Hbase依賴
<!-- 包含HbaseClient和聚合工具類 -->
<dependency>
<groupId>org.apache.hbase</groupId>
<artifactId>hbase-endpoint</artifactId>
<version>2.2.2</version>
</dependency>
配置文件
hbase:
config:
properties:
hbase.zookeeper.quorum : hadoop-master:2181,hadoop-slave01:2181,hadoop-slave02:2181
配置Bean
package com.aimsphm.nuclear.hbase.config;
import lombok.extern.slf4j.Slf4j;
import org.apache.hadoop.hbase.HBaseConfiguration;
import org.apache.hadoop.hbase.client.Connection;
import org.apache.hadoop.hbase.client.ConnectionFactory;
import org.springframework.boot.context.properties.ConfigurationProperties;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;
import org.springframework.util.Assert;
import java.io.IOException;
import java.util.Map;
import java.util.concurrent.ExecutorService;
import java.util.concurrent.Executors;
/**
* @Package: com.aimsphm.nuclear.hbase.config
* @Description: <hbase配置信息>
* @Author: MILLA
* @CreateDate: 2020/3/5 13:16
* @UpdateUser: MILLA
* @UpdateDate: 2020/3/5 13:16
* @UpdateRemark: <>
* @Version: 1.0
*/
@Configuration
@ConfigurationProperties(prefix = HbaseConfig.CONF_PREFIX)
@Slf4j
public class HbaseConfig {
public static final String CONF_PREFIX = "hbase.config";
private Map<String, String> properties;
public Map<String, String> getProperties() {
return properties;
}
public void setProperties(Map<String, String> properties) {
this.properties = properties;
}
@Bean
public Connection connection() throws IOException {
org.apache.hadoop.conf.Configuration conf = HBaseConfiguration.create();
Assert.isTrue(properties != null && !properties.isEmpty(), "Hbase config can not be null");
for (Map.Entry<String, String> confEntry : properties.entrySet()) {
conf.set(confEntry.getKey(), confEntry.getValue());
}
System.setProperty("zookeeper.sasl.client", "false");//消除sasl認證
ExecutorService executor = Executors.newScheduledThreadPool(20);
Connection connection = ConnectionFactory.createConnection(conf, executor);
return connection;
}
}
HbaseUtils工具類
package com.aimsphm.nuclear.hbase.utill;
import com.google.common.collect.Lists;
import com.google.common.collect.Maps;
import com.google.common.collect.Sets;
import lombok.extern.slf4j.Slf4j;
import org.apache.hadoop.hbase.Cell;
import org.apache.hadoop.hbase.CellUtil;
import org.apache.hadoop.hbase.HTableDescriptor;
import org.apache.hadoop.hbase.TableName;
import org.apache.hadoop.hbase.client.*;
import org.apache.hadoop.hbase.client.coprocessor.AggregationClient;
import org.apache.hadoop.hbase.client.coprocessor.LongColumnInterpreter;
import org.apache.hadoop.hbase.filter.CompareFilter;
import org.apache.hadoop.hbase.filter.RowFilter;
import org.apache.hadoop.hbase.filter.SubstringComparator;
import org.apache.hadoop.hbase.io.compress.Compression;
import org.apache.hadoop.hbase.util.Bytes;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.stereotype.Component;
import org.springframework.util.Assert;
import org.springframework.util.StopWatch;
import org.springframework.util.StringUtils;
import java.io.IOException;
import java.math.BigDecimal;
import java.nio.ByteBuffer;
import java.util.*;
import static org.apache.hadoop.hbase.CompareOperator.EQUAL;
/**
* @Package: com.aimsphm.nuclear.hbase.utill
* @Description: <Hbase操作工具類>
