WAS v8.0 > Migration and coexistence > Distributed operating systems > Scenario 2: Migrating cells across operating systems using the command-line tools > 13. Manually synchronize all nodes
Start and stop the dmgr
The dmgr is an administration application that runs in a special application server. The dmgr is created when WAS ND is installed or when a management profile is created using the dmgr profile template. With the dmgr, we can administer multiple WAS nodes.
Start the dmgr to manage the WAS cell. Stop the dmgr as needed, such as when migrating to a new version of the WAS ND or installing fixpacks.
Start the dmgr
cd WAS_HOME/bin
./startManager command.
If startup throws the following
WSVR0629I: The request buffer for thread pool "server.startup" has reached its capacity
...into...
DMGR_PROFILE/logs/SystemOut.log
...increase the size of the server.startup thread pool request buffer. We can leverage wsadmin.sh for this task.
- Use Jython:
print AdminConfig.list('ThreadPool') tp = "server.startup(thread_pool_path_from_list_command)" print AdminConfig.showall(tp) AdminConfig.modify(tp, [['maximumSize', new_size_value]]) print AdminConfig.showall(tp) AdminConfig.save()
- Use Jacl:
$AdminConfig list ThreadPool set tp "server.startup(thread_pool_path_from_list_command)" $AdminConfig showall $tp $AdminConfig modify $tp {{maximumSize new_size_value}} $AdminConfig showall $tp $AdminConfig save
Set the tp variable to the dmgr thread pool path; for example...
"server.startup(cells/myCell01/nodes/cellMgr/servers/dmgr|server.xml#ThreadPool_startup)"
Run showall twice: First to see the maximumSize value and then, after running the modify command, to ensure that the maximumSize value changed.
On each federated node, start node agents...
startNode.sh
Stop the dmgr
cd $DMGR_PROFILE/bin
./stopManager
To stop the dmgr from the administrative console:
System administration | Deployment manager | Configuration tab | Stop
On each federated node, stop node agents...
stopNode.sh