As others have stated, use a file to tag to see if the process completed or is still running. If you can modify your java code, you can have the java code write the little status files. If that's not possible, just wrap it all up in a shell script. Something like;
For the 1st java process
Code:
#!/bin/sh
echo "running" > /var/tmp/process-status.txt
## Insert the method to launch your java code here.
## e.g. $JAVA_HOME/bin/java -jar /path/to/jarfile.jar
echo "complete" > /var/tmp/process-status.txt
And for the 2nd process, something like;
Code:
#!/bin/sh
COUNTER=0
# Do this 10 times. Keeps the process from running forever...
while [ $COUNTER -lt 11 ]; do
if [ -f /var/tmp/process-status.txt ]; then
# Nothing fancy, just snag the return code from grep.
# send the output to /dev/null, because we don't really
# care about it.
grep complete /var/tmp/process-status.txt &> /dev/null
# if grep found complete, $? will be 0, otherwise 1.
if [ $? -eq 0 ]; then
# Insert the code to run your 2nd java app.
# e.g. $JAVA_HOME/bin/java -jar /path/to/jar.file
# remove the status file, so we don't get confused
# the next time we run.
rm /var/tmp/process-status.txt
# the break here will "break out" of the loop, regardless of the
# status of the $COUNTER variable.
break
fi
fi
# the file either didn't exist, or the file doesn't have "complete" in it, so
# we'll wait for 60 seconds and try again.
sleep 60
# increase our counter variable by 1.
let COUNTER=COUNTER+1
done
if [ $COUNTER -eq "11" ]; then
# Our counter variable got too big, which means that the
# loop would stopped. With the sleep in there, we waited 10 minutes
# for the earlier code to populate the file with "complete". We should
# Probably generate an error of some type.
#
# If cron is setup properly, it'll email all output to root@localhost, so
# we'll just let cron handle notifying the operator.
#
echo "WARNING: $0 failed to run. Something is broken."
fi
So the 1st script creates a /var/tmp/process-status.txt file, with the word "running" in it. It then launches your java process. When the java process exits, it writes "complete" into the file.
The 2nd script will only start the 2nd java process if it can find the word "complete" in the /var/tmp/process-status.txt file. If it finds the word "complete" in the file, it will start up the 2nd java process. It'll do this check 10 times, waiting 60 seconds between checks. If all 10 checks fail, then it'll spit out an error. If cron is configured properly, cron will automatically generate an email to root@locahost with any output. You could probably run "mailx" or something to generate a nicer email message or something.
Note I didn't test any of this, so there are probably quite a few bugs, or better ways to do it (e.g. inside a for loop perhaps). I'm sure other posters will point those out. But it should give you a good idea on how to do it.
Hi,
I am a total newbie to all things Unix. I've worked out I need to set up something that will allow me to automatically backup a DB for me, the DB is for a foum system I run.
Now, I've only found out I need to use telnet for this, and worked out hwo to log into telnet today. From here... (4 Replies)
I need to monitor my cron jobs with another unix machine since occasionally the cron will go down on the main server but there are no errors. Can anyone help with a script to write to use the cron on the back up machine to monitor the main server?
I am using SCO and the cron jobs have been... (3 Replies)
I'm trying to run cron jobs to start any inhibited processes after a system reboot.
I can schedule th cron, but i'm confused as to how to incorporated the reboot, since reboot is scheduled at different times, once every month.
How can I write this to start every 15 min after after a reboot
... (2 Replies)
how to Put a cron entry which should be same script triggered on every Saturday and 1st of every month at 01.00 GMT.
0 2 1 * 6 --( At 2.00 GMT every sat & on 1st of every month)
the above syntax is correct?
Thanks (1 Reply)
Hi,
We have a group of hosts using which the cron jobs are submitted...
Few days ago i had submitted a cron job in of these hosts, but unfortunately forgot the host name :(
Can anyone please help me out in finding this host name from which the cron s submitting the job, i dont want the... (2 Replies)
Hi,
I am a Linux administrator (newbie) in my company. The distro being used in the servers here is Centos 5.3
Just need to know, as a Linux administrator is it better for me to use /etc/crontab to set my cron jobs. I do not want to use the crontab -e to schedule my cron jobs.
