Using the HBase Get Operator


The TIBCO StreamBase® Adapter for Apache HBase is implemented as a suite of five global Java operators, including the HBase Admin, Delete, Get, Put, and Scan operators.

This page describes the HBase Get operator, which allows a StreamBase application to extract rows by ID from the connected HBase database. The operator uses property values in the project's server configuration file to set up the connection to the HBase database, as described in Configuration File Settings. Multiple HBase operators can share a single instance of an HBase connection by selecting the same HBase configuration setting ID.

Configuration File Settings

This section describes the configuration for an HBase database connection instance that you specify in your project's sbconf file. This configuration is the same for all HBase operator types.

HBase Operator Configuration

The <adapter-configurations> element of a project's sbconf file, despite its name, is used to specify configuration value groups for either operators or adapters.

The HBase configuration section of the sbconf file starts with an <adapter-configurations> element that contains one <adapter-configuration name="hbase"> element. This element, in turn, contains one or more <setting> elements.

Each <section name="hbase"> element must contain one element in the form <setting name="id" val="HBaseConfigName"/>, where HBaseConfigName is the name you assign to a group of settings that uniquely define an individual HBase database connection. All other <setting> elements are optional.

The example configuration below shows a basic configuration to connect to an HBase server. You can have as many configurations as your application requires, but each configuration must have a unique id.

Example 1. Example <adapter-configuration> Section for HBase

    <adapter-configuration name="hbase">
      <section name="hbase">
        <setting name="id" val="HBase Sample"/>
        <setting name="connectAtStartup" val="true" />

        <!-- All values below are passed directly to the 
             HBaseConfiguration class unmodified --/>
        <setting name="hbase.master" val="" />
        <setting name="hbase.zookeeper.quorum" val="" />
        <setting name="" val="2181" />
        <setting name="hbase.client.retries.number" val="5" />
        <setting name="zookeeper.session.timeout" val="5000" />
        <setting name="zookeeper.recovery.retry" val="5" />

Configuration Settings

Setting Type Description
id string The value of the id setting displays in the drop-down list in the adapter's Properties view, and is used to uniquely identify this section of the configuration file.
connectAtStartup true or false If true, this operator instance connects to HBase on startup of this operator's containing module.
*** string All other values are directly sent to the HBaseConfiguration class, which is responsible for setting up a connection to the HBase server. See the Apache HBase documentation for the available client configuration options and for further information on setting up a connection to HBase.

Properties View Settings

This section describes the properties you can set for an HBase Get operator, using the various tabs of the Properties view in StreamBase Studio.

General Tab

Name: Use this required field to specify or change the name of this instance of this component, which must be unique in the current EventFlow module. The name must contain only alphabetic characters, numbers, and underscores, and no hyphens or other special characters. The first character must be alphabetic or an underscore.

Operator: A read-only field that shows the formal name of the operator.

Class: Shows the fully qualified class name that implements the functionality of this operator. If you need to reference this class name elsewhere in your application, you can right-click this field and select Copy from the context menu to place the full class name in the system clipboard.

Start with application: If this field is set to Yes (default) or to a module parameter that evaluates to true, this instance of this operator starts as part of the JVM engine that runs this EventFlow module. If this field is set to No or to a module parameter that evaluates to false, the operator instance is loaded with the engine, but does not start until you send an sbadmin resume command, or until you start the component with StreamBase Manager.

Enable Error Output Port: Select this check box to add an Error Port to this component. In the EventFlow canvas, the Error Port shows as a red output port, always the last port for the component. See Using Error Ports to learn about Error Ports.

Description: Optionally enter text to briefly describe the component's purpose and function. In the EventFlow canvas, you can see the description by pressing Ctrl while the component's tooltip is displayed.

Operator Properties Tab

Property Type Description
HBase Config drop-down list The name of the HBase configuration to use with this operator. The value selected by this drop-down list determines the database connection this operator works against. The values that populate this list are stored in the project's sbconf file, as described in the Configuration File Settings section.
Table Name string The HBase table that this operation is to be performed against.
Max Versions int The maximum number of versions to fetch for each row from the database. If this value is 1, the output schema represents a schema for a single version output. If, however, this value is greater than one, the output schema will be modified to represent a list of values with their corresponding timestamp values.
Enable Pass Through Fields check box If enabled, the fields passed into the operator are to be copied into a user-defined field in the output schema.
Pass Through Field Name string The name of the field in the output schema that is to hold the pass-through fields.
Filter Factory Class string The fully qualified name of the class that implements to create the org.apache.hadoop.hbase.filter.FilterList to perform filtering on the rows returned by this operation. Leave blank for no filtering. See the Filter Factory Interface section for more on creating a FilterList class.
Serializer Class string The fully qualified name of the class which implements If set this class will be called to serialize and deserialize to and from tuples.
Row Id Field Name string The field in the inbound schema that represents the field that contains the Row ID to get.
Enable Status Port check box If enabled, a status port is made available for this operator instance, which will emit status tuples for various events from this operator.
Log Level INFO Controls the level of verbosity the adapter uses to issue informational traces to the console. This setting is independent of the containing application's overall log level. Available values, in increasing order of verbosity, are: OFF, ERROR, WARN, INFO, DEBUG, TRACE.