* @Author: MILLA
* @CreateDate: 2020/3/5 13:40
* @UpdateUser: MILLA
* @UpdateDate: 2020/3/5 13:40
* @UpdateRemark: <>
* @Version: 1.0
*/
@Slf4j
@Component
public class HbaseUtils {
@Autowired
private Connection connection;
/**
* 創建表
*
* @param tableName 表名
* @param families 列族
*/
public void createTable(String tableName, List<String> families, Compression.Algorithm type) throws IOException {
TableName name = TableName.valueOf(tableName);
try (Admin admin = connection.getAdmin()) {
Assert.isTrue(!admin.tableExists(name), "This table already exists");
TableDescriptorBuilder desc = TableDescriptorBuilder.newBuilder(name);
for (String cf : families) {
ColumnFamilyDescriptorBuilder builder = ColumnFamilyDescriptorBuilder.newBuilder(cf.getBytes());
if (type != null) {
builder.setCompressionType(type);
}
builder.setMaxVersions(1);//指定最大版本1,值會被覆蓋
desc.setColumnFamily(builder.build());
}
admin.createTable(desc.build());
} catch (IOException e) {
throw e;
}
}
public void addFamily2Table(String tableName, List<String> families, Compression.Algorithm type) throws IOException {
TableName table = TableName.valueOf(tableName);
try (Admin admin = connection.getAdmin()) {
Assert.isTrue(admin.tableExists(table), "This table not exists");
for (String cf : families) {
ColumnFamilyDescriptorBuilder builder = ColumnFamilyDescriptorBuilder.newBuilder(cf.getBytes());
if (type != null) {
builder.setCompressionType(type);
}
builder.setMaxVersions(1);//指定最大版本1,值會被覆蓋(當時間戳大於等於已存數據的時間戳時會覆蓋原來的值)
admin.addColumnFamily(table, builder.build());
}
} catch (IOException e) {
throw e;
}
}
/**
* 刪除表操作
*
* @param tableName
*/
public void deleteTable(String tableName) throws IOException {
TableName name = TableName.valueOf(tableName);
try (Admin admin = connection.getAdmin()) {
Assert.isTrue(admin.tableExists(name), "This table not exists");
admin.disableTable(name);
admin.deleteTable(name);
} catch (IOException e) {
throw e;
}
}
/**
* 插入記錄(單行單列族-單列單值-指定時間戳)
*
* @param tableName 表名
* @param rowKey 行名
* @param family 列族名
* @param column 列名
* @param value 值
* @param timestamp 值對應的時間戳
*/
public void insertDouble(String tableName, String rowKey, String family, String column, Double value, Long timestamp) throws IOException {
insertObject(tableName, rowKey, family, column, value, timestamp);
}
/**
* 插入記錄(單行單列族-單列單值-不指定時間戳)
*
* @param tableName 表名
* @param rowKey 行名
* @param family 列族名
* @param column 列名
* @param value 值
*/
public void insertDouble(String tableName, String rowKey, String family, String column, Double value) throws IOException {
insertObject(tableName, rowKey, family, column, value, null);
}
/**
* 插入記錄(單行單列族-單列單值)
*
* @param tableName 表名
* @param rowKey 行名
* @param family 列族名
* @param column 列名
* @param value 值
* @param timestamp 值對應的時間戳
*/
public void insertObject(String tableName, String rowKey, String family, String column, Object value, Long timestamp) throws IOException {
TableName name = TableName.valueOf(tableName);
try (Table table = connection.getTable(name)) {
Put put = new Put(Bytes.toBytes(rowKey));
if (timestamp != null) {
put.setTimestamp(timestamp);
}
put.addColumn(Bytes.toBytes(family), Bytes.toBytes(column), getBytes(value));
table.put(put);
} catch (IOException e) {
throw e;
}
}
/**
