Why ?
In
WebLogic admin console you can see the list of servers and it is health/state ,
if you don’t have a monitoring tools that is connected to WebLogic and check
these servers ( like OEM with WebLogic pack) you have to write a script to get
this status in the file for example. Most monitoring tool will be able to be
configured with the file or shell script and report an alert in certain
condition.
How ?
Below
are steps and a sample code to monitor certain servers in WebLogic, WebLogic
script use a jython which mostly like a python programing language, tab is
considered in the jython code:
1- Create a shell script .sh file with executing
permission to call the WebLogic file and store the result in a text file , this
can be run in a crontab for example every 5 minutes to be like an archived logs
or appended somewhere in a log file:
#!/bin/bash
now=$(date
+%d-%m-%Y_%H-%M)
/oracle/Middleware/Oracle_Home/oracle_common/common/bin/wlst.sh
/export/home/oracle/get_wls_serverstate_all.py
&>/export/home/oracle/weblogic_status.txt
cp
-p /export/home/oracle/weblogic_status.txt /export/home/oracle/archive_weblogic_status/weblogic_status_${now}.txt
2-
Create a jython code file to monitor and exclude certain servers as below
example:
redirect('/dev/null',
'false')
connect('weblogic','<password>','t3://<serverIP>:7001')
import
datetime
servers
= cmo.getServers()
domainRuntime()
for
server in servers:
if server.getName() not in
["",""]:
try:
cd('/ServerRuntimes/' +
server.getName());
currentState =
get('OverallHealthState').getState()
now =
datetime.datetime.now()
current_time =
now.strftime("%d-%b-%y_%H:%M:%S")
dump_file =
'/export/home/oracle/threadDump/'+current_time+server.getName()+'-ThreadDump.log'
if currentState == 0:
print
server.getName() + ': ' + get('State') + ': HEALTH_OK'
elif currentState == 1:
try:
threadDump('true',dump_file,server.getName())
messg1
= 'ThreadDump has been generated on:'+dump_file
except
WLSTException, e:
messg1
= 'Exectpiont occur in threadDump'
print
server.getName() + ': ' + get('State') + ': HEALTH_WARN :'+messg1
elif currentState == 2:
try:
threadDump('true',dump_file,server.getName())
messg1
= 'ThreadDump has been generated on:'+dump_file
except
WLSTException, e:
messg1
= 'Exectpiont occur in threadDump'
print
server.getName() + ': ' + get('State') + ': HEALTH_CRITICAL :'+messg1
elif currentState == 3:
try:
threadDump('true',dump_file,server.getName())
messg1
= 'ThreadDump has been generated on:'+dump_file
except
WLSTException, e:
messg1 = 'Exectpiont
occur in threadDump'
print
server.getName() + ': ' + get('State') + ': HEALTH_FAILED :'+messg1
elif currentState == 4:
try:
threadDump('true',dump_file,server.getName())
messg1
= 'ThreadDump has been generated on:'+dump_file
except
WLSTException, e:
messg1 = 'Exectpiont occur
in threadDump'
print
server.getName() + ': ' + get('State') + ': HEALTH_OVERLOADED :'+messg1
else:
try:
threadDump('true',dump_file,server.getName())
messg1
= 'ThreadDump has been generated on:'+dump_file
except
WLSTException, e:
messg1 = 'Exectpiont occur in threadDump'
print
server.getName() + ': ' + get('State') + ': UNKNOWN_HEALTH_STATE (' +
str(currentState) + ') :'+messg1
except WLSTException, e:
print
server.getName()+': SHUTDOWN: Not_Running'
disconnect()
3- To use a crontab for example :
####
#Weblogic
status
####
0,5,10,15,20,25,30,35,40,45,55
* * * * /export/home/oracle/check_weblogic_all.sh