Range Tab

Property Type Description
Time Range Min Field Name string Specifies the field in the incoming schema to be used to specify the starting point timestamp of a time range. Only columns and versions of columns within the specified time range are retrieved; a value of null or empty means do not use.

The default Max Versions setting to return is 1. If your time range spans more than one version and you want all versions within the time range returned, increase the Max Versions setting to a value greater than one.

Time Range Max Field Name string Specifies the field in the incoming schema to be used to specify the end point timestamp of a time range. Only columns and versions of columns within the specified time range are retrieved; a value of null or empty means do not use.

The default Max Versions setting to return is 1. If your time range spans more than one version and you want all versions within the range returned, increase the Max Versions setting to a value greater than one.

Column Filter Tab

Property Type Description
Column Filter Type "Range" or "Columns" Indicates the type of column filtering to perform.
Column Filter Min Range string The minimum value for the column range. If left blank, there is no lower bound.
Column Filter Min Range Inclusive check box If enabled, the value of the minimum column itself is to be included in the range.
Column Filter Max Range string The maximum value for the column range. If left blank, there is no upper bound.
Column Filter Max Range Inclusive check box If enabled, the value of the maximum column itself is to be included in the range.
Columns field grid The columns to return with this query. The Family column of the field grid must have a value, but the Column column can be blank, which directs the operator to include all columns of the specified Family.

Schema Tab

Property Type Description
Row Id Type drop-down list The data type used for the Row ID field.
Convert Value To String check box This option only applies if no user-defined schema is provided. If enabled, this option tries to convert all cell values to a string; if not enabled, all values will be a blob.
HBase Schema schema grid The schema used for scan data results. If no schema is provided, the schema is a list of tuples containing the family column value sets.

Concurrency Tab

Use the Concurrency tab to specify parallel regions for this instance of this component, or multiplicity options, or both. The Concurrency tab settings are described in Concurrency Options, and dispatch styles are described in Dispatch Styles.


Concurrency settings are not suitable for every application, and using these settings requires a thorough analysis of your application. For details, see Execution Order and Concurrency, which includes important guidelines for using the concurrency options.

Filter Factory Interface

Interface Description

The filter factory interface ( is used to create a FilterList for an HBase Scan or Get operation. The Filter list allows HBase to return only those rows that match the criteria provided by the specified filters. The interface allows an interaction between the inbound tuple and the Scan or Get operations. See the HBase API documentation for further information about the org.apache.hadoop.hbase.filter.FilterBase class and its sub-classes, and for how-to information on filtering data.

When you have compiled a Java class to serve as a filter, provide the fully qualified name of your class in the Filter Factor Class field in the Operator Properties tab in your HBase Scan or Get operator's Properties view.

The next section provides an example class file that takes the inbound tuples and uses some of the fields of each tuple to create a filter to use for the read operation.

Filter Factory Example


import org.apache.hadoop.hbase.filter.CompareFilter.CompareOp;
import org.apache.hadoop.hbase.filter.FilterList;
import org.apache.hadoop.hbase.filter.SingleColumnValueFilter;
import org.apache.hadoop.hbase.filter.SubstringComparator;
import org.apache.hadoop.hbase.util.Bytes;


public class DemoFilterFactory implements IHBaseFilterFactory {

  public FilterList createFilterList(Tuple tuple) throws Exception {
    FilterList filterList = 
      new FilterList(FilterList.Operator.MUST_PASS_ALL);
      Bytes.toBytes(tuple.getString("column")), CompareOp.EQUAL, 
      new SubstringComparator(tuple.getString("matchSubString"))));    
    return filterList;

Serialize Interface

Interface Description

The serialize interface ( is used create custom serialization for the tuples. This class will allow you to store and receive the tuples in any format you create to and from the hbase server.

