org.apache.hadoop.yarn.server.timelineservice.storage.common

Class ColumnHelper<T>

java.lang.Object

org.apache.hadoop.yarn.server.timelineservice.storage.common.ColumnHelper<T>

Type Parameters:

T - refers to the table.

public class ColumnHelper<T>
extends Object

This class is meant to be used only by explicit Columns, and not directly to write by clients.

Constructor Summary

Constructors

Constructor and Description
ColumnHelper(ColumnFamily<T> columnFamily)
ColumnHelper(ColumnFamily<T> columnFamily, ValueConverter converter)
ColumnHelper(ColumnFamily<T> columnFamily, ValueConverter converter, boolean needSupplementTs)

Method Summary

Methods

Modifier and Type	Method and Description
ColumnFamily<T>	getColumnFamily()
static byte[]	getColumnQualifier(byte[] columnPrefixBytes, byte[] qualifier)
static byte[]	getColumnQualifier(byte[] columnPrefixBytes, long qualifier)
static byte[]	getColumnQualifier(byte[] columnPrefixBytes, String qualifier)
ValueConverter	getValueConverter()
Object	readResult(org.apache.hadoop.hbase.client.Result result, byte[] columnQualifierBytes) Get the latest version of this specified column.
<K> Map<K,Object>	readResults(org.apache.hadoop.hbase.client.Result result, byte[] columnPrefixBytes, KeyConverter<K> keyConverter)
<K,V> NavigableMap<K,NavigableMap<Long,V>>	readResultsWithTimestamps(org.apache.hadoop.hbase.client.Result result, byte[] columnPrefixBytes, KeyConverter<K> keyConverter)
void	store(byte[] rowKey, TypedBufferedMutator<?> tableMutator, byte[] columnQualifier, Long timestamp, Object inputValue, Attribute... attributes) Sends a Mutation to the table.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

ColumnHelper

public ColumnHelper(ColumnFamily<T> columnFamily)

ColumnHelper

public ColumnHelper(ColumnFamily<T> columnFamily,
            ValueConverter converter)

ColumnHelper

public ColumnHelper(ColumnFamily<T> columnFamily,
            ValueConverter converter,
            boolean needSupplementTs)

Parameters:

columnFamily - column family implementation.

converter - converter use to encode/decode values stored in the column or column prefix.

needSupplementTs - flag to indicate if cell timestamp needs to be modified for this column by calling TimestampGenerator.getSupplementedTimestamp(long, String). This would be required for columns(such as metrics in flow run table) where potential collisions can occur due to same timestamp.

Method Detail

store

public void store(byte[] rowKey,
         TypedBufferedMutator<?> tableMutator,
         byte[] columnQualifier,
         Long timestamp,
         Object inputValue,
         Attribute... attributes)
           throws IOException

Sends a Mutation to the table. The mutations will be buffered and sent over the wire as part of a batch.

Parameters:

rowKey - identifying the row to write. Nothing gets written when null.

tableMutator - used to modify the underlying HBase table

columnQualifier - column qualifier. Nothing gets written when null.

timestamp - version timestamp. When null the current timestamp multiplied with TimestampGenerator.TS_MULTIPLIER and added with last 3 digits of app id will be used

inputValue - the value to write to the rowKey and column qualifier. Nothing gets written when null.

attributes - Attributes to be set for HBase Put.

Throws:

IOException - if any problem occurs during store operation(sending mutation to table).

getColumnFamily

public ColumnFamily<T> getColumnFamily()

Returns:

the column family for this column implementation.

readResult

public Object readResult(org.apache.hadoop.hbase.client.Result result,
                byte[] columnQualifierBytes)
                  throws IOException

Get the latest version of this specified column. Note: this call clones the value content of the hosting Cell.

Parameters:

result - from which to read the value. Cannot be null

columnQualifierBytes - referring to the column to be read.

Returns:

latest version of the specified column of whichever object was written.

Throws:

IOException - if any problem occurs while reading result.

readResultsWithTimestamps

public <K,V> NavigableMap<K,NavigableMap<Long,V>> readResultsWithTimestamps(org.apache.hadoop.hbase.client.Result result,
                                                                   byte[] columnPrefixBytes,
                                                                   KeyConverter<K> keyConverter)
                                                               throws IOException

Type Parameters:

K - identifies the type of column name(indicated by type of key converter).

V - the type of the values. The values will be cast into that type.

Parameters:

result - from which to reads data with timestamps

columnPrefixBytes - optional prefix to limit columns. If null all columns are returned.

keyConverter - used to convert column bytes to the appropriate key type.

Returns:

the cell values at each respective time in for form {idA={timestamp1->value1}, idA={timestamp2->value2}, idB={timestamp3->value3}, idC={timestamp1->value4}}

Throws:

IOException - if any problem occurs while reading results.

readResults

public <K> Map<K,Object> readResults(org.apache.hadoop.hbase.client.Result result,
                            byte[] columnPrefixBytes,
                            KeyConverter<K> keyConverter)
                          throws IOException

Type Parameters:

K - identifies the type of column name(indicated by type of key converter).

Parameters:

result - from which to read columns

columnPrefixBytes - optional prefix to limit columns. If null all columns are returned.

keyConverter - used to convert column bytes to the appropriate key type.

Returns:

the latest values of columns in the column family. If the column prefix is null, the column qualifier is returned as Strings. For a non-null column prefix bytes, the column qualifier is returned as a list of parts, each part a byte[]. This is to facilitate returning byte arrays of values that were not Strings.

Throws:

IOException - if any problem occurs while reading results.

getColumnQualifier

public static byte[] getColumnQualifier(byte[] columnPrefixBytes,
                        String qualifier)

Parameters:

columnPrefixBytes - The byte representation for the column prefix. Should not contain Separator.QUALIFIERS.

qualifier - for the remainder of the column. Separator.QUALIFIERS is permissible in the qualifier as it is joined only with the column prefix bytes.

Returns:

fully sanitized column qualifier that is a combination of prefix and qualifier. If prefix is null, the result is simply the encoded qualifier without any separator.

getColumnQualifier

public static byte[] getColumnQualifier(byte[] columnPrefixBytes,
                        long qualifier)

Parameters:

columnPrefixBytes - The byte representation for the column prefix. Should not contain Separator.QUALIFIERS.

qualifier - for the remainder of the column.

Returns:

fully sanitized column qualifier that is a combination of prefix and qualifier. If prefix is null, the result is simply the encoded qualifier without any separator.

getValueConverter

public ValueConverter getValueConverter()

getColumnQualifier

public static byte[] getColumnQualifier(byte[] columnPrefixBytes,
                        byte[] qualifier)

Parameters:

columnPrefixBytes - The byte representation for the column prefix. Should not contain Separator.QUALIFIERS.

qualifier - the byte representation for the remainder of the column.

Returns:

fully sanitized column qualifier that is a combination of prefix and qualifier. If prefix is null, the result is simply the encoded qualifier without any separator.