* 插入記錄(單行單列族-多列多值)
*
* @param tableName 表名
* @param rowKey 行名
* @param family 列族名
* @param columns 列名(數組)
* @param values 值(數組)(且需要和列一一對應)
*/
public void insertDoubles(String tableName, String rowKey, String family, List<String> columns, List<Double> values) throws IOException {
TableName name = TableName.valueOf(tableName);
try (Table table = connection.getTable(name)) {
Put put = new Put(Bytes.toBytes(rowKey));
for (int i = 0; i < columns.size(); i++) {
put.addColumn(Bytes.toBytes(family), Bytes.toBytes(columns.get(i)), Bytes.toBytes(values.get(i)));
table.put(put);
}
} catch (IOException e) {
throw e;
}
}
/**
* 查找一行記錄
*
* @param tableName 表名
* @param rowKey 行名
*/
public List<Map<String, Object>> selectData(String tableName, String rowKey) throws IOException {
return this.selectData(tableName, rowKey, null);
}
/**
* 查找一行記錄
*
* @param tableName 表名
* @param rowKey 行名
* @param family 列族
*/
public List<Map<String, Object>> selectData(String tableName, String rowKey, String family) throws IOException {
return this.selectData(tableName, rowKey, family, null);
}
/**
* 查找一行記錄
*
* @param tableName 表名
* @param rowKey 行名
* @param family 列族
* @param qualifier 列名
*/
public List<Map<String, Object>> selectData(String tableName, String rowKey, String family, String qualifier) throws IOException {
TableName name = TableName.valueOf(tableName);
try (Table table = connection.getTable(name)) {
Get g = new Get(rowKey.getBytes());
if (!StringUtils.isEmpty(family)) {
g.addFamily(Bytes.toBytes(family));
if (!StringUtils.isEmpty(qualifier)) {
g.addColumn(Bytes.toBytes(family), Bytes.toBytes(qualifier));
}
}
Result rs = table.get(g);
Map<String, Set<String>> familyMap = Maps.newHashMap();
Map<String, List<Object>> qualifierMap = Maps.newHashMap();
assembleCellData(familyMap, qualifierMap, rs);
return assembleReturnData(qualifierMap, familyMap.entrySet());
} catch (IOException e) {
throw e;
}
}
/**
* @param tableName 表格名稱
* @param rowKeyStart 開始rowKey
* @param rowKeyEnd 結束rowKey
* @return
* @throws IOException
*/
public List<Map<String, Object>> selectDataList(String tableName, String rowKeyStart, String rowKeyEnd) throws IOException {
return this.selectDataList(tableName, rowKeyStart, rowKeyEnd, null);
}
/**
* @param tableName 表格名稱
* @param rowKeyStart 開始rowKey
* @param rowKeyEnd 結束rowKey
* @param family 列族
* @return
* @throws IOException
*/
public List<Map<String, Object>> selectDataList(String tableName, String rowKeyStart, String rowKeyEnd, String family) throws IOException {
return this.selectDataList(tableName, rowKeyStart, rowKeyEnd, family, null);
}
/**
* @param tableName 表格名稱
* @param rowKeyStart 開始rowKey
* @param rowKeyEnd 結束rowKey
* @param family 列族
* @param qualifier 列
* @return
* @throws IOException
*/
public List<Map<String, Object>> selectDataList(String tableName, String rowKeyStart, String rowKeyEnd, String family, String qualifier) throws IOException {
TableName name = TableName.valueOf(tableName);
try (Table table = connection.getTable(name)) {
Scan scan = new Scan();
boolean isHasFamily = StringUtils.isEmpty(family);
boolean isHasQualifier = StringUtils.isEmpty(qualifier);
if (!isHasFamily) {
scan.addFamily(Bytes.toBytes(family));
if (!isHasQualifier) {
scan.addColumn(Bytes.toBytes(family), Bytes.toBytes(qualifier));
