Table of Contents
Java 版的
Hbase 工具类，配置 zookeeper 的地址
public class HbaseUtil {
public static Configuration getHbaseConfig() {
Configuration config = HBaseConfiguration.create();
// 显式配置 hbase zookeeper 地址
config.set("hbase.zookeeper.quorum", "192.168.78.135");
// config.set("hbase.table.sanity.checks", "false");
// hdfs 配置和 hbase 配置 (hbase-site.xml, core-site.xml)
// config.addResource(new Path(CommUtils.getBasicPath() + "/config/hbase-site.xml"));
// config.addResource(new Path(CommUtils.getBasicPath() + "/config/core-site.xml"));
return config;
}
}建表操作
private static void createSchemaTables(Admin admin) throws IOException {
// 设置 table name
HTableDescriptor table = new HTableDescriptor(TableName.valueOf(TABLE_NAME));
// 设置 列族，压缩
table.addFamily(new HColumnDescriptor(CF_DEFAULT)
.setCompressionType(Compression.Algorithm.NONE));
// 预设 region 数量 和 region 算法
table.setRegionReplication(2);
table.setRegionSplitPolicyClassName(RegionSplitter.UniformSplit.class.getName());
System.out.print("Creating table Start");
// 如果表存在，先 disable 表，在 delete 表
if (admin.tableExists(table.getTableName())) {
admin.disableTable(table.getTableName());
admin.deleteTable(table.getTableName());
}
// 实际的创建表函数
admin.createTable(table);
System.out.println("Create table Done");
}修改表结构
private static void modifySchema(Admin admin) throws IOException {
TableName tableName = TableName.valueOf(TABLE_NAME);
// 修改表之前，判断表是否存在，防止报异常
if (!admin.tableExists(tableName)) {
System.out.println("Table does not exist.");
System.exit(-1);
}
// 创建表对象
HTableDescriptor table = admin.getTableDescriptor(tableName);
// 创建列族对象
HColumnDescriptor newColumn = new HColumnDescriptor("NEWCF");
// 设置压缩格式
newColumn.setCompactionCompressionType(Compression.Algorithm.GZ);
// 执行最大版本号
newColumn.setMaxVersions(HConstants.ALL_VERSIONS);
// 新增表列族
admin.addColumn(tableName, newColumn);
// 修改存在 column family 的压缩格式和版本
HColumnDescriptor existingColumn = new HColumnDescriptor(CF_DEFAULT);
existingColumn.setCompactionCompressionType(Compression.Algorithm.GZ);
existingColumn.setMaxVersions(HConstants.ALL_VERSIONS);
table.modifyFamily(existingColumn);
admin.modifyTable(tableName, table);
// Disable an existing table
admin.disableTable(tableName);
// Delete an existing column family
admin.deleteColumn(tableName, CF_DEFAULT.getBytes("UTF-8"));
// Delete a table (Need to be disabled first)
admin.deleteTable(tableName);
}列出所有表
private static void listTable(Admin admin) throws IOException {
TableName[] tableNames = admin.listTableNames();
for (TableName tableName : tableNames) {
System.err.println("tableName is : " + tableName);
}
}删表
private static void dropTable(Admin admin, String tableName) throws IOException {
TableName table = TableName.valueOf(tableName);
admin.disableTable(table);
System.err.println("disable table is success");
admin.deleteTable(table);
System.err.println("drop table is success");
}写表
private static void insertData(HTable table) throws IOException {
// rowKey 前缀
String rowKey = "rowKey";
// 列族名称
String family = "f1";
// 造一批假数据
for (int i = 0; i < 50; i++) {
Put put = new Put(Bytes.toBytes(rowKey + i));
for (int j = 0; j < 10; j++) {
put.add(family.getBytes(), Bytes.toBytes(rowKey + i + "-key" + j), Bytes.toBytes(rowKey + i + "-value" + j));
}
table.put(put);
}
}根据 rowKey 删数据
private static void deleteData(HTable table) throws IOException {
