Thursday, July 17, 2003, Kevan Edwards wrote:

> We have not experienced any performance issues when working with smaller > data sets (n < 100,000) in stand alone mode, but when we access larger > data sets stored on our Unix/Sun server in the client/sever mode our > performance begins to nose dive at a dramatic rate when performing multiple > tasks.

We have a performance problem too, but not quite the same situation as you.

We are working with SPSS 9.0 in a Windows NT 4.0 environment with data sets of about 200,000 records and a file size of about 100 MB. Processing the data in a client/server mode is very slow, and therefore we copy the data to the local PCs before processing, saving about 80% of time.

So the difference between local mode and server based mode seems to be a general problem. But of course it doesn't explain your experiences with SPSS 11. If you can't make local copies of your data, may be you can downgrade to an older version of SPSS?

-- Best regards, Gerd