Big question...

[hudhwk]

Hi guys,

I have a big problem, I have a process that executes after 7 PM everyday, from time to time it is suspended because of communication problems. What I need is a process that executes on background and monitors the main process. What I've been doing is every 10 mins. I create a text file with the information of the progress of the execution of the process, and I am able to know the status, in this case I need to do a script that checks if the ERROR status is in the text file, then restart the main process. I have checked the WC command but it shows the filename and I just can't find someother way.

Do you have any idea of how to solve this? Thanks in advance.

Daniel Buentello.

 

[PHV]

man grep

Hope This Helps, PH.
Want to get great answers to your Tek-Tips questions? Have a look at FAQ219-2884 or FAQ222-2244

 

[hfaix]

If it was my script........

I would kick it off, writing the PID to a file somewhere and grep for that PID every few minutes. If it's up, great, if it's not, decide if I need to start it back up.

try

ps -ef

or

ps -ef | grep $SCRIPTPID

( where SCRIPTPID equals the PID of the executed script )

 

[SamBones]

If you're on Solaris, then [tt]pgrep[/tt] is the way to go, not [tt]ps[/tt] and [tt]grep[/tt]. Just see the man pages.

Hope this helps.

 

[hudhwk]

Thanks guys for your answers,

Let me be mor specific. I have a file named vg01.log which contains information as follows:

Group   Seq#,LDEV#.P/S,Status,Fence,   % ,P-LDEV# M
VG01    32081  167..S-VOL PSUE ASYNC ,  100    22  -
VG01    51067   22..P-VOL PSUE ASYNC ,   89   167  -
VG01    32081  17d..S-VOL PSUE ASYNC ,  100   105  -
VG01    51067  105..P-VOL PSUE ASYNC ,   98   17d  -
VG01    32081  183..S-VOL PSUE ASYNC ,  100   203  -
VG01    51067  203..P-VOL PSUE ASYNC ,   96   183  -

I have to do a process the searches this file and identifies when a PSUE status is stored, I just don't know how to search the word and then, once it is identified, be able to run the process again.

I know it is a simple IF command but don't know how to identify the PSUE with an script.

I hope this info works and thanks again for your help.

Sincerely,

Daniel Buentello

 

[tbohon]

I have a similar situation - we have a process running which, if it fails, causes us no end of grief in cleanup time and effort.

I wrote the following script in ksh and put it in crontab every 30 minutes. The process I'm looking for is called adt_in .

#!/usr/bin/ksh
 
 /usr/local/bin/cl_stat live | /usr/bin/grep adt_in > tempchk
 FSIZE=`ls -l tempchk | awk '{print $5}'`
 if [ "$FSIZE" = "0" ]; then
         echo "Trxd process adt_in is not active.\n"

(script then sends me an urgent email about the failure)

cl_stat is a script which simply checks the live environment for running processes - so you could probably use ps in it's place. It's proprietary to our vendor so I can't share it here, sorry.

Maybe this will give you a clue ... ???

Tom

"My mind is like a steel whatchamacallit ...

 

[hudhwk]

It worked just PERFECT.

Thanks Tom.

Sincerely,

Daniel Buentello

 

[tbohon]

Glad to be of assistance! I've learned a lot here in this (and other) forums, good to be able to share.

Take care.

Tom

"My mind is like a steel whatchamacallit ...