Delete delete = new Delete("rowKey1".getBytes());
// 指定 列族 名称，删除整个列族
delete.addFamily("f1".getBytes());
// 指定列族下的一个列，删除列
delete.addColumn("f1".getBytes(), "key1".getBytes());
// 什么也不指定，则删除整个 rowKey 下的数
table.delete(delete);
}查表，所有数据
private static void scanData(HTable table) throws IOException {
// 初始化 startRow 和 endRow
Scan scan = new Scan("123".getBytes(), "rowKey11".getBytes());
ResultScanner scanner = table.getScanner(scan);
for (Result result : scanner) {
// 获取 rowKey
String rowKey = new String(result.getRow());
// 获取 key
NavigableMap<byte[], byte[]> familyMap = result.getFamilyMap("f1".getBytes());
for (Map.Entry<byte[], byte[]> entry : familyMap.entrySet()) {
String key = new String(entry.getKey());
String value = new String(entry.getValue());
// 打印结果
System.err.println("rowKey is : " + rowKey + ", key is : " + key + ", value is : " + value);
}
}
}根据 rowKey 查数据
private static void selectDataByRowKey(HTable table) throws IOException {
Get get = new Get("rowKey1".getBytes());
// 获取单个 rowKey 下整个结果
Result result = table.get(get);
System.err.println("result is : " + result.toString());
// 根据列族，获取 key
NavigableMap<byte[], byte[]> familyMap = result.getFamilyMap("f1".getBytes());
for (Map.Entry<byte[], byte[]> entry : familyMap.entrySet()) {
System.err.println("key is : " + new String(entry.getKey()) + ", value is : " + new String(entry.getValue()));
}
// 获取 value
byte[] value = result.getValue("f1".getBytes(), "key1".getBytes());
System.err.println("value is : " + new String(value));
}Scala 版的
Hbase 工具类
object HbaseUtil {
// 已过时
// val conf = new HBaseConfiguration()
val conf = HBaseConfiguration.create()
conf.set("hbase.zookeeper.quorum", "192.168.78.135")
val connection = ConnectionFactory.createConnection(conf)
}列出所有表
def listTables(admin: Admin) = {
val names = admin.listTableNames()
names.foreach(println(_))
}建表
def createTable(admin: Admin) = {
if (admin.tableExists(this.tb)) {
println("表已存在，请换个表名重新建表")
} else {
val descriptor = new HTableDescriptor
descriptor.setName(TableName.valueOf(tableName))
descriptor.addFamily(new HColumnDescriptor("f1"))
admin.createTable(descriptor)
println("create table done")
}
}查看表结构
def describeTable(admin: Admin) = {
val descriptor = admin.getTableDescriptor(TableName.valueOf(tableName))
val families = descriptor.getFamilies
println(families)
}删表
def dropTable(admin: Admin) = {
if (admin.tableExists(this.tb)) {
admin.disableTable(this.tb)
admin.deleteTable(this.tb)
} else
println("表不存在，请确认表名")
}写数据
def insertData(tableName: String): Unit = {
// 新建表对象
val table = new HTable(conf, tableName)
// 造假数据开始
for (a <- 51 to 150) {
// 创建 put 对象，用来存放数据，构造方法里放 rowKey 的值
val put = new Put(("rowKey" + a).getBytes())
for (b <- 1 to 10) {
val key = ("rowKey" + a + "-key" + b).getBytes()
val value = ("rowKey" + a + "-value" + b).getBytes()
// addColumn 方法参数：列族，key，value
put.addColumn("f1".getBytes(), key, value)
}
// put 数据
table.put(put)
}
}扫描整个表
def scanData(tableName: String): Unit = {
// 创建 scan 对象
val scan = new Scan
// 创建表对象，构造方法：配置对象，表名
val table = new HTable(conf, tableName)
// 获取 scanner 对象
val scanner: ResultScanner = table.getScanner(scan)
// 从 scanner 对象中取所有结果
val result = scanner.next()
// 遍历全部结果
while (result != null) {
// 获取 rowKey
val row = new String(result.getRow)
// 获取 rowKey 下所有数据，返回的结构是 NavigableMap<byte[], byte[]>