When you have compiled a Java class to serve as a serializer, provide the fully qualified name of your class in the Serializer Class field in the Operator Properties tab in your HBase Scan, Get, or Put operator's Properties view.

The next section provides an example class file that takes the tuples implements the methods required to perform a serialize call to and from the custom data format.

Serializer Example


import java.util.ArrayList;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
import java.util.Map.Entry;
import java.util.NavigableMap;

import org.apache.hadoop.hbase.util.Bytes;


public class DemoSerializer implements IHBaseSerializer {

    private static final String FIELD_COLUMNS = "columns";
    private static final String FIELD_COLUMN = "column";
    private static final String FIELD_FAMILY = "family";
    private static final String FIELD_VALUE = "value";

    private static final String FAMILY1 = "Family1";    
    private static final String COLUMN1 = "Column1";
    private static final String COLUMN2 = "Column2";
    private static final String COLUMN3 = "Column3";
    private static final String COLUMN4 = "Column4";
    private static final String COLUMN5 = "Column5";
    private static final String COLUMN6 = "Column6";
    private static final String COLUMN7 = "Column7";
    Map<String, Map<String, CompleteDataType>> families;
    public DemoSerializer() {
        families = new HashMap<String, Map<String, CompleteDataType>>();
        Map<String, CompleteDataType> family1Columns = new HashMap<String, CompleteDataType>();
        family1Columns.put(COLUMN1, CompleteDataType.forInt());
        family1Columns.put(COLUMN2, CompleteDataType.forLong());
        family1Columns.put(COLUMN3, CompleteDataType.forDouble());
        family1Columns.put(COLUMN4, CompleteDataType.forString());
        family1Columns.put(COLUMN5, CompleteDataType.forBlob());
        family1Columns.put(COLUMN6, CompleteDataType.forTimestamp());
        family1Columns.put(COLUMN7, CompleteDataType.forBoolean());
        families.put(FAMILY1, family1Columns);        
     * This method is used to determine if the class should serialize and deserialize this family column
     * @param family - Will this class handle this family column 
     * @return true if the class can handle this family otherwise false
    public boolean canSerializeFamily(String family) { 
        return families.containsKey(family);
     * This method is used to get the schemas that should be set for the families during deserialize.  This method is called at typecheck time as well as runtime.
     * @param maxVersions - The maximum number of versions that the end user is going to request from the server.
     * @return A map of schemas for the families that this class supports
    public Map<String, Schema> getSchemas(int maxVersions) {
        Map<String, Schema> schemas = new HashMap<String, Schema>();
        for (Entry<String, Map<String, CompleteDataType>> family : families.entrySet()) {
            List<Field> fields = new ArrayList<Field>(); 
            for (Entry<String, CompleteDataType> columns : family.getValue().entrySet()) {
                fields.add(new Schema.Field(columns.getKey(), maxVersions <= 1 ? columns.getValue() : CompleteDataType.forList(columns.getValue())));
            Schema schema = new Schema("", fields);
            schemas.put(family.getKey(), schema);
        return schemas;
     * This method is called to deserialize a family
     * @param schema - The schema of the family
     * @param family - The family
     * @param columnBytes - The family column bytes for each column under the family.  The value NavigableMap<byte[], NavigableMap<Long, byte[]>>. the byte[] is the column name (convert to String via Bytes.toString({value})), the Long is the version timestamp which can be converted to a Streambase timestamp via Timestamp.msecs(Timestamp.TIMESTAMP, {Long time value})
     * @param maxVersions - The maximum number of versions that the end user has requested from the server.
     * @return A tuple fill with the family column data
    public Tuple deserialize(Schema schema, String family, NavigableMap<byte[], NavigableMap<Long, byte[]>> columnBytes, int maxVersions) throws StreamBaseException {
        if (family.equals(FAMILY1)) {
            Tuple tuple = schema.createTuple();
            for (Entry<byte[], NavigableMap<Long, byte[]>> entry : columnBytes.entrySet()) {
                String column = Bytes.toString(entry.getKey());
                List<Object> data = new ArrayList<Object>();
                for (Entry<Long, byte[]> dataEntry : entry.getValue().entrySet()) {
                    if (dataEntry.getValue() == null || dataEntry.getValue().length <= 0) {
                    if (column.equals(COLUMN1)) {
                    } else if (column.equals(COLUMN2)) {
                    } else if (column.equals(COLUMN3)) {
                    } else if (column.equals(COLUMN4)) {
                    } else if (column.equals(COLUMN5)) {
                    } else if (column.equals(COLUMN6)) {
                        data.add(Timestamp.msecs(Timestamp.TIMESTAMP, Bytes.toLong(dataEntry.getValue())));
                    } else if (column.equals(COLUMN7)) {
                if (data != null && data.size() > 0) {
                    tuple.setField(column, maxVersions <= 1 ? data.get(0) : data);

            return tuple;
        return null;

