6.20.08 - Snappy Compression for Aster-Hadoop Connectors - Aster Database

Teradata Aster® Database 6.20 Release Notes

Product
Aster Database
Release Number
6.20.08
Published
November 2019
Language
English (United States)
Last Update
2019-11-12
dita:mapPath
vux1500568440584.ditamap
dita:ditavalPath
External-only.ditaval
The following procedure represents a new sub-section for the Configuring Aster Database and HCatalog section in the Teradata Aster® Database User Guide for Aster Appliances.
  1. If CDH Hadoop version is higher than the CDH Hadoop version used by Aster, then replace the following files/symbolic links from Hadoop cluster to /home/beehive/partner/hadoop/<hadoop-version>/native on the Aster queen node:
    /var/opt/teradata/cloudera/parcels/<hadoop-version>/lib/hadoop/lib/native/libsnappy.so
    /var/opt/teradata/cloudera/parcels/<hadoop-version>/lib/hadoop/lib/native/libsnappy.so.1
    /var/opt/teradata/cloudera/parcels/<hadoop-version>/lib/hadoop/lib/native/libsnappy.so.1.1.1
    /var/opt/teradata/cloudera/parcels/<hadoop-version>/lib/hadoop/lib/native/libsnappy.a

    libsnappy.so and libsnappy.so.1 are symbolic links to libsnappy.so.1.1.1 on Hadoop machines. When you replace them to the Aster native directory, they become files that have the same content as libsnappy.so.1.1.1 and will work whether they are symbolic links are files.

  2. Distribute the native libraries to all nodes by running: ncli node clonefile <filepath>

    Your Aster-Hadoop connector query should now run successfully.