Hadoop SSTable is an InputFormat implementation that supports reading and splitting Cassandra SSTables. Leveraging this input format MapReduce jobs can make use of Cassandra data for offline data analysis.
Currently Cassandra 2.1.x is not supported.
See a full example to get a feel for how to write your own jobs leveraging hadoop-sstable.
Cassandra Create Table Statement (used for table metadata)
hadoop.sstable.cql="CREATE TABLE foo..."
The Compressed SSTable Reader uses off-heap memory which can accumulate when task JVMs are reused.
Additionally, each MapReduce job written using this input format will have it's own set of constraints. We currently tune the following settings when running our jobs.
io.sort.mb io.sort.factor mapred.reduce.tasks hadoop.sstable.split.mb mapred.child.java.opts
Binaries and dependency information for Maven, Ivy, Gradle and others can be found at [http://search.maven.org]
Example for Gradle:
$ git clone [email protected]:fullcontact/hadoop-sstable.git $ cd hadoop-sstable $ ./gradlew build
For bugs, questions and discussions please use the Github Issues.
Copyright 2014 FullContact, Inc.
Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at
Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.