Configuring Cloudera ODBC Driver for Impala

[Jlochoa]

Dear SAS Community,

My SAS environment is running on Linux servers (9.4M4). I need to configure SAS to connect to a Cloudera Hadoop data source. I have reviewed the SAS documentation "SAS 9.4 Hadoop Configuration Guide for Base SAS and SAS/ACCESS, 4th Edition". I understand from the documentation that I can use the SAS Deployment Manager to pull the necessary set of Hadoop JAR and configuration files. I plan to follow the instructions to allow the full functionality of the SAS/ACCESS drivers that I have. 

However, Is it possible to just download and configure the Cloudera ODBC Driver for Impala without running the SAS Deployment Manager? Will SAS Foundation/BASE interpret the Cloudera ODBC Driver for Impala as a typical ODBC driver connection to Oracle for example?

Thanks,
Jose

[SamLoflin · Accepted solution]

Hi, Jose.

We offer a SAS ACCESS TO HADOOP product for Linux and the Hadoop JAR and configuration files come into play when using this product.

...but we also offer a SAS ACCESS TO IMPALA product which works with an ODBC driver manager and whichever Impala ODBC driver you opt to use.  The assumption is that you will use SAS ACCESS TO IMPALA with a free ware unixODBC driver manager and a Cloudera Impala ODBC driver, but you can configure your ODBC client environment on Linux to accommodate other Impala ODBC drivers such as the one provided by ODBC driver vendor Progress/Data Direct.  See the "SAS/ACCESS Interface to Impala" section of the "Configuration Guide for SAS 9.4 Foundation for UNIX Environments at:

http://support.sas.com/documentation/installcenter/en/ikfdtnunxcg/66380/PDF/default/config.pdf

You can also use SAS ACCESS TO ODBC with an Impala ODBC driver on Linux, including the Cloudera-provided one which you mention.  SAS ACCESS TO ODBC also works with Impala ODBC drivers via an ODBC driver manager.  If you use SAS ACCESS TO ODBC with a Cloudera ODBC driver, it will need to be paired with a unixODBC driver manager.  If you use SAS ACCESS TO ODBC with a Progress/Data Direct Impala ODBC driver, it comes with its own proprietary ODBC driver manager.

The advantage of using SAS ACCESS TO IMPALA is that this engine has been engineered and optimized specifically for Impala.  SAS ACCESS TO ODBC can be used with ant ODBC standard-compliant ODBC driver.

Whether you use SAS ACCESS TO IMPALA or ACCESS TO ODBC, you must on your own install and configure an ODBC driver manager and ODBC driver for Impala and this does not require using the SAS Deployment Manager.  If you currently use SAS ACCESS TO ODBC with other drivers, you will need to take care to configure so that all of the ODBC client components in your Linux environment do not conflict with one another.

We would be more than happy to assist you with any of this.

Thanks,

Sam Loflin

Sr. Technical Support Engineer

SAS ACCESS Technical Support

[Jlochoa]

Thanks Sam, Peter from SAS Tech Support is providing guidance on the driver configurations.