DSXchange

HTH means 'Hope That Helps'.

You need to design 'checkpoints' or restartability into your jobs, there's no magic switch.

It will if the spaces are already part of the field. However, it will not automatically add them for you if they are not there.

Welcome Aboard. Unfortunately, not something I've ever tried or worried about. Until real help comes along, I'm curious... why jump through hoops for a 'silent' install? Typically this would be something needed when multiple installs across servers would be needed, yes? I'm also curious where you fo...

You need to post the complete errors, not just selected snippets. To me this sounds like a resource issue, that your uvconfig parameters may need to be tuned but the error messages would confirm that.

3) 2GB, actually. Unless the hashed file was created with the -64BIT option in which case it is, for all practical purposes, limited only by your available disk space. I'm not exactly following your other questions. If you asked if several links in a job can write to the same hashed file, the answer...

You don't need to pull the value like that. Not for this situation.

Just use it:

I'm gonna have to get me to Amsterdam someday to try out this infamous 'hash file' Ray keeps talking about.

You need to explicitly pad the field if you want to add spaces to it, it won't happen automatically just because a field has been defined as a Char. The Char will allow them to be retained, however. For example: FMT(Link.XXX,"3L") Makes your field left justified and pads it out to ...

If you want to rollback all work when an error occurs, you cannot use Transaction Grouping. Use it when you have multiple related links and you need each 'group' of links to either all be individually committed or rolled back. The Transaction Size is forced to 1 when this option is enabled, which sh...

You got lucky. Sequential files are by their nature multiple read but single write operations. This is not a 'DataStage thing'. More than likely you won't get any kind of error but the output would be... unpredictable... at best. Junk at worst case. Use a larger data set. Run the job more than once...

As noted - you need to keep the hashed file up to date by writing to it any new keys generated during the job run. They then become available for further processing if you disable caching of the hashed file lookup or if you choose 'Enabled, locked for updates'. The former is generally safer.

'Multiple instances' of hashed files will not be generated simply by making a job multi-instance - unless you ensure that by assigning unique names to them. One way you can do this by incorporating the Invocation ID into the hashed file name. Otherwise, the jobs will all create / delete / write to t...

Do you currently support UNIX servers in your organization? Or would this be something new? That could be a deciding factor as well.

XP is not a supported Server platform??? thats not true.. DataStage can be installed on XP and there is no problem with the OS. May be I am missing something which I dont understand. Right. So you really don't have a problem? Open up the Install and Upgrade Guide to Page 2-4. If we've got the same ...