Page 1 of 1
Invalid BROWSE Connection
Posted: Wed Feb 15, 2006 1:07 am
by Daddy Doma
Using ProfileStage as a local install on Windows XP Pro SP2. Our PSDB is on SQL Server 2005. My data sources are fixed width text files.
During Column Analysis, I can access Analysis Report, Column Distribution and View Sample with no problems. However, when I try to examine specific records from within the Distribution results I get an error message:
Code: Select all
ProfileStage
Invalid connection, please reset the BROWSE Connection for database xxx
<OK>
Looking through the Help index I know the BROWSE Connection is setup via Connection Manager, and it will Test okay. Yet my error persists.
Thoughts/Suggestions/Answers?
Posted: Wed Feb 15, 2006 3:27 pm
by ray.wurlod
It may have something to do with XP SP2 - check with your support provider as to whether there are any known issues on this platform.
Invalid BROWSE Connection - ProfileStage w/SQL Server 2005
Posted: Wed Feb 15, 2006 6:21 pm
by Daddy Doma
Talking with IBM/Ascential Australia and they indicate the problem may be due to our use of SQL Server 2005. ProfileStage has only been certified with 2k.
We were a little naive when starting our profiling exercise and didn't realise we would need a database server to host the PSDB. The (incorrect) assumption was that ProfileStage would provide its own. Read the installation manual!
In a mad rush, we downloaded a 180-day evaluation copy of SQL Server 2005 off the Internet and, once XP Pro was upgraded to SP2, the only issue was our missing IIS. There was so much trouble getting admin rights for our PCs and the PS software that we couldn't be bothered running the gauntlet again for a Windows install disk. The SQL Server 2005 installation did mention we may lose some functionality...
Our issue with the (missing) BROWSE function is viewing the data we are profiling. The data in the PSDB cannot be viewed through SQL Server because it is encrypted as hexadecimal. We cannot see it through PS. And our data sources are not TRUE fixed width files.
The far right columns are sometimes incomplete and there are no trailing spaces. This would be fine if the final column was ragged right, but the record may end a few columns previously. Hence, my import to SQL Server "wraps" the records and misaligns them.
Zac.
Posted: Wed Feb 15, 2006 11:26 pm
by ray.wurlod
The far right columns are sometimes incomplete and there are no trailing spaces. This would be fine if the final column was ragged right, but the record may end a few columns previously. Hence, my import to SQL Server "wraps" the records and misaligns them.
Pre-process with DataStage to true fixed width. 10 minutes' design work max. Probably under one minute run time.