RMI registry.
S80iwod:20030305:1700: OpenDeploy server home is /apps/opendeploy/5.6.0/OpenDeployNG
Get RMI server binding hostname/IP address from
/apps/opendeploy/5.6.0/OpenDeployNG/etc/deploy.cfg.
deploy.cfg file did not specify a RMI server binding
hostname/IP address.
Will use default which is the local hostname utlb33
WARNING: OpenDeploy server or its RMI registry is already running.
WARNING: Please stop OpenDeploy server before starting.


utlb33# ./S80iwod stop
S80iwod:20030305:1700: Stopping OpenDeploy server and its RMI registry.
S80iwod:20030305:1700: OpenDeploy server home is /apps/opendeploy/5.6.0/OpenDeployNG

OpenDeploy server and its RMI registry stopped.

utlb33# ./S80iwod start
S80iwod:20030305:1701: Starting OpenDeploy server and its RMI registry.
S80iwod:20030305:1701: OpenDeploy server home is /apps/opendeploy/5.6.0/OpenDeployNG
Get RMI server binding hostname/IP address from
/apps/opendeploy/5.6.0/OpenDeployNG/etc/deploy.cfg.
deploy.cfg file did not specify a RMI server binding
hostname/IP address.
Will use default which is the local hostname utlb33
Started RMI registry on port 9173 for OpenDeploy server.
Starting OpenDeploy server.
It may take a few moments. Please wait....
OpenDeploy server and its RMI registry started.

SYSTEM INFO:
Sun Microsystems Inc. SunOS 5.8 Generic Patch October 2001

Migrateduser

The OpenDeploy development group is aware of this problem.

Normally, when Solaris is shutdown, it executes the RC
shutdown script for OpenDeploy which takes OpenDeploy
down cleanly.

When the Solaris server crashed, it was not possible to
execute the shutdown RC script so OpenDeploy doesn't
get cleaned up properly.

During the boot time right after the Solaris server crash, OpenDeploy startup detects this and mistakenly thinks it is still running and issues the messages you are seeing.

To recover from this, issuing "iwod stop" will do the cleanup
needed and the subsequent "iwod start" will bring OpenDeploy up. I think this is what you have done.

This problem occurs only when the Solaris server goes down
abnormally.

This problem will be fix in OpenDeploy 5.6 Service Pack 1 so during boot time, if OpenDeploy startup detects this situation, it will do the necessary cleanup then continue to bring OpenDeploy up.

- Anna