Hivemetastore源码阅读(⼀)
不要问我为什么,因为爱,哈哈哈哈。。。进⼊正题,最近做项⽬顺带学习了下hive metastore的源码,进⾏下知识总结。 hive metastore的整体架构如图:
⼀、组成结构:
如图我们可以看到,hive metastore的组成结构分为客户端服务端,那么下来我们逐⼀进⾏分析: 1、客户端
从代码的⾓度来看:太多了。。我们从⼊⼝HIVE开始看,可以到MetaStoreClient客户端的创建: 1private IMetaStoreClient createMetaStoreClient() throws MetaException {
2
3 HiveMetaHookLoader hookLoader = new HiveMetaHookLoader() {
4 @Override
5public HiveMetaHook getHook(
有天下第一行书6 org.apache.astore.api.Table tbl)
7throws MetaException {
8
9try {
10if (tbl == null) {
11return null;
12 }
13 HiveStorageHandler storageHandler =
14 StorageHandler(conf,
15 Parameters().get(META_TABLE_STORAGE));
16if (storageHandler == null) {
17return null;
18 }
MetaHook();
20 } catch (HiveException ex) {
21 (StringUtils.stringifyException(ex));
22throw new MetaException(
23 "Failed to load storage handler: " + ex.getMessage());
24 }
25 }
26 };
Proxy(conf, hookLoader, metaCallTimeMap,
28 Name());
29 }
我们可以看到,创建MetaStoreClient中,创建了HiveMetaHook,这个Hook的作⽤在于,每次对meta进⾏操作的时候,⽐如createTable的时候,如果建表的存储⽅式不是⽂件,⽐如集成hbase,HiveMetaStoreClient会调⽤hook的接⼝⽅法preCreateTable,进⾏建表前的准备,⽤来判断外部表与内部表,如果中途有失败的话,依旧调⽤hook中的rollbackCreateTable进⾏回滚。
1public void createTable(Table tbl, EnvironmentContext envContext) throws AlreadyExistsException,
2 InvalidObjectException, MetaException, NoSuchObjectException, TException {
3 HiveMetaHook hook = getHook(tbl);
4if (hook != null) {
5 hook.preCreateTable(tbl);
6 }
7boolean success = false;
8try {
9// Subclasses can override this step (for example, for temporary tables)
10 create_table_with_environment_context(tbl, envContext);
11if (hook != null) {
12 hookmitCreateTable(tbl);
13 }
14 success = true;
15 } finally {
16if (!success && (hook != null)) {
17 llbackCreateTable(tbl);
18 }
19 }
20 }
在hbase表不存在的情况下,不能create external table ,会报doesn't exist while the table is declared as an external table,那么需直接创建create table 创建⼀个指向hbase的hive表。 建表语句如下:
CREATE TABLE hbase_table_1(key int, value string) /
STORED BY 'org.apache.hadoop.hive.hbase.HBaseStorageHandler'
WITH SERDEPROPERTIES ("lumns.mapping" = ":key,cf1:val")
TBLPROPERTIES ("hbase.table.name" = "tableName", "hbase.mapred.output.outputtable" = "tableName");
散射粒子
代码请查看HBaseStorageHandler的preCreateTable⽅法,这⾥就不贴出来啦。
随之回归Hive类,Hive类可以说是整个元数据DDL操作的最顶层抽象。HiveMetaStoreClient实现了IMetaStoreClient的接⼝,在创建HiveMetaStoreClient时,会创建于server段HiveMetaStore的链接,并且会通过检查astore.local是否为true,来决定是在本地创建服务端,这⾥为在本地: 1public HiveMetaStoreClient(HiveConf conf, HiveMetaHookLoader hookLoader)
2throws MetaException {
3
4this.hookLoader = hookLoader;
5if (conf == null) {
6 conf = new HiveConf(HiveMetaStoreClient.class);
7 }
9 filterHook = loadFilterHooks();
10
11 String msUri = Var(HiveConf.ConfVars.METASTOREURIS);
12 localMetaStore = HiveConfUtil.isEmbeddedMetaStore(msUri);
13if (localMetaStore) {集韵
14// instantiate the metastore server handler directly instead of connecting
15// through the network
16 client = wRetryingHMSHandler("hive client", conf, true);
17 isConnected = true;
18 snapshotActiveConf();
19return;
20 }
随后,创建server端的HiveMetaStore.HMSHandler,HMSHandler继承⾃IHMSHandler,⽽IHMSHandler⼜继承⾃ThriftHiveMetastore.Iface,在HMSHandler中实现了所有操作的对外⽅法:
public class ThriftHiveMetastore {
/**
* This interface is live.
*/
public interface Iface extends com.facebook.fb303.FacebookService.Iface {
public String getMetaConf(String key) throws MetaException, org.apache.thrift.TException;
public void setMetaConf(String key, String value) throws MetaException, org.apache.thrift.TException;
public void create_database(Database database) throws AlreadyExistsException, InvalidObjectException, MetaException, org.apache.thrift.TException;
public Database get_database(String name) throws NoSuchObjectException, MetaException, org.apache.thrift.TException;
public void drop_database(String name, boolean deleteData, boolean cascade) throws NoSuchObjectException, InvalidOperationException, MetaException, org.apache.thrift.TException; public List<String> get_databases(String pattern) throws MetaException, org.apache.thrift.TException;
public List<String> get_all_databases() throws MetaException, org.apache.thrift.TException;
public void alter_database(String dbname, Database db) throws MetaException, NoSuchObjectException, org.apache.thrift.TException;
public Type get_type(String name) throws MetaException, NoSuchObjectException, org.apache.thrift.TException;
public boolean create_type(Type type) throws AlreadyExistsException, InvalidObjectException, MetaException, org.apache.thrift.TException;
public boolean drop_type(String type) throws MetaException, NoSuchObjectException, org.apache.thrift.TException;
......
