multiple servers defined, need one as defualt

[deleted_user]

I have two servers defined that I can see in EG.

I'm trying to process some code on one of them . . . how do I 'point' the code
to execute on the one of choice?

[deleted_user]

To set the default server:

Menu "Tools"
Options...
Respository and server -- under Administration
Select the default server from the list of two that should be there.

To explicity run a process on a specific server, right-click on the code/query and then in the popup menu, there should be choice for "Server" or "Run on ..." or something like that (my EG just broke).

[deleted_user]

OK, EG is un-broke now.

When you select a library from a server, the data is associated with that server.
If you run a query on that data, it will run on the same server.

If you create a code block/item. Then, you have the choice of servers via a right click and "Select Server".

Unfortunately, I have not found out how to select a library from one server, create a query, and point the output of the query to a different server, which is something I have a need for doing.

[deleted_user]

Aha!

I remember now.

In the Query dialog, there is the abilty to set "Options" for that query.
In there you can choose the server the query is to run on.
If you select a different server, it will popup a warning that the data will be copied to the work library of the other server, and that this may take a while.

Again, per some other postings of mine, I wish EG would take better advantage of SAS/Connect, for remote processing and the passing of data from one server to another, in addition to allowing asynchronous running of desparate items in a project.

[deleted_user]

Well if you're really desperate Chuck, there is a workaround to make SAS/Connect work. I set it up for a client under EG 3. I don't recall why it was necessary, possibly the eternal problem of getting data from or to a workstation.

At the time there was some talk about that particular functionality being available around the next release, so it is worth hanging on I think.

As for running asynchronously, that has been identified as desirable.

Kind regards

David

[ChrisHemedinger]

Referring to the "eternal problem" of getting data to/from a workstation, perhaps you should check out the Upload/Download tasks in

http://www.sas.com/apps/demosdownloads/setupcat.jsp?cat=SAS+Enterprise+Guide

These tasks allow you to select local *.sas7bdat files and copy them to a remote server (Upload) and select data sets from libraries on a remote server and copy them to your PC (download).

These are fairly efficient file transfers, so it works fast even for large data. You do not need a local SAS install for these tasks to work.

Chris