Sushil, The answer to your question, should come from comparing the listed capabilities/features of each of these products. SAS DI : SAS Data Integration Studio SAS VA Data Builder: https://support.sas.com/documentation/cdl/en/vaug/67500/PDF/default/vaug.pdf The list of features are too long to put in one answer, but I would recommend DI Studio when you have to - Collaborate amongst multiple developers (Change Management - Check in/out) - Scheduling Jobs - Developing User Custom Transformations Just my 2 cents. Ahmed ... View more

Reading the Requirements again, I believe "Dashboards targeting Desktop system and iOS and Android based tablets" makes the choice tilt towards SAS VA, as it has native support for mobile, whereas SAS EBI would/used to require additional software (Roambi ES for SAS) for such requirements http://support.sas.com/software/products/wrs/index.html#s1=1 ... View more

Either would work for your Reporting/Visualization requirements. SAS VA gives the flexibility of avoiding being tied down to particular/fixed Dimensions and Drill Hierarchies. You or your Power users can compose run-time Drill paths during the Report Design phase. SAS EBI & SAS VA have settle hardware & Infrastructure requirement differences. SAS VA is about 100% RAM centric because of the LASR Analytic Server, but still requires CPU horse power, where as SAS EBI strikes a little bit more balanced dependencies on CPU/RAM/Disk I/O. I would recommend you review the differences using broader set of aspects and considerations beyond Reporting/Visualization needs. Just a thought! Ahmed ... View more

The way I understand it, SAS University Edition runs in its own virtual (Linux) machine, which is "completely" isolated from your Windows (I'm guessing) machine, and the only link between them is how you setup the SAS Shared Folder, based on the installation instructions. With that said, I don't know of a way, whereby you can modify the pre-installed SAS software, nor the Virtual Linux Environment variables, in order to configure the SAS/ACCESS product. By the Way, you'll need SAS/ACCESS to MySQL and NOT SAS/ACCESS to ODBC!! Having SAS/ACCESS to ODBC listed in the proc setinit listing/output, does not mean it is installed! You probably will get better answer posting your question to the Good luck, Ahmed ... View more

Rwan, SAS/ACCESS to ODBC means that you SAS installation can access data stored in ODBC compliant software, such as MS SQL Server, ORACLE, and any other RDBMS that support ODBC connections. Accessing SAS tables via ODBC is completely different. You'll need - Software licenses for either SAS/SHARE or SAS Scalable Performance Data Server (SPDS), or SAS Integration Technology - A running server instance (SAS/SHARE, SPDS, Metadata Server/Object Spawner) - The Corresponding SAS ODBC drivers installed on the Windows machine desired to access the SAS tables Hope this helps, Ahmed ... View more

I would go with Xia Keshan approach (Using Hash Objects) if all the tables were stored in SAS tables/data sets. You can control the the Memory allocated to your SAS Job/Session/Process using the -MEMSIZE xxxM option on the invocation command line. SAS comes with default settings that is specified in the *.cfg  (SAS Configuration) file, and using the -MEMSIZE will override the default setting. Now a days, the minimum requirement for installing SAS v9.x is 4GB RAM/CPU or Core, and that's plenty of RAM!! Otherwise, if your tables are scattered around and stored in mix environments (SAS, Data Base, Flat Files, ....Etc) , Then I would recommend you standardize your storage platform first, then use the best manipulation technique(s). Just my 2 cents, Ahmed ... View more

I have used shell script with SQL statements before, where User Input was passed into the embedded SQL statements within the *.sh file on Linux, you can use Bash shell to display prompts with predetermined options and validation as a mechanism to gather user input, then pass those values to the embedded SQL statements, where applicable...! Just a thought, Hope it helps, Ahmed ... View more

Hi danny, If you want to restrict SAS to only use 1 GB of RAM, then use the -memsize 1GB option within your SAS Windows shortcut. I'm guessing that's how you launch/start your SAS session. If you cannot create unique index for your smaller data set "SEQ_LST" (1,763,513) on CUST_ID TMP_CUST, then I would suggest looking into using Hash Object for as alternative. There are several papers showing how to do that. Check www.lexjansen.com Hope this helps, Ahmed ... View more

This link may have the answers you are looking for SAS® OnDemand for Academics ... View more

@juanvg1972, I think your only option is using SAS® OnDemand for Academics Hope this helps, Ahmed ... View more

Sergio, First, I would recommend you to check out the amount of Memory allocated to the SAS session running on your PC. - Open SAS Session, and submit the following statement      Proc oprions group=memory; run; Second, Once you checked the output of the above in the log, you can increase the allocated Memory amount by specifying the following options in the SAS Command (Probably within the Windows Short-cut) -memsize xxM -sortSize xxM Note: -sortSize should not exceed 75% of the -memSize amount. Third, As I mentioned this before in earlier reply (June 5), Change the value of the   - buffsize= option within your SQL Pass-thru connection clause.   - readbuff= option within your ORACLE Libname statement Forth, Re-Run your test and compare results. Hope this helps, Ahmed ... View more

Hi Sergio, This is typically what I do when it comes to debugging and resolving Performance issues with Queries that involves RDBMS source tables in SAS. 1. Always test the SQL statement(s) within a Native RDBMS Query Tool. This allows you to validate and Profile the query, which in turn, gives you better insight into the execution path and the true cost of the query (Performance Impact). In your case, using ORACLE you are spoiled for choices    1.1 Oracle SQL Developer (Free)    1.2 Toad for Oracle (Commercial)    1.3 .... (Other tools that I'm not familiar with) 2. Use SQL Pass-Through. This allows you to use RDBM's specific syntax that may not be surfaced by the SAS Libname statement options. 3. Change the default BUFFSIZE= option value.           "BUFFSIZE= specifies the number of rows to retrieve from an Oracle table or view with each fetch. Using this argument can improve the performance of any query to Oracle. By setting the value of the BUFFSIZE= argument in your SAS programs, you can find the optimal number of rows for a given query on a given table. The default buffer size is 250 rows per fetch. The value of BUFFSIZE= can be up to 2,147,483,647 rows per fetch, although a practical limit for most applications is less, depending on the available memory." 4. Use the SASTRACE= System Option to determine whether an operation is processed by SAS or is passed to the DBMS for processing. I hope this helps, Ahmed ... View more

Quentin, Here is my take on this interesting subject. I would strongly recommend using Parametrized SAS Stored Processes (SAS Macro Programs) developed using MVC (Model-Viewer-Controller) design pattern. Here is why Model: The application data, business rules, logic, and functions. (Custom query(ies), Data step & Procedures flows) View: The screen/output presentation (Reformat the results depending on the Front-End Client) Controller: Defines the way the user interface reacts to user input (Custom validation and value manipulation) MVC decouples these three objects to increase flexibility and reuse. I personally have been using the above approach for several years now, and I've had many success in re-using the same developed macros to serve different Front-End applications developed in different languages and technologies (Flex, JavaScript, .Net, HTML). The typical protocol I've been using is HTTP Restful Web Services to communicate with SAS via good old SAS/IntrNet and SPWA. Most if not all current programming languages and web frameworks support Restful Web Services, so the world is your oyster when it comes to choosing what to learn. just my two cents, Ahmed ... View more