Are there any batch jobs, user reports, etc running during the times
you are taking the exports or anything cpu-intensive running during
those times?
I've found exp to be fairly cpu-intensive along with the fact that the
version of exp you are using is going to be competing for the buffer
cache along with anything else that may be running during that same
time period, including the other export process.
I've quickly modified a script I wrote for our exports based upon what
you provided that you may want to try. You can log what is happening
on the OS if you like or could even add logic to log into the database
to see if there is contention in the 'while loop'. It will also create
1 log file and 1 dump file per schema so you can see where the time is
being spent per export. I would move the smaller schemas to the front
and the larger to the end of the list so at least the smaller ones
complete.
You'll need to modify and set your environment but it is a start.
#!/bin/ksh
SCHEMAS='GOED HEND HONT MIDD MOOR GRAV OOST SPEN TEGE MIER ODBC WERK
VEER TOAD MEER BERG SOBR ASTE KESS SOED TWEE MILL'
for S in $SCHEMAS
do
EXP_RUN=`ps -ef | grep oracle | grep exp | wc -l`
while [ $EXP_RUN -gt 2 ]
do
sleep 20
# ps -ef | grep oracle | grep exp >> log.file
# vmstat 5 4 >> log.file
EXP_RUN=`ps -ef | grep oracle | grep exp | wc -l`
done
exp owner=$S logfile=$S.LOG file=$S.DMP buffer=4000000 grants=n
&
done
# Wait until exp processes are complete before exiting.
EXP_RUN=`ps -ef | grep oracle | grep exp | wc -l`
while [ $EXP_RUN -gt 1 ]
do
sleep 20
# ps -ef | grep oracle | grep exp >> log.file
# vmstat 5 4 >> log.file
EXP_RUN=`ps -ef | grep oracle | grep exp | wc -l`
done
- "Weerd de E.C. Kirsten" <Kirsten.deWeerd_at_Oranjewoud.nl> wrote:
> Hi,
>
> I'm puzzled. I haven't got a clue. Here's the situation :
>
> During the night we take an export. Due to problems with dmp-files
> getting
> bigger than 2Gig we stopped taking a full export. We now do 2
> exports, each
> with an owner-clause so that all owners are exported in one of both
> files.
> Works great.
>
> The problem is this :
>
> The first export takes about 2 hours, the second about 3 hours. Some
> nights
> however, for a reason I can not think off, the second export takes
> well over
> 6 hours. It has happened before for two nights in a row, then it went
> back
> to aprox 3 hours for a couple of nights. Last night however it was
> still
> running after nearly 6 hours. We had to abort it all for our users to
> go to
> work.
>
> Whatever could be going on, causing the export to perform rather well
> one
> night, and trigger it to be slow the next. We never had problems with
> the
> first export so far ...
>
> I'll unclude the contents of the par-file :
>
> buffer=4000000
>
owner=(GOED,HEND,HONT,MIDD,MOOR,GRAV,OOST,SPEN,TEGE,MIER,ODBC,WERK,VEER,TOAD
> ,MEER,BERG,SOBR,ASTE,KESS,SOED,TWEE,MILL)
> grants=N
> log=/var/opt/export/tax/fulltax_2.log
> file=/var/opt/export/tax/fulltax_2.dmp
>
> We are going to try a different approach tonight. We cut this export
> in two
> exports again, each with about half of the owners stated above.
> However, I'm
> not certain this will solve anything.
>
> So, again I turn to you all. Anyone have any clue what could be going
> on ???
>
> Oracle 7.2.3 on HP-UX 10.20
>
> Thanx in advance, again ...
>
> Greets,
>
> Kirsten
>
> --
> Author: Weerd de E.C. Kirsten
> INET: Kirsten.deWeerd_at_Oranjewoud.nl
>
> Fat City Network Services -- (858) 538-5051 FAX: (858) 538-5051
> San Diego, California -- Public Internet access / Mailing
> Lists
> --------------------------------------------------------------------
> To REMOVE yourself from this mailing list, send an E-Mail message
> to: ListGuru_at_fatcity.com (note EXACT spelling of 'ListGuru') and in
> the message BODY, include a line containing: UNSUB ORACLE-L
> (or the name of mailing list you want to be removed from). You may
> also send the HELP command for other information (like subscribing).
Received on Wed Aug 09 2000 - 08:05:16 CDT