DSXchange

There is no way of which I am aware to "bulletproof" the requirement that only some jobs are to be read-only, particularly in version 8. There will always be someone with access. You might like to experiment with using file system permissions to protect some or all of the job-numbered objects in the...

Perhaps you need first to re-build the indexes in the Repository. CATEGORY is one of the columns indexed. It may be that the delete operation is trying to use that index.

You could make use of protected project. But there will always be someone with rights enough to delete; in this case anyone who has been added to the Business Manager role. Someone with a big enough magnet in close enough proximity to the disk could also delete things - are you beginning to see that...

Hotel room. Airport lounge is tonight (local time).

This function is for parallel jobs, not for server jobs.

For server jobs look in the SDK functions - there are even some there based on the 1970-01-01 base, which may fulfil your requirements.

Is it just me, or wasn't the message "no rows affected"? Or is this a different thread? Odd, it's not even Monday!

No. The reasoning behind the default is that, in a sequential file, *something* actually has to be there, even if it is "".

Check also that your servers are accessible (names can be resolved) and that your DHCP lease on the client has not expired.

Do you get the same error when running the job independently of the job sequence?

You get the asterisks when processing a date that does not conform to the format string. You need to check your data for such situations.

Don't use a capital "M" in the format string. Use "%dd-%mmm-%yyyy" as the format string. Note, however, that this format string is not applicable if the day number is a single digit; you may need to check for this possibility.

A partition is that subset of the Data Set's rows that are associated with a single processing node. For example, if you created the Data Set with a four-node configuration file, it will have four partitions (if it is large enough). A segment is an individual data file within a partition. For exampl...

You can set the Quote Character property on the Format tab in the Sequential File stage.

If you know what the code point for the character is you can intercept it with a Char() function perhaps in a Convert() function.

Yes, you will need an ODBC driver to access any database using the ODBC stage. However, an ODBC driver for Teradata ships with DataStage. All you ought to need to do is configure the DSN (by editing the .odbc.ini file) and make it available to a project (by editing uvodbc.config in the project).