Hello,
We are preparing to upgrade our db from UDB v7 to v8. We run on in an AIX environment, performing our work via ksh shells. Within these shells we perform the usual db2 sql commands (insert, delete, update, commit etc.).

As part of this process we monitor/check results by piping text into a log which we then parse using awk/sed/grep. The results we get determine our subsequent course of action.

In testing under v8, we've found that, for awk at least, our technique produces a multi line return. Where under v7 we scanned along a single line extracting what we needed, we've had to perform a scan on multiple lines to extract our data.

Has anyone ecountered a similar problem and/or does anyone know if there is some parm I can give to DB2 to tell it to return output in a single line format?
Your thoughts are appreciated.

mr620

Can you give an example, please ?

Sathyaram

Yes.

we call a stored proc:

db2 CALL "get_ctl_id ('${COALITION}', '${SCRIPT_NM}', '${SCRIPT_KEY}', 0)" >${TMP_FILE}

which returns info and redirects into the TMP_FILE

TMP_FILE used to come out on one line.
Coalition is vfj CTL ID is 1014.
We'd then parse this one line for the vfj and 1014 codes.

Under v8 it comes out on multiple lines. TMP_FILE now looks like
Coalition is vfj
CTL ID is 1014

We can of course parse the multiple lines, but thought it might be nice if we could get the DB2v8 system to give the result of the SP back on one line.