compare dsx files

[datastagehunter]

Hi all
We have a four production servers, we are deploying codes in each servers, and we are in assumption that all servers are in same code level, because we deploy the same incremental package,
Is there any way to compare the backups of these production servers and can we can say these all servers code in same level?
Thx

[ray.wurlod]

Probably the easiest is to use diff and ignore the couple of properties (host name, timestamp) that you would expect to be different.

In the next ("Hawk") release there is a side-by-side comparison tool where you can perform this task at a component level (job, routine, etc.) and generate a differences report that is hyperlinked (in Designer) to the actual locations of the difference(s).

[sb_akarmarkar]

Hi,

I think you need to use some third party to to compare two dsx file.
Like VSS, where you can compare two version like that....


Thanks,

Anupam

[vmcburney]

I like Exam Diff which you can get from any shareware download site like the one at zdnet. It has different highlighting of rows that are new, updated or missing. It has inline highlighting of differences.

[trobinson]

I have found diffs on dsx files to be difficult because of all the noise in the files caused by changes that are not really changes. I'm guessing Hawk will do the diff while at the same time filtering out differences that make no difference.
One housekeeping method I try to follow is to track each job with a Release number in the beginning of the short description annotation. If maintained accurately, a BIG IF, this field is visible in Director and Manager so that it becomes relatively easy to see the release level of jobs in different projects easily. Everytime I change ANYTHING on a job I increment the release of that job (ex. 3.2.05 - Fifth change to the 3.2 release). For me this works well.

[ray.wurlod]

Yes it's a big IF, particularly in a team of more than one developer. However, it's a problem that can be avoided if you make it known that you have a baseball bat ready to hand.

[datastagehunter]

Hi
Could anybody please tell me how Hawk version is called in number like data stage 7.5, how this Hawk version called, and please tell me when this version is going to released, please tell me what advanced features they are adding more for this version

Thanks
Ajay

[ray.wurlod]

The number has not yet been announced. The major release number is likely to be 8. We're not sure about the minor release number; a rumour exists that IBM does not like 0 in version numbers, so it will either be 8.1 or 8.0.

Release date depends on what's found during beta testing. "Late 2006" is about as close as one can get.

There is a lot of information available from IBM about what's in it. It's a major change - far too many differences to post here. The entire architecture has been re-vamped. You can find posts that describe some of the new architecture by searching, both here and at IBM. Look particularly for papers from the IBM Information Integration Live event held in Las Vegas in November 2005 and/or search for "Serrano-Hawk".