Hi 

I have an environment where i am running TDP for Oracle V2.2 on Solaris 7. My 
TSM server version is 4.2.1.11 and is running on an AIX 4.3.3 server.

My output of "tdpoconf showenvironment" command on the solaris box is shown 
below.

TDP FOR ORACLE INFORMATION
  Version:              2
  Release:             2
  Level:                  0    
  Sublevel:             0
  Platform:             TDP Oracle SUN

TSM SERVER INFORMATION
  Server Name:          TEST_ORACLE
  Server Address:       100.2.3.44
  Server Type:            AIX-RS/6000
  Server Port:             1500
  Communication Method: TCP/IP

SESSION INFORMATION
  Owner Name:         oracle
  Node Name:           TEST_ORACLE
  Node Type:            TDP Oracle SUN
  DSMI_DIR:             /opt/tivoli/tsm/client/api/bin
  DSMI_ORC_CONFIG:      /opt/tivoli/tsm/client/oracle/bin/dsm.opt
  TDPO_OPTFILE:         /opt/tivoli/tsm/client/oracle/bin/tdpo.opt
  Password Directory:   /opt/tivoli/tsm/client/oracle/bin
  Compression:          FALSE

POLICY DATA
  Domain Name:         DBDOMAIN
  Management Class:  STANDARD
  Active Policy Set:     STANDARD
  Backup Retention:    30

The same server also has a backup-archive client running on it. The scheduler 
for the backup-archive clients runs fine and backs up the system.

As soon as I start the scheduler for the TDP, I get these messages in 
dsmsched.log

04/05/2002 17:06:38 Node Name: TEST_ORACLE
04/05/2002 17:06:38 Please enter your user id <TEST_ORACLE>: 

I also get the following message on the console screen

[2] + Stopped (SIGTTOU)        nohup /usr/bin/dsmc sched > /dev/null 2>&1 &

At this point when I do a "ps -ef | grep dsm", I do see two separate "dsmc 
sched" processes running, however the scheduler for TDP is unable to pick up 
any schedule information. Consequently the event goes from a state of PENDING 
to a state of MISSED.

My /opt/tivoli/tsm/client/ba/bin/dsm.sys is given below.

SErvername  TSM
   COMMmethod         TCPip
   TCPPort            1500
   TCPServeraddress   100.2.3.44
   passwordaccess     generate
   schedlogname       /opt/tivoli/tsm/client/ba/logs/dsmsched.log
   schedlogretention  7 D
   errorlogname       /opt/tivoli/tsm/client/ba/logs/dsmerror.log
   errorlogretention  7 D
   largecommbuffers   yes
   inclexcl           /opt/tivoli/tsm/client/ba/bin/Exclude.list
*
*
*
SErvername  test_oracle
   COMMmethod         TCPip
   TCPPort            1500
   TCPServeraddress   100.2.3.44
   passwordaccess     prompt
   schedlogname       /opt/tivoli/tsm/client/oracle/logs/dsmsched.log
   schedlogretention  7 D
   errorlogname       /opt/tivoli/tsm/client/oracle/logs/dsmerror.log
   errorlogretention  7 D
   largecommbuffers   yes
   inclexcl           /opt/tivoli/tsm/client/oracle/bin/Exclude.list
   nodename           test_oracle


As shown, I have a separate stanza for the TDP client and I also have a 
separate  dsm.opt for TDP that points to it. Before starting the scheduler for 
TDP, I issue the command "export 
DSM_CONFIG=/opt/tivoli/tsm/client/oracle/bin/dsm.opt" so it points to the 
correct file.

Lastly, i have already generated a password file in the default install 
directory using "tdpoconf password" command.  This file is named 
TDPO.test_oracle. The oracle user account also has access rights to it.

Why does the scheduler prompt for a password when the password file exists. The 
manual says to set the passwordaccess parameter to prompt and thats why it is 
prompting but shouldnt it pick it up from the password file that has been 
generated. I've tried it a couple of times but same result.

Should the second scheduler session be started from a different directory or 
using a different method? Am i missing something. The manual is very vague on 
the complete process for automating TDP schedules.

Thanks.