Thank you, Jeff. That's a really nice
summary. I wanted to make sure to get the part you said correct,
how you said, "Some processes require manual intervention to stop
before they can be started." What I do is make sure that I'm
signed in as the "ONLY execute as "userid" ", so for example, with
Appworx that would be signed in as appworx. Then I do a "ps -ef |
grep appw" to see everything running as appworx, then do the stop
commands, then do the "ps -ef | grep appw" and kill anything still
running if not now stopped, then do the start command -- that's
basically what you mean by "manual intervention", just as one
example w/ appworx, correct?
I'd prefer that these
jobs are coordinated through the dba group however, here are a
few systems which have to be started whenever the production
database goes down and a dba might not be available. Some
processes require manual intervention to stop before they can
be started.
Check if TOMCAT is running on banapp and banapp[2-3-4-5]:
ONLY
execute as "oracle":
/u01/app/oracle/admin/tools/startup/tomcat_start_stop.shl
{start|stop}
banprod
Oracle Weblogic on baninb01, baninb02 and baninb03
ONLY
execute as "oracle":
/u01/app/oracle/admin/tools/startup/start_stop.shl
{start|stop}
Workflow production
on banapp3 and banapp4 only:
ONLY execute as "oracle":
cd
/u05/app/banner/workflow/banprod
stop.sh
start.sh
Appworx production
on chemungv:
ONLY
execute as "appworx":
$HOME/start_stop/start_stop.shl
{start|stop}
Banner Job
Submission etcetera on chemungv:
ONLY execute as "banjob":
$HOME/stop_all
wait and ensure all
"banjob" processes are terminated; manual intervention may
be required.
$HOME/start_all
Tomcat (bu -
secure apps) on ORION:
ONLY execute as "badm":
/u01/app/badm/apps/prod/apache-tomcat-prod/bin/catalina.sh
stop
wait
10 seconds
/u01/app/badm/apps/prod/apache-tomcat-prod/bin/catalina.sh
start
Start scripts are
usually found here:
/u01/app/oracle/admin/tools/startup/
Look
for "start_stop.shl" AND only run this if needed after a
reboot.
Chuck
and I are most familiar with the intricacies of these
systems and will start other services such as Workflow,
etcetera.
I'm not concerned about the application server nodes as those can be easily restarted however, data-recovery is far more problematic for a corrupt database.
- CHARON - banner production (banprod)
- GENESEE - ODS production (odsprod)
- BUDB - bing custom database (bprod), Applications Manager (awxprov) and HR DB (hrdb)
- MAXDB - Facilities Assets (maxprod)
- JADE - Telecom (pinnprod)
- R25DB - Resource 25 (r25prod)
- RSCDB - Events Management