val map = result.getFamilyMap("f1".getBytes)
val iter = map.entrySet().iterator()
// 开始遍历 map 结果数据
while (iter.hasNext) {
val entry = iter.next()
val key = new String(entry.getKey)
val value = new String(entry.getValue)
println("rowKey is : " + row + ", key is : " + key + ", value is : " + value)
}
}
}根据 rowKey 查数据
def scanData(tableName: String): Unit = {
// 创建 scan 对象
val scan = new Scan
// 创建表对象，构造方法：配置对象，表名
val table = new HTable(conf, tableName)
// 获取 scanner 对象
val scanner: ResultScanner = table.getScanner(scan)
// 从 scanner 对象中取所有结果
val result = scanner.next()
// 遍历全部结果
while (result != null) {
// 获取 rowKey
val row = new String(result.getRow)
// 获取 rowKey 下所有数据，返回的结构是 NavigableMap<byte[], byte[]>
val map = result.getFamilyMap("f1".getBytes)
val iter = map.entrySet().iterator()
// 开始遍历 map 结果数据
while (iter.hasNext) {
val entry = iter.next()
val key = new String(entry.getKey)
val value = new String(entry.getValue)
println("rowKey is : " + row + ", key is : " + key + ", value is : " + value)
}
}
}根据 rowKey 删数据
def deleteDataByRowKey(tableName: String, row: String): Unit = {
val table = new HTable(conf, tableName)
val delete = new Delete(row.getBytes())
table.delete(delete)
}查询所有记录数 count*
def countData(tableName: String): Unit = {
val table = new HTable(conf, tableName)
val descriptor = table.getTableDescriptor
// 定义协调处理器名称
val aggre = "org.apache.hadoop.hbase.coprocessor.AggregateImplementation"
// 如果表没有该协同处理器，先关表，addCoprocessor，开启表
if (!descriptor.hasCoprocessor(aggre)) {
val tbn = TableName.valueOf(tableName)
admin.disableTable(tbn)
descriptor.addCoprocessor(aggre)
admin.modifyTable(TableName.valueOf(tableName), descriptor)
admin.enableTable(tbn)
}
// 创建处理器客户端
val client = new AggregationClient(conf)
// 计算 count 时间
val watch: StopWatch = new StopWatch()
watch.start()
// 开始 rowCount 操作
val count = client.rowCount(table, new LongColumnInterpreter, new Scan())
watch.stop()
// 打印
println(watch.getTime)
println(tableName + " count si : " + count)
}Scala 操作表代码
class TableSchemaDemo1(val tableName: String) {
val tb = TableName.valueOf(tableName)
def dropTable(admin: Admin) = {
if (admin.tableExists(this.tb)) {
admin.disableTable(this.tb)
admin.deleteTable(this.tb)
} else
println("表不存在，请确认表名")
}
def describeTable(admin: Admin) = {
val descriptor = admin.getTableDescriptor(TableName.valueOf(tableName))
val families = descriptor.getFamilies
println(families)
}
def createTable(admin: Admin) = {
if (admin.tableExists(this.tb)) {
println("表已存在，请换个表名重新建表")
} else {
val descriptor = new HTableDescriptor
descriptor.setName(TableName.valueOf(tableName))
descriptor.addFamily(new HColumnDescriptor("f1"))
admin.createTable(descriptor)
println("create table done")
}
}
def listTables(admin: Admin) = {
val names = admin.listTableNames()
names.foreach(println(_))
}
def listTables(): Unit = {
}
override def toString = s"TableSchemaDemo1($tableName)"
}
object TableSchemaDemo1 {
def main(args: Array[String]): Unit = {
// 获取 hbase 配置
val conf = HbaseUtil.conf
// 获取连接
val connection = HbaseUtil.connection
// 获取管理员
val admin = connection.getAdmin()
val demo = new TableSchemaDemo1("tbs")
demo.listTables(admin)
demo.createTable(admin)
demo.describeTable(admin)
// demo.dropTable(admin)
println(demo.toString)
}
}
280
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
