DataStage may not have shut down cleanly. If there are processes still out there waiting on a port after it was shut down, then DataStage will exhibit the symptoms you are describing (no listener after restart).
If you can do so, try the following:
1) Check using the web client (
http://servername:9080) to insure that no-one is logged on by checking "Active Sessions". If they are on, ask them to logout and wait for the active session list to only show your web admin session (hit refresh). Then logout of the web client.
2) Go into the Director and make sure there are no jobs active in any project. Stop schedules if you have to.
3) Once you know everyone is out of DataStage completely, check for slave processes on the server (ps -ef | grep dsapi_slave). If you see slave processes still on, those are what is causing your problem. They are leftover processes from crashed connections (network blip, PC crashed, etc.). To get rid of them, very carefully have the administrator issue a "nice kill" (kill -15 pid) on them. This "nice kill" will allow the process to stop semi-cleanly, cleanup any locks and then notifiy its parent client process (will be a "dscs" process if you look for it with ps) to also logout cleanly. Processes should go away almost immediately. If they don't then you may have to do a hard kill (kill -9 pid) on them. You need to insure that all dscs and dsapi_slave processes are off the system.
4) Next shutdown the DataStage engine (bin/uv -admin -stop).
5) Now do a "netstat -a | grep dsrpc" and see if any "listener" processes are stuck on the port. If you are lucky, it will come back "clean" (empty). If you see any processes out there (usually in a FINWAIT state) they are in timeout mode and will go away in about 10 to 15 minutes. The goal of doing step 3 (getting rid of hung processes) is to hopefully get everything off the system semi-cleanly so you don't have to wait 15 minutes.
6) Once the "netstat -a | grep dsrpc" comes back clean (empty) then you can restart DataStage (bin/uv -admin -start) and it should start cleanly.
[* EDIT - Hmm - took a while to type that response, and after posting I noticed you just posted that there are two DataStage instances on this server - DO NOT follow these instructions then unless you plan on shutting down BOTH instances *]