Problem Note 64225: Results are returned more slowly than expected when you access Hadoop views with SAS/ACCESS® Interface to Hadoop
Performance might be slow when you use SAS/ACCESS Interface to Hadoop to query Apache Hadoop views. As part of the query compilation phase, SAS/ACCESS Interface to Hadoop generates the following statements: describe formatted table-name|view-name followed by select * from table-name|view-name.
These two statements are also visible in the SAS® log when you enable the SASTRACE= option, as shown in the following example:
options sastrace=',,,ds' sastraceloc=saslog nostsuffix;
...more data...
HADOOP_4: Prepared: on connection 0
DESCRIBE FORMATTED `default`.`CLASS_VIEW`
HADOOP_5: Prepared: on connection 0
SELECT * FROM `default`.`CLASS_VIEW` LIMIT 0 -- /* initial prepare */
The SELECT statement might run slowly when the query is for a Hadoop view.
Click the Hot Fix tab in this note to access the hot fix for this issue.
Operating System and Release Information
SAS System | SAS/ACCESS Interface to Hadoop | Microsoft® Windows® for x64 | 9.45 | 9.46 | 9.4 TS1M5 | 9.4 TS1M6 |
64-bit Enabled AIX | 9.45 | 9.46 | 9.4 TS1M5 | 9.4 TS1M6 |
64-bit Enabled Solaris | 9.45 | 9.46 | 9.4 TS1M5 | 9.4 TS1M6 |
HP-UX IPF | 9.45 | 9.46 | 9.4 TS1M5 | 9.4 TS1M6 |
Linux for x64 | 9.45 | 9.46 | 9.4 TS1M5 | 9.4 TS1M6 |
Solaris for x64 | 9.45 | 9.46 | 9.4 TS1M5 | 9.4 TS1M6 |
*
For software releases that are not yet generally available, the Fixed
Release is the software release in which the problem is planned to be
fixed.
Type: | Problem Note |
Priority: | high |
Date Modified: | 2019-05-21 14:36:49 |
Date Created: | 2019-05-21 04:53:57 |