     * This method is called only for the PUT operations to determine if the input schema provided by the user is valid for the serialize method
     * @param family - The family, this value may be an empty string if the user input is a generic input schema which consists of tuple(columns list((column string, family string, value string))).
     * @param schema - The schema given by the end user for the family
     * @throws TypecheckException - Throw a typecheck exception if the schema is invalid
    public void typecheckSerialize(String family, Schema schema) throws TypecheckException {
        if (family.isEmpty()) { // no family may happen if a generic schema is given tuple(columns list((column string, family string, value string))), you can choose to support this or not
            // we support the generic case only with string as the value type
            try {
                if (!schema.getField(FIELD_COLUMNS).getElementType().getSchema().getField(FIELD_VALUE).getCompleteDataType().equals(CompleteDataType.forString())) {
                    throw new TypecheckException("Generic fields are only supported with a string type");    
            } catch (TupleException e) {
                throw new TypecheckException("Error checking generic fields: " + e.getMessage(), e);
        Map<String, CompleteDataType> family1Columns = families.get(family);
        if (family1Columns != null) { // check to make sure that the schema field types match
              for (Field field : schema.fields()) {
                  CompleteDataType completeDataType = family1Columns.get(field.getName());
                  if (completeDataType == null) {
                      throw new TypecheckException("Invalid field [" + field.getName() + "] found in family [" + family + "]");
                  if (!field.getCompleteDataType().equals(completeDataType)) {
                      throw new TypecheckException("Field [" + field.getName() + "] must have a datatype of [" + completeDataType + "]");

     * This method is called to serialize a tuple into data for the put operation
     * @param family - The family to serialize
     * @param tuple - The input tuple to convert
     * @return A map of family column names and their associated data 
    public Map<String, byte[]> serialize(String family, Tuple tuple) throws StreamBaseException {        
        if (family.equals(FAMILY1) && tuple != null) {
            Map<String, byte[]> familyData = new HashMap<String, byte[]>();
            if (tuple.getSchema().hasField(FIELD_COLUMNS)) { // check for the generic form of a tuple that we have stated via typecheck that we support
                List<?> columns = tuple.getList(FIELD_COLUMNS);
                for (Object familyColumnObject : columns) {
                    Tuple familyColumnTuple = (Tuple)familyColumnObject;
                    String genericFamily = familyColumnTuple.isNull(FIELD_FAMILY) ? null : familyColumnTuple.getString(FIELD_FAMILY);
                    String genericColumn = familyColumnTuple.isNull(FIELD_COLUMN) ? null : familyColumnTuple.getString(FIELD_COLUMN);
                    String genericValue = familyColumnTuple.isNull(FIELD_VALUE) ? null : familyColumnTuple.getString(FIELD_VALUE);
                    if (genericFamily != null && genericFamily.equals(family) && genericColumn != null && genericValue != null) {
                        familyData.put(genericColumn, Bytes.toBytes(genericValue));
            } else {                
                for (Field field : tuple.getSchema().fields()) {
                    if (tuple.isNull(field)) {
                    if (field.getName().equals(COLUMN1)) {
                        familyData.put(field.getName(), Bytes.toBytes(tuple.getInt(field)));                        
                    } else if (field.getName().equals(COLUMN2)) {
                        familyData.put(field.getName(), Bytes.toBytes(tuple.getLong(field)));                        
                    } else if (field.getName().equals(COLUMN3)) {
                        familyData.put(field.getName(), Bytes.toBytes(tuple.getDouble(field)));                        
                    } else if (field.getName().equals(COLUMN4)) {
                        familyData.put(field.getName(), Bytes.toBytes(tuple.getString(field)));
                    } else if (field.getName().equals(COLUMN5)) {
                        familyData.put(field.getName(), tuple.getBlobBuffer(field).copyBytes());
                    } else if (field.getName().equals(COLUMN6)) {
                        familyData.put(field.getName(), Bytes.toBytes(tuple.getTimestamp(field).toMsecs()));
                    } else if (field.getName().equals(COLUMN7)) {                        
                        familyData.put(field.getName(), Bytes.toBytes(tuple.getBoolean(field)));
            return familyData;
        return null;