DataStage BackUp process

sandeepgs · Post by **sandeepgs** » Wed Sep 25, 2013 3:55 am

Hi All,

I would like to know is there any process that helps to automate the process of taking backup's of the code in datastage and maintain versions.

I started wring a batch script but my installation shows that that there is no command "dscmdexport" to take an export of the code. So I stated exploring option with "IBM Information Server manager".

As I am using "IBM Information Server manager" console to create and manage versions manaually within the repository itself.

I have two questions:
1. Is there any command that datastage supports to take backup's
2. Can we schedule the process od taking the backups with in the tool itself if not batch script.

Thanks for your response.

Regards,
sandeep.

rohitagarwal15 · Post by **rohitagarwal15** » Wed Sep 25, 2013 5:04 am

Try using "dscc.exe" command. You can find it where your Datastage clients are instlaled, location like "C:\IBM\InformationServer\Clients\Classic"

priyadarshikunal · Post by **priyadarshikunal** » Wed Sep 25, 2013 5:28 am

dscmdexport is available on client machine and not on server machine and should be executed from the machine where client is installed. Since you are on Windows server, you can very well have client installed there as well. Or you can explore istools.

chulett · Post by **chulett** » Wed Sep 25, 2013 7:44 am

The 'new' utility in the 8.x world is istool. Andy posted a nice overview of the syntax here.

IBM Analytics Champion 2009 - 2020 · Post by **asorrell** » Wed Sep 25, 2013 8:02 am

I'd recommend istool for several reasons.

1) It is a server-based tool so it is significantly faster than using client-based backups.
2) It can backup more than just jobs (important if you are using some of the other products in the suite)
3) It can be automated via shell-script (or batch file), then scheduled

The only drawback is the lack of granularity on an import. It means that you must import the entire project at once. To get a single job you need to import the project into a temporary location then copy the job(s) you need to the correct project.

chulett · Post by **chulett** » Wed Sep 25, 2013 8:13 am

Not to be confused with iStool. And yes, it's perfectly safe to click on the link. I swear.

sandeepgs · Post by **sandeepgs** » Wed Sep 25, 2013 12:09 pm

Hi Chulett,

Thanks for your response.

Yes, I have tried the "istool" command line interface and the testing is successful. Andy had defined in very detail the use of command line interface.

One more point is that as I go understood that "IBM Information Server Manager" is designed to be as a deployment tool which can used as a version control tool as well.

I tried digging through all the IBM materials and found that it can only be used as a version control tool only on integration with CVS/ IBM Rational ClearCase even though it maintains versions in Packages.

So can you confirm my below understanding.

1. DataStage doesn't have a configuration management tool as what we have Informatica.
Ex: In Informatica I can check-in and check-out the code and on check-out it will not allow any other users to modify the code until it is checked back in. This is the same functionality as what we have in configuration management tools Ex: CCC harvest etc.
2. Using Information Server Management tool creating different builds will serve the purpose of creating versions but the versions are created Metadata repository which will in turn eat up entire metadata repository space and can result in degraded server performance.
3. There is no way we can automate the backup process apart from creating a script and scheduling it.

So what will be the recommendation to use separate configuration management tool if at all i have to maintain versions and integrate as what IBM document defines.

Thanks & Regards,
Sandeep

chulett · Post by **chulett** » Wed Sep 25, 2013 12:11 pm

What exact version are you on?

sandeepgs · Post by **sandeepgs** » Wed Sep 25, 2013 12:18 pm

Hi chulett,

I am using Version 9.1.

Regards,
Sandeep.

kduke · Post by **kduke** » Thu Sep 26, 2013 8:01 pm

I would not want to check in and out a full export of the project. You need to export one job at a time. You can script this on the client but you need dsjob to get a list of job names then export these one at a time. You can easily do a check in. Check in on most of these will ignore it if it has not changed.

sandeepgs · Post by **sandeepgs** » Sat Sep 28, 2013 4:20 pm

Thanks for reply,

For writing a script, I got the answer as to use "istool".

But I would more of like to know whether I can use IBM Information Server Management tool as a configuration management tool.

If so, then do I need any third party tool like IBM Rational ClearCase etc.,

Thanks & Regards,
Sandeep

IBM Analytics Champion 2009 - 2020 · Post by **asorrell** » Mon Sep 30, 2013 9:11 am

Sandeep, I find this thread a bit confusing because you keep referring to backups and then switching to talk about source code control products.

Though they both are used to keep copies of source code (or jobs in this case), they are usually not treated as interchangeable or inter-operable products.

I am not aware of any company that depends on source-code control systems for their backups. Though sccs can be used to regenerate one or two items, there are too many drawbacks to depend on them for full system or project restores.

If you have questions on both areas, please split out the source code control discussion to a separate thread and continue discussing backups here.

DSXchange

DataStage BackUp process

DataStage BackUp process

Re: DataStage BackUp process