If you suspect some Client sessions did not cleanly end or disconnect, you can perform a clean restart to make sure the DataStage Engine is starting from scratch. Please make sure all clients are logged off and no jobs are running before performing the below steps.
- Log on to server as dsadm or root
- cd $DSHOME
- source the environment:
. ./dsenv # that is dot space dot slash dsenv - Check for DataStage client connections
ps -ef | grep dsapi_slave
ps -ef | grep dscs
If there are any connections found, use kill -9 <process number> to remove them. - Check for jobs processes:
ps -ef | grep phantom | grep java | grep agentf
If there are any connections found, use kill -9 <process number> to remove them. - Check datastage shared memory segments:
ipcs -mop | grep ade
If there any shared memory segments returned, remove them, use command:
ipcrm -m ID , where ID is the number from the second column of ipcs -mop
NOTE: On Red Hat Linux, use option "-mp" instead of "-mop". - Check if there is any port hung using:
netstat -a | grep dsrpc
If you get any results and the dsrpc is in WAIT then you need to wait until this it disappears (you can also reboot the machine). - At this point you can stop or start the DataStage Engine.
bin/uv -admin -stop
bin/uv -admin -start
When you have cleanly restarted the DataStage Engine, try logging in with the DataStage clients