Archives of the TeradataForum

Message Posted: Fri, 24 Jan 2003 @ 18:45:08 GMT

	<Prev	Next>		<<First	<Prev	Next>	Last>>

we run this through bteq using unix sys v cron to kick it off once a week in the wee hours

delete from dbc.sw_event_log where thedate lt date-30;
delete from dbc.hw_event_log where thedate lt date-30;
delete from dbc.logonoff where logdate lt date-60;
delete from dbc.resusagespma where thedate lt date-30;
delete from dbc.resusageipma where thedate lt date-30;
delete from dbc.resusagescpu where thedate lt date-30;
delete from dbc.resusageicpu where thedate lt date-30;
delete from dbc.resusagesctl where thedate lt date-30;
delete from dbc.resusagesvpr where thedate lt date-30;
delete from dbc.resusagesvpr2 where thedate lt date-30;
delete from dbc.resusageivpr where thedate lt date-30;
delete from dbc.resusageshst where thedate lt date-30;
delete from dbc.resusagesobj where thedate lt date-30;
delete from dbc.resusagesldv where thedate lt date-30;

there are macros that do this.... we just clean out what we use....

we have lots of little scripts that do stuff for us.... like this

i justed finished playing with this one it tails any mail file with size greater then 0 keep in mind i have no users on SYS V unix boxes so this is not alot of crap. It has helped me to find cron processes that do not run correctly and spew out errors into mail....

#!/bin/ksh
#
# simple mail checker to see if we have mail on any node
#
dirit=`ls -s /var/mail | grep -v ':'| grep -v 0 | grep -v total | awk
'{print $2}' |  xargs `
whoit=`uname`
print $dirit
for i in `print ${dirit}`
do
tail -40 /var/mail/$i | mailx -s " $whoit /var/mail/$i"
[email protected]
done

Christopher Platt
Teradata Systems Administration

	<Prev	Next>		<<First	<Prev	Next>	Last>>