To build a C / S architecture, you can solve the limitations of desktop hardware. The hardware resource of the call server can process other jobs in parallel when the client is data analysis, otherwise the massive data will be "fake" frequently.
There are two modes of the SAS C / S architecture:
First, Integration Technologies: It is actually a bit like a three-layer B / S mode, and the client does not need to install SAS, just use the browser to access the SAS server site. Here I used SAS EG, but unfortunately I can't find EG3.0 to install it (I heard that EG has begun to come with the disk). Which friend can provide a test, please contact me, thank you. The interface is very beautiful. But EG does not seem to support Enterprise Miner.
You can go to the SAS website http://support.sas.com/documentation/index.html to download "How to Start and Test The Sas Integration Technology ".pdf" Reference
Second, Connect: In this mode, the client must install SAS, using the Server login to remotely access the server. First install the SAS software on the server, then start the SPAWNER listening service, and then you can establish a connection through Login on the client. The following detailed how to allow customers to access SAS via SAS SPAWNER, which is relatively simple without rewriting the browser interface, can also achieve data mining of the call server hardware resources, of course, you have to have 1 can be in Windows Server SAS license running on the operating system. This remote access method can use EM, as long as your SAS license is an EM.
SAS SPAWNER
use
:
1. Install the SPAWNER service:
CD / D "SASROOT DIR" spawner -c tcp -security -install
If you change the 23rd port of the default listening, you must define a service port, such as
Myportspawner -c tcp -security -service myport -install
2. Interactive spawner:
Spawner -c tcp -security
3. Add a user who is used for Signon on Windows, which must have "allowed by approved job login" permission
4.SIGNON
Time
If 23, then
% let tcpsec = _prompt_;
or
% let tcpsec = user.pass; signon hostname;
If not 23, then
% let tcpsec = _prompt_;
or
% let tcpsec = user.pass; signon hostname.myport;
5. Submit a statement to the remote SAS to use rsubmit, such as:
Rsubmit; libname crlibm "& libpath"; endrsubmit;
6. To use the far-end library to the local, you can use the following command
LibName local library name server = server name Slibref = remote library name;
7. Turn off the remote connection and use the following command SIGNOFF HOSTNAME;
Thank you for your advice!
