DSXchange

Yes, I found that also, but the description is incomplete. We've been experimenting with it, and are getting close. Will advise conclusions in a day or three. Actually, that's not true - the description is complete as far as it goes; HHJLDATE generates a Julian date form. What the customer wants is ...

You may need to remove all three components of the hashed file manually before you can proceed. Something, somewhere, has gotten out of sync. Remove the data portion of the hashed file with rm -rf hashedfile Remove the dictionary portion of the hashed file with rm -f D_hashedfile Remove the VOC entr...

No. This is not a DataStage issue (since the attachments are actually there). It is an issue with your email client.

Depends what you mean by "corrupt". Physical corruption can be tested on a hashed file by hashed file basis with tools such as fixtool or uvfixfile. There is a handy trick where you don't take a backup using uvbackup (that is, you discard the backup image) - the report advises whether there are any ...

If you specify "-" in the format string, then the data must also have this delimiter character. A format string of "%yyyy%mm%dd" would be more appropriate (in both Modify and Transformer stage types) where your souce date is of the form YYYYMMDD. No delimiters, see?

This sounds like a great question to be asking your support provider. If there's something mentioned in the documentation, then your support provider ought to be able to source that something.

The easiest method would be to use an Execute Command activity in a job sequence.

You almost certainly do not want to call an osh script from within a job stream, because the script would then need to be invoked, start up, execute and finish for every row processed by that job.

A job sequence can, of course, contain just a single Job Activity. The run method of the job can be set to "reset if required, then run". Of course, the automatic restart capabilities of that sequence (skip earlier successful activities) is wasted, but you would still, nonetheless (unless you caused...

In the Modify stage the string_trim[character,direction,justify](string) function can be used to trim occurrences of character from the beginning or end of a string and to justify the result if the target string is a fixed-length string. Here, character is the character to be trimmed; it defaults to...

Recompiling invalidates the checkpoints. You must restart without recompiling for this mechanism to work properly.

This is only possible if the two projects are on the same server. If they are on different servers, then you need networking software such as UV/Net. Unfortunately for you if this is your scenario, when DataStage split from UniVerse this is one of the components that was cut out of the DataStage Eng...

This might be more easily solved with three hashed files, each with a single key column, and appropriate constraint expressions testing the results of the individual lookups.

It's probably a good idea to put the command umask 002 into your dsenv script.

Basically no. There is already a loop of sorts within a job ("process each row").

In Job Sequences we have "activities" rather than stages, and two of these (version 7.5.1 and later) are called the StartLoop and EndLoop activities. With these you can loop around activities 3, 4 and 5.

Not totally certain what you mean by "hierarchy" here but, if you mean the hierarchy of control, you define it on paper and implement it with job sequences (for preferences) or job control code (for versions older than 5.2).