DSXchange

It sounds like your data volume is low and well within the capabilities of SQL-Server. DS has ODBC drivers for Unix so that it can talk to SQL-Server. For your volumes, Oracle is more trouble than its worth. It's power is in materialized views, parallel query, partitioned tables, etc, all of which i...

You can search the forum or read the DS BASIC manual for hash file OPEN/READ/WRITE/CLOSE operations. There are tons of examples posted on this forum as well. Give it a shot and then post some followup questions if you have issues.

NO. You have an ETL tool, not a database. It's up to you to figure out how a row of source data needs to affect the target. If you put triggers into a database, you're removing T logic from your ETL tool and putting it into the database, where you cannot control performance, processing, or keep trac...

Yes you can, but once you're going down this path, why do you need a GUI ETL tool? Can you state your reasons why you think you need to do this?

SQL-Server is cheap and easy. However, it doesn't have the shear power to deal with large volumes. It also matters how you're going to use the database, OLAP, transactional, whatever. Oracle is expensive and requires careful infrastructure architecture for high volumes, whch SQL-Server can't handle....

Hey, thanks for all the answers Called tech support and they actually called back before anyone on the forum could answer. Come on folks, we can't let tech support show us up!! 2.2 gig files are currently not supported in the CFF stage on linux, but they are now fixed on the Sequential stage (I've v...

Make sure you haven't enabled row-buffering. You don't need to use the locking mechanisms either. As long as only one job is processing against the hash file, you don't need to lock the row because no one else is looking at the file. You don't need the aggregator stage.

Anyone know if there's a 2.2 gigabyte limit on the CFF stage?

Please be patient, my website is down right now and is being recovered. I'll make everything available shortly.

Your issue is one of a recursive traversal using a linked-list. There is no ability using inherent DS stages to do what you wish. You could accomplish what you want two ways: custom DS BASIC function to traverse a hash lookup instead of using a hash stage; or put your data into a table and use SQL (...

Use of TRIM(link.column) removes all leading and trailing spaces; and where multiple spaces reduces to one space as long as it's not a leading or trailing space. Check your Online Documentation DS BASIC manual for the TRIM function.

Check this out:
http://www.dsxchange.com/viewtopic.php?t=85364

Try here:
http://www.pickwiki.com/cgi-bin/wiki.pl

also, do a google on Malcolm Bull for his site (can't get to it right now myself to give you the link, blocked by firewall today).

Also, everything is available on your Online Documentation folder under the Start button for Ascential.

In order to parse the text message being return, you have to look in the job log. This is extremely inefficient. While you can get the DBMSCODE, translating it to the offending column is nearly impossible without inspecting the log message, and even then you may not be able to extract the column nam...

I believe OCONV(link.column, "MCT") converts everything to title case, which will capitalize the first letter of every word in a string.