Some stuff:-
1> Create a common mount point to access log  out files from both concurrent nodes.
     Ex:  /u01/share_drive/conc/log
           /u01/share_drive/conc/out
2>  Set up APPLPTMP  & APPLTMP to a common shared location for both CM nodes.

    Ex :- /u01/share_drive/conc/applptmp
            /u01/share_drive/conc/appltmp
3> Make sure u define utl parameter -/u01/share_drive/conc/applptmp --> First parameter should be this variable
4> Set the value APPLDCP to ON in context file
On both CCM Nodes verify FDCPTRW to 2.
echo $FDCPTRW
If FDCPTRW is not define as environment variable, the you need to update .profile in directory /home/<sid>mgr and add FDCPTRW.  See line below.
export FDCPTRW=2

On 11i database server set SQLNET Expire Time to 1.  The default setting was 10 in sqlnet.ora.
5> Make sure all patches are applied -
6495206
7164202
10145294
9074947

6> Modify $FND_TOP/bin/startmgr file in CM node 

 From:
eval $@ batchmgr
To:
eval $@ batchmgr pmon=2 sleep=30 restart=2

5>Execute Autoconfig on both nodes

Performance patch - 15981176 , 1590009
7> Reset the env
8> Check ur tns entries on both CM nodes to reflect both FNDFS and FNDSM variables in tnsnames.ora file.
9> register if any missing CM nodes using Install  node navigation
10> If you want to configure database instance fail over please set the profile option
Concurrent  PCP instance Check  to ON – so that if any instance is having problem then the managers pointing to that instance will failover to other instance
11> Define Internal Moniors for two CM nodes, like below..  & Service Managers similar to below.
       Manager : Internal Monitor: <PCP1_HOSTNAME>
       ShortName: FNDIM_<PCP1_HOSTNAME>
       Application Name : Application Object Library
      Description : Internal Monitor
     Type: Internal Monitor
     Primary :- <PCP1_HOSTNAME>  ; Platform :- Linux x86-64 (64-bit)
     Program Library : Name : FNDIMON ; Application: Application Object Library

     Workshift:- Standard  Description:- Active 24 hours every day; Processes :- 1 ; Sleep Seconds : 60

Similar to above define another manager - Internal Monitor: PCP2_HOSTNAME

12> Edit $ORACLE_HOME/dbs/_ifile.ora. Add the following parameters:
_lm_global_posts=TRUE
_immediate_commit_propagation=TRUE ( Check any other parameters-- pending only for RAC)

13> Please set TCP/IP parameters to settings listed below on database servers…
TCP KeepAlive Time=200
TCP KeepAlive Intvl=20
TCP KeepAlive Probes=2
TCP retries1=2
TCP retries2=2
TCP Syn Restries=2


Follow the instructions below to verify TCP/IP settings.
1. Log into database node and run the commands below.
a. cd /sbin
b. ./sysctl -a | grep tcp
2. Verify TCP/IP parameters have been changed


14>  Doc ID 1525233.1; 1057802.1