Start/stop scripts for Oracle
Applications
All scripts get kicked off from the
"/etc/init.d/appshost_applications.sh"
that has links as follows:
-rwxr----- 1 root root 1316 Aug 23 07:27
/etc/init.d/appshost_application_control
lrwxrwxrwx 1 root root 38 Aug 23 13:34
/etc/rc1.d/K10appshost_application_control ->
../init.d/appshost_application_control
lrwxrwxrwx 1 root root 38 Aug 23 13:34
/etc/rc2.d/K10appshost_application_control ->
../init.d/appshost_application_control
lrwxrwxrwx 1 root root 38 Aug 23 13:34
/etc/rc3.d/S99appshost_application_control ->
../init.d/appshost_application_control
lrwxrwxrwx 1 root root 38 Aug 23 13:34
/etc/rc4.d/K10appshost_application_control ->
../init.d/appshost_application_control
lrwxrwxrwx 1 root root 38 Aug 23 13:34
/etc/rc5.d/S99appshost_application_control ->
../init.d/appshost_application_control
lrwxrwxrwx 1 root root 38 Aug 23 13:34
/etc/rc6.d/K10appshost_application_control ->
../init.d/appshost_application_control
The "appshost_applications.sh" start the
"start_applications.sh" which is located in
"/mnt/sol_stage/eb/scripts/start_stop" and logs all output
to "/$ENVIRONMENT/log/start_stop"
for overview please see the graphics ...
There are two parts of the "start_applications.sh":
1. "RDBMS" part
1.1 set up the various needed variables, ensure that all
log-files gets created in "/$ENVIRONMENT/log/start_stop",
and
then kicks of the "start_db_part.sh"
2. "Apps" part
2.1 set up the various needed variables, ensure that all
log-files gets created in "/$ENVIRONMENT/log/start_stop",
and
kicks off "start_apps_part.sh"
The sequence for "stop_applications.sh" is exactly the
same, apart for the fact that "Apps" gets run first and
then the "RDBMS" part.
3. "start_db_part.sh"
3.1 set up the various needed variables, ensure that all
log-files gets created in "/$ENVIRONMENT/log/start_stop",
and
test for type of environment: single node, dual node or
RAC. Dependent on type of environment the script will
either:
1. for single-node: "start_local_before_db*.sh" scripts,
start actual rdbms and listener(addbctl.sh and
addlnctl.sh),
"start_local_after_db_*"
2. for dual node: "start_local_before_db*.sh" scripts,
start actual rdbms and listener(addbctl.sh and
addlnctl.sh),
(correction here: have to have "ssh" built in) and then
"start_local_after_db_*.sh"
3. for RAC-nodes: loop through all necessary nodes, and
then do a "ssh" to the node(NOTE: passphrases are ok, but
initial
exchange have to be manual as there are no initial
knowledge of all systems... could be fixed by the system
guys,
but that is not likely to happen!), whereafter the sequence
is for the single-node....
4. "start_apps_part.sh"
4.1