Periodic Routines

lbegley · Post by **lbegley** » Thu Jan 20, 2005 6:30 am

I am responsible for the DataStage server and the DataStage application software. I am not a DS developer.
Are then any routines, that should be executed periodically, to insure the integrity of DataStage (indexes, cleanup of projects, etc).
I want to be pro-active to my approach in providing DataStage server support.

Thanks in advance.

Sainath.Srinivasan · Post by **Sainath.Srinivasan** » Thu Jan 20, 2005 7:27 am

Steps that you can follow are
1.) Daily export and backup of projects used frequently
2.) Clearing &PH& on periodic intervals
3.) Backup of file system which DataStage developers use
4.) Keeping track of space usage
5.) Backup of db used by DataStage developers

Apart from this, you may require more DataStage expertise such as library files and configuration maintenance.

lbegley · Post by **lbegley** » Thu Jan 20, 2005 7:32 am

Thanks for your response.

Could you elaborate a bit on:
"Apart from this, you may require more DataStage expertise such as library files and configuration maintenance."

Sainath.Srinivasan · Post by **Sainath.Srinivasan** » Fri Jan 21, 2005 8:12 am

Some 'extra' steps are to have your dbms connection defn.s available for DataStage. For e.g. if you use Oracle, keep the tnsnames.ora up-to-date.

Also entries in ODBCINI file updated with your connections.

Have datastage startup and shutdown scripts part of your Unix boot mechanism

Check entries in dsenv file to reflect all relevant libraries to be loaded.

Update datastage config file to have MFILES and T30FILES in an optimal size and also regen whenever modified.

Have steps to release locks whenever DataStage jobs / processes are locked.

If you are using ERP systems such as SAP or Siebel, have the ports opened for developers and provide them a place to land their extracted files.

There are loads of them but may not pertain to your situation. Hence if you let us know your environment, we will be able to inform more and best usages.

lbegley · Post by **lbegley** » Fri Jan 21, 2005 8:16 am

We have a Windows environment.

Thanks!

ray.wurlod · Post by **ray.wurlod** » Fri Jan 21, 2005 3:48 pm

The only piece of advice above that does not apply in the Windows environment is that about the dsenv file. All the rest applies just as much in the Windows environment. ODBC.INI is a Registry entry rather than a file, and maintained through the ODBC Administrator on Windows, but it's still relevant - you need to have out ODBC data sources configured.

One more idea is to spread your disk I/O over as many physical disks (not drive letters) as you can. Use different disks for paging, hashed files, staging areas, DataStage projects.

DataStage is tuned to be reasonable self-managing, and its defaults are good in the majority of cases so that little tuning is required.

DataStage server acts as a client to database servers so that, if you want to use "native" connectivity, that database's client software must be installed and properly configured on the DataStage server machine. You should be able to get DBA assistance in this.