That means... (1 Reply)
Hi,
please help on this am trying to exec the below mentioned cron jobs
but its getting failed fro the past two days
###but when am trying to execte the cron by the times
23,29 18 * * * /export/home/inrvgo/thelak/China.sh
its getting exec properly
please help on this
#... (8 Replies)
There are two jobs in Solaris , Cron and at jobs..
I know how to disable or enable cron jobs.
How can I enable at jobs and disable it.
Kindly help.
Rj (2 Replies)
I'm running cronjobs on a redhat 5.X. Cronjobs are getting failed frequently so how to find the root cause (2 Replies)
Discussion started by: karthik9358
2 Replies
LEARN ABOUT SUNOS
cron
cron(1M) System Administration Commands cron(1M)NAME
cron - clock daemon
SYNOPSIS
/usr/sbin/cron
DESCRIPTION
cron starts a process that executes commands at specified dates and times.
You can specify regularly scheduled commands to cron according to instructions found in crontab files in the directory
/var/spool/cron/crontabs. Users can submit their own crontab file using the crontab(1) command. Commands which are to be executed only once
can be submitted using the at(1) command.
cron only examines crontab or at command files during its own process initialization phase and when the crontab or at command is run. This
reduces the overhead of checking for new or changed files at regularly scheduled intervals.
As cron never exits, it should be executed only once. This is done routinely by way of the svc:/system/cron:default service. The file
/etc/cron.d/FIFO file is used as a lock file to prevent the execution of more than one instance of cron.
cron captures the output of the job's stdout and stderr streams, and, if it is not empty, mails the output to the user. If the job does not
produce output, no mail is sent to the user. An exception is if the job is an at(1) job and the -m option was specified when the job was
submitted.
cron and at jobs are not executed if your account is locked. Jobs and processses execute. The shadow(4) file defines which accounts are not
locked and will have their jobs and processes executed.
Setting cron Jobs Across Timezones
The timezone of the cron daemon sets the system-wide timezone for cron entries. This, in turn, is by set by default system-wide using
/etc/default/init.
If some form of daylight savings or summer/winter time is in effect, then jobs scheduled during the switchover period could be executed
once, twice, or not at all.
Setting cron Defaults
To keep a log of all actions taken by cron, you must specify CRONLOG=YES in the /etc/default/cron file. If you specify CRONLOG=NO, no log-
ging is done. Keeping the log is a user configurable option since cron usually creates huge log files.
You can specify the PATH for user cron jobs by using PATH= in /etc/default/cron. You can set the PATH for root cron jobs using SUPATH= in
/etc/default/cron. Carefully consider the security implications of setting PATH and SUPATH.
Example /etc/default/cron file:
CRONLOG=YES
PATH=/usr/bin:/usr/ucb:
This example enables logging and sets the default PATH used by non-root jobs to /usr/bin:/usr/ucb:. Root jobs continue to use
/usr/sbin:/usr/bin.
The cron log file is periodically rotated by logadm(1M).
FILES
/etc/cron.d Main cron directory
/etc/cron.d/FIFO Lock file
/etc/default/cron cron default settings file
/var/cron/log cron history information
/var/spool/cron Spool area
/etc/cron.d/queuedefs Queue description file for at, batch, and cron
/etc/logadm.conf Configuration file for logadm
ATTRIBUTES
See attributes(5) for descriptions of the following attributes:
+-----------------------------+-----------------------------+
| ATTRIBUTE TYPE | ATTRIBUTE VALUE |
+-----------------------------+-----------------------------+
|Availability |SUNWcsu |
+-----------------------------+-----------------------------+
SEE ALSO svcs(1), at(1), crontab(1), sh(1), logadm(1M), svcadm(1M), queuedefs(4), shadow(4), attributes(5), smf(5)NOTES
The cron service is managed by the service management facility, smf(5), under the service identifier:
svc:/system/cron:default
Administrative actions on this service, such as enabling, disabling, or requesting restart, can be performed using svcadm(1M). The ser-
vice's status can be queried using the svcs(1) command.
DIAGNOSTICS
A history of all actions taken by cron is stored in /var/cron/log and possibly in /var/cron/olog.
SunOS 5.10 5 Aug 2004 cron(1M)
04-15-2009
But it should give you a good idea on how to do it.