UNIX for Dummies Questions & Answers

Hi,

I am a c# developer and new to linux/unix environment after the prev developer resigned and is expected to swim but I am drowning

My problem is with keepalive. Currently there are 90+ nodes/jobs that run under keepalive. I managed to get code that shows me the count of these nodes/jobs that are running under keepalive viz.

Code:

ps -eaf | grep -w "sh" | grep -v grep | grep -v $$ | wc -l

This command is excecuted every 5 minutes to make sure that the number of jobs/nodes under keepalive stay consistant. Now the bossman wants to know the name of the jobs/nodes that at anytime no longer run under keepalive. The logic is at follows:

step 1: get all the jobs/nodes running under keepalive in the morning (isit keepalive )
step 2: monitor the count of jobs/nodes running under keepalive (have this )
step 3: if the count (step 2) change record the names of the jobs/nodes that is no longer in keepalive into a file

Please help. Thanx.

Moderator's Comments:

Use code tags, thanks.

hi,
have a look at man diff, diff may help to check for differences in files, line by line.

On the other hand, if you post a couple of example output files from the above checks and the desired output someone may give you more specific directions.
see ya
fra

Sorry, the code should be

Code:

ps -eaf | grep -w "/keepalive.sh" | grep -v grep | grep -v $$ | wc -l

The result of this command is a three digit number like the keepalivecount.gif thats attached.

I need a command that will check if a certain .sh file is still running under the keepalive monitoring.

Example :

Code:

$ keepalive.sh keepalive247.sh newmq.cfg risk_engine_mqin 60

This says that the keepalive process will check every 60 seconds if the process/jobs/node "risk_engine_mqin", that is part of the newmq.cfg configuration, is running. If not, it will be started.

It happens from time to time that the server gives a problem such as permissions or space etc and that causes certain process/jobs/nodes or even the keepalive process itself to fail.

I need code to give me a list of all the process/jobs/node that is managed by keepalive so that I can compare the list if the count starts to go down in order to identify the problematic process/jobs/node.

Hope this makes more sense.

Hi, in the

Code:

ps -eaf ...

that you showed before, if you try to delete the last redirection (that is, "| wc -l") it will list all the processes containing keepalive.sh in their command definition, plus many other infos.

also, if you replace "wc -l" with:

Code:

awk ' { print $NF } '

you will see a list of processes only.

You may start from there to write instructions to address the issue.

see ya
fra

thanx Frappa !