}
}
scan.withStartRow(Bytes.toBytes(rowKeyStart));
scan.withStopRow(Bytes.toBytes(rowKeyEnd));
ResultScanner scanner = table.getScanner(scan);
Map<String, Set<String>> familyMap = Maps.newHashMap();
Map<String, List<Object>> qualifierMap = Maps.newHashMap();
for (Result rs : scanner) {
assembleCellData(familyMap, qualifierMap, rs);
}
return assembleReturnData(qualifierMap, familyMap.entrySet());
} catch (IOException e) {
throw e;
}
}
/**
* @param qualifierMap 列集合
* @param entries 列族集合
* @return 經過轉換的返回值
*/
private List<Map<String, Object>> assembleReturnData(Map<String, List<Object>> qualifierMap, Set<Map.Entry<String, Set<String>>> entries) {
List<Map<String, Object>> families = Lists.newArrayList();
for (Iterator<Map.Entry<String, Set<String>>> it = entries.iterator(); it.hasNext(); ) {
Map.Entry<String, Set<String>> next = it.next();
String key = next.getKey();
Set<String> value = next.getValue();
Map<String, Object> familyData = Maps.newHashMap();
List<Map<String, Object>> qualifiers = Lists.newArrayList();
familyData.put("family", key);
familyData.put("qualifiers", qualifiers);
families.add(familyData);
for (String cell : value) {
Map<String, Object> qualifierData = Maps.newHashMap();
qualifierData.put("qualifier", cell);
qualifierData.put("times", qualifierMap.get(cell + "Times"));
qualifierData.put("values", qualifierMap.get(cell + "Values"));
qualifiers.add(qualifierData);
}
}
return families;
}
/**
* 組裝數據
*
* @param familyMap 列族集合
* @param qualifierMap 列集合
* @param rs
*/
private void assembleCellData(Map<String, Set<String>> familyMap, Map<String, List<Object>> qualifierMap, Result rs) {
if (rs.size() == 0) {
return;
}
for (Cell cell : rs.listCells()) {
String family = Bytes.toString(CellUtil.cloneFamily(cell));
String qualifier = Bytes.toString(CellUtil.cloneQualifier(cell));
double value = Bytes.toDouble(CellUtil.cloneValue(cell));
long timestamp = cell.getTimestamp();
if (!qualifierMap.containsKey(qualifier + "Times")) {
qualifierMap.put(qualifier + "Times", Lists.newArrayList());
qualifierMap.put(qualifier + "Values", Lists.newArrayList());
}
if (!familyMap.containsKey(family)) {
familyMap.put(family, Sets.newHashSet());
}
familyMap.get(family).add(qualifier);
qualifierMap.get(qualifier + "Times").add(timestamp);
qualifierMap.get(qualifier + "Values").add(value);
}
}
/**
* 刪除一行記錄
*
* @param tableName 表名
* @param rowKey 行名
*/
public void deleteRow(String tableName, String rowKey) throws IOException {
TableName name = TableName.valueOf(tableName);
try (Table table = connection.getTable(name)) {
Delete d = new Delete(rowKey.getBytes());
table.delete(d);
} catch (IOException e) {
throw e;
}
}
/**
* 刪除單行單列族記錄
*
* @param tableName 表名
* @param rowKey 行名
* @param family 列族名
*/
public void deleteFamily(String tableName, String rowKey, String family) throws IOException {
TableName name = TableName.valueOf(tableName);
try (Table table = connection.getTable(name)) {
Delete d = new Delete(rowKey.getBytes()).addFamily(Bytes.toBytes(family));
table.delete(d);
} catch (IOException e) {
throw e;
}
}
/**
* 刪除單行單列族單列記錄
*
* @param tableName 表名
* @param rowKey 行名
* @param family 列族名
* @param qualifier 列名
*/
public void deleteColumn(String tableName, String rowKey, String family, String qualifier) throws IOException {
