Thank you Carel:
"It depends how you have defined the ODBC SQL statement. Is the PF keyed
and index rebuild set to *IMMED? We had a similar procedure to update
about 16 files, the smallest with about 100 records (halve of the files)
and the largest about 75000 records, others under 30000 records. This took
more than 12 hours. We changed it to AS/400's impor/excport function
(CPY__IMPF) and it ran within 2,5 hours. Do you really need to do this
with ODBC?"
No, we do not have to do this with ODBC. It was the preferred option at
the time. We can explore this iSeries import option. However, the data
we are building right now is greater than 65000, out to a MS Access
database, which is linked to a DB on the iSeries. I know that we can
export the data into a comma delimited file but we can't export it because
it says that is too big for such a file. We are bullring the database
with a VB program and or issue is that the throughput of data is
generating a bottle neck of access, slowing down the process so bad that
we can even count the records processed on the screen! We were able to
process other tasks with a little less number of records with no
speed/performance issues but when we get to this file, the writing to the
linked file (iSeries DB file) is so tremendously slow, that we do not know
what may be causing this.
Peter Vidal
PALL Corporation / SR Programmer Analyst, IT Development Group
10540 Ridge Rd., Ste 203, New Port Richey, FL 34654-5111
http://www.pall.com
"Imagination is more important than knowledge…"
Albert Einstein (1879-1955)
midrange.com
