TM1 DataSource Name and DataSource Name On Server Parameter

[CapAmerica]

Question:

Now for the DataSource Name and DataSource Name On Server with TM1 10.1.1 I am retrieving a .txt file. Now I have created a process control cube to store the paths and file name and I am using the Prlog. My question though to the TM1 Forum group is on the Data Source tab I have to put a dir path for the DataSource Name and DataSource Name On Server, correct? I believe this is yes. With that said, I have hardcoded my DEV dir path and file name. Now I really don't care about the content in the DataSource Name and DataSource Name On Server as I use the Prolog, but if I migrate this to different environments, I am hard coding my DataSource Name and DataSource Name On Server with DEV info. Though I update my process control cube in the different environments, if you eye ball the TI process by looking at the Data Source tab it could throw someone off. If there anyway I can add a parameter for the DataSource Name and DataSource Name On Server?




TM1Newbie

[lotsaram]

There's no reason to update DataSourceName on the prolog. This is used only to populate the query preview window on the data source tab. The only parameter of relevance is DatasourceNameForServer which dictates the actual file which the TI will process. If you are setting this from a CellGetS on the prolog then whatever is saved in the .pro file and visible on the data source tab is irrelevant as it won't be used at runtime. If you want to avoid potentially confusing a support or dev person looking at the process then you could just DELETE the text in the data source name on server box as it isn't doing didly squat anyway, and make sure it is well commented on the prolog that the datasource is dynamically assigned at runtime from a control cube variable.

[Wim Gielis]

There's no reason to update DataSourceName on the prolog. This is used only to populate the query preview window on the data source tab. The only parameter of relevance is DatasourceNameForServer which dictates the actual file which the TI will process. If you are setting this from a CellGetS on the prolog then whatever is saved in the .pro file and visible on the data source tab is irrelevant as it won't be used at runtime. If you want to avoid potentially confusing a support or dev person looking at the process then you could just DELETE the text in the data source name on server box as it isn't doing didly squat anyway, and make sure it is well commented on the prolog that the datasource is dynamically assigned at runtime from a control cube variable.

Correct, but don't we lose the Variables if we have no Preview?

[qml]

Correct, but don't we lose the Variables if we have no Preview?

No, but you will lose the ability to preview variables. They will still be there, invisible yet unaffected, which you can confirm by the sheer fact that there should be no errors when you save the process, which means all your references to variables defined on the Variables tab are still valid.

Of course If you ever want to see the variables list or edit it, you will need to populate the DataSourceName field again. These are some of the quirks of the good old Editosaurus.

[Wim Gielis]

Correct, but don't we lose the Variables if we have no Preview?

No, but you will lose the ability to preview variables. They will still be there, invisible yet unaffected, which you can confirm by the sheer fact that there should be no errors when you save the process, which means all your references to variables defined on the Variables tab are still valid.

Of course If you ever want to see the variables list or edit it, you will need to populate the DataSourceName field again. These are some of the quirks of the good old Editosaurus.

Hello Kamil

I'm aware of that - it just works fine but we cannot preview our variables. Which is often needed. I agree it's a quirk in the editor.

[lotsaram]

You would only lose the variables if you had a vague moment and clicked on keep derived variables only and then saved changes to the process. Otherwise nothing will change with the saved variables.

TIVarWarning.png (11.75 KiB) Viewed 6702 times

The OP is talking about a situation of migrating from D to P. In a production system there should be no need to preview datasources, or for that matter to open TI processes in edit mode. So really whether the data source preview window is populated or not shouldn't be of any consequence or concern.

Yeah an unknowing admin could refresh the variables and save the process with no variables, but then should such a person have admin rights in the first place???