TableName name = TableName.valueOf(tableName);
try (Table table = connection.getTable(name)) {
Delete d = new Delete(rowKey.getBytes()).addColumn(Bytes.toBytes(family), Bytes.toBytes(qualifier));
table.delete(d);
} catch (IOException e) {
throw e;
}
}
/**
* 查找單行單列族單列記錄
*
* @param tableName 表名
* @param rowKey 行名
* @param family 列族名
* @param column 列名
* @return
*/
public Object selectValue(String tableName, String rowKey, String family, String column) throws IOException {
TableName name = TableName.valueOf(tableName);
Table table = connection.getTable(name);
Get g = new Get(rowKey.getBytes());
g.addColumn(Bytes.toBytes(family), Bytes.toBytes(column));
Result rs = table.get(g);
return getObject(rs.value(), Double.class);
}
/**
* 根據rowKey關鍵字查詢報告記錄
*
* @param tableName
* @param rowKeyword
* @return
*/
public List scanReportDataByRowKeyword(String tableName, String rowKeyword) throws IOException {
List list = new ArrayList<>();
Table table = connection.getTable(TableName.valueOf(tableName));
Scan scan = new Scan();
//添加行鍵過濾器,根據關鍵字匹配
RowFilter rowFilter = new RowFilter(CompareFilter.CompareOp.EQUAL, new SubstringComparator(rowKeyword));
scan.setFilter(rowFilter);
ResultScanner scanner = table.getScanner(scan);
try {
for (Result result : scanner) {
//TODO 此處根據業務來自定義實現
list.add(null);
}
} finally {
if (scanner != null) {
scanner.close();
}
}
return list;
}
/**
* 根據rowKey關鍵字和時間戳範圍查詢報告記錄
*
* @param tableName
* @param rowKeyword
* @return
*/
public List scanReportDataByRowKeywordTimestamp(String tableName, String rowKeyword, Long minStamp, Long maxStamp) throws IOException {
ArrayList list = new ArrayList<>();
Table table = connection.getTable(TableName.valueOf(tableName));
Scan scan = new Scan();
//添加scan的時間範圍
scan.setTimeRange(minStamp, maxStamp);
RowFilter rowFilter = new RowFilter(EQUAL, new SubstringComparator(rowKeyword));
scan.setFilter(rowFilter);
ResultScanner scanner = table.getScanner(scan);
try {
for (Result result : scanner) {
//TODO 此處根據業務來自定義實現
list.add(null);
}
} finally {
if (scanner != null) {
scanner.close();
}
}
return list;
}
/**
* 利用協處理器進行全表count統計
*
* @param tableName
*/
public Long countRowsWithCoprocessor(String tableName) throws Throwable {
TableName name = TableName.valueOf(tableName);
Admin admin = connection.getAdmin();
HTableDescriptor descriptor = admin.getTableDescriptor(name);
String coprocessorClass = "org.apache.hadoop.hbase.coprocessor.AggregateImplementation";
if (!descriptor.hasCoprocessor(coprocessorClass)) {
admin.disableTable(name);
descriptor.addCoprocessor(coprocessorClass);
admin.modifyTable(name, descriptor);
admin.enableTable(name);
}
//計時
StopWatch stopWatch = new StopWatch();
stopWatch.start();
Scan scan = new Scan();
//Hbase服務端需要開啓該服務
AggregationClient aggregationClient = new AggregationClient(connection.getConfiguration());
Long count = aggregationClient.rowCount(name, new LongColumnInterpreter(), scan);
stopWatch.stop();
log.debug("RowCount:{} ,全表count統計耗時:{}", count, stopWatch.getTotalTimeMillis());
return count;
}
/**
* 根據不同的類型獲取不同的byte數組
*
* @param object
* @return
*/
private static byte[] getBytes(Object object) {
Class className = object.getClass();
if (className.equals(java.lang.Boolean.class)) {