在创建HiveMetaStore的init⽅法中,同时创建了三种Listener---
MetaStorePreEventListener,MetaStoreEventListener,MetaStoreEndFunctionListener⽤于对每⼀步事件的监听。
1 initListeners = MetaStoreListeners(
特勤机甲队62 MetaStoreInitListener.class, hiveConf,
3 Var(HiveConf.ConfVars.METASTORE_INIT_HOOKS));
4for (MetaStoreInitListener singleInitListener: initListeners) {
5 MetaStoreInitContext context = new MetaStoreInitContext();
6 Init(context);
7 }
8
9 String alterHandlerName = ("astore.alter.impl",
10 Name());
11alterHandler = (AlterHandler) Class(
12 alterHandlerName), hiveConf);
抚顺育才中学
13 wh = new Warehouse(hiveConf);
14
15synchronized (HMSHandler.class) {
16if (currentUrl == null || !currentUrl.ConnectionURL(hiveConf))) {
17 createDefaultDB();
18 createDefaultRoles();
19 addAdminUsers();
20 currentUrl = ConnectionURL(hiveConf);
21 }
22 }
23
证券监督管理条例
24if (Boolean("abled", false)) {
25try {
26 Metrics.init();
27 } catch (Exception e) {
28// log exception, but ignore inability to start
29 ("error in Metrics init: " + e.getClass().getName() + " "
30 + e.getMessage(), e);
31 }
32 }
33
34preListeners = MetaStoreListeners(MetaStorePreEventListener.class,
35 hiveConf,
36 Var(HiveConf.ConfVars.METASTORE_PRE_EVENT_LISTENERS));
37listeners = MetaStoreListeners(MetaStoreEventListener.class, hiveConf,
38 Var(HiveConf.ConfVars.METASTORE_EVENT_LISTENERS));
39listeners.add(new SessionPropertiesListener(hiveConf));
40 endFunctionListeners = MetaStoreListeners(
41 MetaStoreEndFunctionListener.class, hiveConf,
42 Var(HiveConf.ConfVars.METASTORE_END_FUNCTION_LISTENERS));
同时创建了AlterHandler,它是HiveAlterHandler的接⼝,是将修改表和修改partition的操作抽离了出来单独实现(修改表很复杂的。。)。
1public interface AlterHandler extends Configurable {
2
3/**
4 * handles alter table
5 *
6 * @param msdb
7 * object to get metadata
8 * @param wh
9 * TODO
10 * @param dbname
11 * database of the table being altered
12 * @param name
13 * original name of the table being altered. same as
14 * <i>newTable.tableName</i> if alter op is not a rename.
15 * @param newTable
16 * new table object
17 * @throws InvalidOperationException
18 * thrown if the newTable object is invalid
19 * @throws MetaException
20 * thrown if there is any other error
21*/
22public abstract void alterTable(RawStore msdb, Warehouse wh, String dbname,
23 String name, Table newTable) throws InvalidOperationException,
24 MetaException;
25
26/**
27 * handles alter table, the changes could be cascaded to partitions if applicable
28 *
29 * @param msdb
30 * object to get metadata
31 * @param wh
32 * Hive Warehouse where table data is stored
33 * @param dbname
34 * database of the table being altered
35 * @param name
36 * original name of the table being altered. same as
37 * <i>newTable.tableName</i> if alter op is not a rename.
38 * @param newTable
39 * new table object
40 * @param cascade
41 * if the changes will be cascaded to its partitions if applicable
42 * @throws InvalidOperationException
43 * thrown if the newTable object is invalid
44 * @throws MetaException
45 * thrown if there is any other error
46*/
47public abstract void alterTable(RawStore msdb, Warehouse wh, String dbname,
48 String name, Table newTable, boolean cascade) throws InvalidOperationException,
49 MetaException;
最重要的是RawStore的创建。RawStore不光是定义了⼀套最终的物理操作,使⽤JDO将⼀个对象当作表进⾏存储。ObjectStore中的transaction机制也是通过JDO提供的transaction实现的。当commit失败时,将rollback所有操作。
1 @Override
2public void createDatabase(Database db) throws InvalidObjectException, MetaException {
3boolean commited = false;
4 MDatabase mdb = new MDatabase();
5 mdb.Name().toLowerCase());
6 mdb.LocationUri());
7 mdb.Description());
8 mdb.Parameters());
9 mdb.OwnerName());
10 PrincipalType ownerType = db.getOwnerType();
11 mdb.setOwnerType((null == ownerType ? PrincipalType.USER.name() : ownerType.name()));
12try {
13 openTransaction();
14 pm.makePersistent(mdb);
15 commited = commitTransaction();
16 } finally {
17if (!commited) {
18 rollbackTransaction();
19 }
20 }
21 }
22
23 @SuppressWarnings("nls")
24private MDatabase getMDatabase(String name) throws NoSuchObjectException {
25 MDatabase mdb = null;
26boolean commited = false;
27try {
28 openTransaction();
29 name = alizeIdentifier(name);
30 Query query = pm.newQuery(MDatabase.class, "name == dbname");
31 query.declareParameters("java.lang.String dbname");
32 query.setUnique(true);
33 mdb = (MDatabase) ute(name);
34 pm.retrieve(mdb);
35 commited = commitTransaction();
36 } finally {
37if (!commited) {
38 rollbackTransaction();
39 }
40 }
41if (mdb == null) {
42throw new NoSuchObjectException("There is no database named " + name);
43 }
44return mdb;
45 }
今晚就到这⾥。。。⾃⼰摸索,如有错误,还望指出谢谢~