Spark + TableStore

Spark access to TableStore

• Prepare a data table

Let us prepare a pet table as an example, where name is the only primary key.

• The example below demonstrates how Spark consumes table data in TableStore.

private static RangeRowQueryCriteria fetchCriteria() {
    RangeRowQueryCriteria res = new RangeRowQueryCriteria("pet");
    res.setMaxVersions(1);
    List<PrimaryKeyColumn> lower = new ArrayList<PrimaryKeyColumn>();
    List<PrimaryKeyColumn> upper = new ArrayList<PrimaryKeyColumn>();
    lower.add(new PrimaryKeyColumn("name", PrimaryKeyValue.INF_MIN));
    upper.add(new PrimaryKeyColumn("name", PrimaryKeyValue.INF_MAX));
    res.setInclusiveStartPrimaryKey(new PrimaryKey(lower));
    res.setExclusiveEndPrimaryKey(new PrimaryKey(upper));
    return res;
}
public static void main(String[] args) {
    SparkConf sparkConf = new SparkConf().setAppName("RowCounter");
    JavaSparkContext sc = new JavaSparkContext(sparkConf);
    Configuration hadoopConf = new Configuration();
    JavaSparkContext sc = null;
    try {
        sc = new JavaSparkContext(sparkConf);
        Configuration hadoopConf = new Configuration();
        TableStore.setCredential(
                hadoopConf,
                new Credential(accessKeyId, accessKeySecret, securityToken));
        Endpoint ep = new Endpoint(endpoint, instance);
        TableStore.setEndpoint(hadoopConf, ep);
        TableStoreInputFormat.addCriteria(hadoopConf, fetchCriteria());
        JavaPairRDD<PrimaryKeyWritable, RowWritable> rdd = sc.newAPIHadoopRDD(
                hadoopConf, TableStoreInputFormat.class,
                PrimaryKeyWritable.class, RowWritable.class);
        System.out.println(
            new Formatter().format("TOTAL: %d", rdd.count()).toString());
    } finally {
        if (sc != null) {
            sc.close();
        }
    }
}

Appendix

For complete sample code, see:

• Spark access to TableStore