RHEL 8 (CentOS 8) - Rosetta do not start after server reboot / startup

Symptoms

• With RHEL 8 (CentOS 8), Rosetta do not start following server reboot / startup.
• Running 'systemctl start exlibris.service' might shows:
  >> systemctl start exlibris.service
  ...
  Job for exlibris.service failed because the control process exited with error code.
  See "systemctl status exlibris.service" and "journalctl -xe" for details.
  ...
• Running 'systemctl status exlibris.service' might shows:

  >> systemctl status exlibris.service
  
  exlibris.service - LSB: Startup ExLibris products + oracle
  Loaded: loaded (/etc/rc.d/init.d/exlibris; bad; vendor preset: disabled)
  Active: failed (Result: exit-code) since ...
  Docs: man:systemd-sysv-generator(8)
  Process: 55522 ExecStart=/etc/rc.d/init.d/exlibris start (code=exited,status=203/EXEC)

  ...
  systemd[1]: exlibris.service: control process exited, code=exited status=1
  systemd[1]: Failed to start LSB: Startup ExLibris products + oracle.

  systemd[1]: Unit exlibris.service entered failed state.

  systemd[1]: exlibris.service failed.

Cause

Startup mechanism in RHEL/CentOS 8 was changed, causing current '/etc/rc.d/init.d/exlibris' startup script to fail.

Resolution

1. SSH as 'root' user.

2. Backup '/etc/rc.d/init.d/exlibris'.

3. Replace '/etc/rc.d/init.d/exlibris' content with the following:

#!/bin/bash
#
# Exlibris Starts ExLibris products
#
#
### BEGIN INIT INFO
# chkconfig: 2345 98 10
# description: Startup ExLibris products + oracle
# Provides: $exlibris
# Required-Start: $local_fs $network $remote_fs
### END INIT INFO

i=0
while [ $i -lt 10 ]
do
    if [ -d /exlibris/startup/ ] ; then
        cd /exlibris/startup/
        break
    else
        echo "waiting for mount"
        sleep 10
    fi
    i=$[$i+1]
done

if [ ! -d /exlibris/startup/ ] ; then
    echo "/exlibris/startup not found"
    exit 1
fi
cd /exlibris/startup
case "$1" in
'start')
    if [ -f /exlibris/urm/ora_urm/ora_passwd_u1_1.new.lock ] ; then
        rm -f /exlibris/urm/ora_urm/ora_passwd_u1_1.new.lock
    fi

    ./start_stop start
    [[ -d /var/lock/subsys ]] && touch /var/lock/subsys/exlibris
    # Start the Microfocus development license manager
    if [[ -x /etc/mflmrcscript ]] ; then
        /etc/mflmrcscript
    fi
    ;;
'stop')
    if [ -f fast_shutdown ] ; then
        csh -f fast_shutdown
    else
        ./start_stop stop
    fi
    if [[ -x /opt/microfocus/mflmf/lmfgetpv ]] ; then
        /opt/microfocus/mflmf/lmfgetpv k
    fi
    rm -f /var/lock/subsys/exlibris
;;
esac

4. Make sure to replace any '\r\n' with '\n' for 'exlibris' script:

perl -i -pe 's/\r\n/\n/' /etc/rc.d/init.d/exlibris

5. Grant 'exlibris' script the correct ownership ('root:root') and enough permissions (755):

chown root:root /etc/rc.d/init.d/exlibris
chmod 755 /etc/rc.d/init.d/exlibris

6. Stop then start Rosetta startup service:

systemctl stop exlibris.service
systemctl start exlibris.service

7. Repeat steps 1 - 6 for any other affected application server in the environment.

• Article last edited: 17-Aug-2022