To create a report of failures for that paricular day

rajesekhar · Post by **rajesekhar** » 21 May 2008 5:55

Hi,
At the end of day, I have to create a report of failures(logged in control-M) for that paricular day.
Is it possible to do that by using Control-M reporting facility?
I see that it is not much flexible.

Regards,
rajesekhar

gglau · Post by **gglau** » 22 May 2008 2:00

Look at sample4 of Reporting Facility.

Walty · Post by **Walty** » 22 May 2008 9:23

Hi,
At the end of day, I produced some audit reports for day activity.
I use the <ctmlog> utility to archive the Control-M log every day.

ctmlog list 20080521 0000 20080521 2359 /tmp/20080521.log

In this flat-file i extract the activity like jobs Ko, killed,... based on the code status.

Values are:
5101 job STARTED
5103 job FORCED OK
5105 job SUBMITTED
5120 job EXECUTING
5133 job ENDED OK
5134 job ENDED NOTOK
5401 job HELD
5402 job FREED
5403 job DELETED
5404 job RERUN
5405 job CONFIRMED
5408 job UNDELETED
5409 job KILLED

If you want to extract only the KO jobs

ctmlog listmsg "5134" 2008mmdd hhmm 2008mmdd hhmm /tmp/your file

You can also use the <ctmpsm> utility for extract jobs KO

ctmpsm -LISTJOB NOTOK -SORT JOBNAME

Another way is to run SQL query to extract data from CMR_AJF

select jobname from cmr_ajf where state='8' and status='N' and odate='20080522';
where:
state=8 indicate job Post processed
status='N' indicate job NOT OK or NOT ENDED

Regards
Walty

rajesekhar · Post by **rajesekhar** » 22 May 2008 8:54

@gglau & Walty

With that, I could only get the job whose final status as "ENDED NOT OK"

But I want all the failed jobs(eventhough they were fixed and status changed to OK)

and btw, in the above SQL code, I had to give status like '%NOT%' its a varchar column, is it not the same for all of us?

philmalmaison · Post by **philmalmaison** » 26 May 2008 4:27

tu peux essayer un truc dans le genre :
#!/bin/ksh -x
# extraction des traitements par datacenter
# PhV. 04/03/2008
# Note : il est conseille de compiler les statistiques control-m server avant de lancer ce script
# (ctmjsa *) ou de le lancer juste après s il est deja programme sous controlm

datacenter=${1}
datej=`date '+%y%m%d'`
tmp="${HOME}/ctm/scripts_pv/tmp"
result="${HOME}/ctm/scripts_pv"

cd ${result}

FIC_TEMP="${tmp}/tmp_${datacenter}_${datej}.out"
FIC_CSV="${result}/traitements_${datacenter}_${datej}.csv"

# extraction SQL (jointure entre l'AJF et la RUNINF sur l'ORDERNO)

SQL -w999 <<EOF> ${FIC_TEMP}
select a.APPLIC,";",a.APPLGROUP,";",a.JOBNAME,";",a.MEMNAME,";",b.STARTRUN,";",b.ENDRUN,";",b.ELAPTIME,";",b.OSCOMPSTAT,";",b.NODEID
from CMR_AJF a,CMR_RUNINF b
where a.ORDERNO=b.ORDERNO
and
a.STATUS="N"
ORDER BY APPLIC
go
EOF

# Lecture et mise en forme (suppression des espaces)
echo "APPLI;GROUPE;JOBNAME;MEMNAME;STARTRUN;ENDRUN;ELAPTIME;CODE;HOSTID" > ${FIC_CSV}
cat ${FIC_TEMP} | sed 's/ //g' >> ${FIC_CSV}
rm ${FIC_TEMP}

# purge a 7 jours
find ${result}/traitements* -mtime +7 -exec rm {} \;

c'est plutot light et founi un .csv exploitation sous excel

regards
philmalmaison

naiLS1 · Post by **naiLS1** » 21 Nov 2008 4:58

I need to start producing a daily error report as well. My french is very rusty--can you explain in english what the above script is doing?