return Bytes.toBytes((Boolean) object);
}
if (className.equals(java.lang.Byte.class)) {
return Bytes.toBytes((Byte) object);
}
if (className.equals(java.lang.Short.class)) {
return Bytes.toBytes((Short) object);
}
if (className.equals(java.lang.Character.class)) {
return Bytes.toBytes((Character) object);
}
if (className.equals(java.lang.Integer.class)) {
return Bytes.toBytes((Integer) object);
}
if (className.equals(java.lang.Long.class)) {
return Bytes.toBytes((Long) object);
}
if (className.equals(java.lang.Float.class)) {
return Bytes.toBytes((Float) object);
}
if (className.equals(java.lang.Double.class)) {
return Bytes.toBytes((Double) object);
}
if (className.equals(java.lang.String.class)) {
return Bytes.toBytes((String) object);
}
if (className.equals(java.nio.ByteBuffer.class)) {
return Bytes.toBytes((ByteBuffer) object);
}
if (className.equals(java.math.BigDecimal.class)) {
return Bytes.toBytes((BigDecimal) object);
}
return ByteUtil.toBytes(object);
}
/**
* 根據不同的byte數組獲取不同的對象
*
* @param bytes
* @return
*/
private static Object getObject(byte[] bytes, Class className) {
if (className.equals(java.lang.Boolean.class)) {
return Bytes.toBoolean(bytes);
}
if (className.equals(java.lang.Byte.class)) {
return Bytes.toShort(bytes);
}
if (className.equals(java.lang.Short.class)) {
return Bytes.toShort(bytes);
}
if (className.equals(java.lang.Character.class)) {
return Bytes.toString(bytes);
}
if (className.equals(java.lang.Integer.class)) {
return Bytes.toInt(bytes);
}
if (className.equals(java.lang.Long.class)) {
return Bytes.toLong(bytes);
}
if (className.equals(java.lang.Float.class)) {
return Bytes.toFloat(bytes);
}
if (className.equals(java.lang.Double.class)) {
return Bytes.toDouble(bytes);
}
if (className.equals(java.lang.String.class)) {
return Bytes.toString(bytes);
}
if (className.equals(java.nio.ByteBuffer.class)) {
return ByteBuffer.wrap(bytes);
}
if (className.equals(java.math.BigDecimal.class)) {
return Bytes.toBigDecimal(bytes);
}
return ByteUtil.toObject(bytes);
}
}
ByteUtil工具類
package com.aimsphm.nuclear.hbase.utill;
import java.io.*;
/**
* @Package: com.aimsphm.nuclear.hbase.utill
* @Description: <>
* @Author: MILLA
* @CreateDate: 2020/3/6 11:53
* @UpdateUser: MILLA
* @UpdateDate: 2020/3/6 11:53
* @UpdateRemark: <>
* @Version: 1.0
*/
public final class ByteUtil {
/**
* 對象轉數組
*
* @param obj 對象
* @return
*/
public static byte[] toBytes(Object obj) {
byte[] bytes = null;
ByteArrayOutputStream bos = new ByteArrayOutputStream();
try {
ObjectOutputStream oos = new ObjectOutputStream(bos);
oos.writeObject(obj);
oos.flush();
bytes = bos.toByteArray();
oos.close();
bos.close();
} catch (IOException ex) {
ex.printStackTrace();
}
return bytes;
}
/**
* 數組轉對象
*
* @param bytes 字節數據
* @return
*/
public static Object toObject(byte[] bytes) {
Object obj = null;
try {
ByteArrayInputStream bis = new ByteArrayInputStream(bytes);
ObjectInputStream ois = new ObjectInputStream(bis);
obj = ois.readObject();
ois.close();
bis.close();
} catch (IOException ex) {
ex.printStackTrace();
} catch (ClassNotFoundException ex) {
ex.printStackTrace();
}
return obj;
}
}
備註
壓縮算法、聚合需要在服務端進行插件的配置 Collection對象是單例的,但是在admin中獲取對象的時候使用的是連接池技術
PS:如果將maxVersion設置爲1,那麼相同數據在時間戳大於等於原數據時間戳時,數據會被覆蓋