Apache
/
hbase
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

HBASE-1762 Remove concept of ZooKeeper from HConnection interface 

git-svn-id: https://svn.apache.org/repos/asf/hbase/trunk@1480449 13f79535-47bb-0310-9956-ffa450edef68
This commit is contained in:
Michael Stack 2013-05-08 20:49:30 +00:00
parent 6ad28427b1
commit 4dc244ce5a
7 changed files with 67 additions and 63 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

21
hbase-client/src/main/java/org/apache/hadoop/hbase/client/HConnection.java
View File

@ -38,7 +38,6 @@ import org.apache.hadoop.hbase.protobuf.generated.AdminProtos.AdminService;
import org.apache.hadoop.hbase.protobuf.generated.ClientProtos.ClientService;
import org.apache.hadoop.hbase.protobuf.generated.MasterAdminProtos.MasterAdminService;
import org.apache.hadoop.hbase.protobuf.generated.MasterMonitorProtos.MasterMonitorService;
import org.apache.hadoop.hbase.zookeeper.ZooKeeperWatcher;
/**
* A cluster connection. Knows how to find the master, locate regions out on the cluster,
@ -48,8 +47,7 @@ import org.apache.hadoop.hbase.zookeeper.ZooKeeperWatcher;
* connections at a lower level.
*
* <p>HConnections are used by {@link HTable} mostly but also by
* {@link HBaseAdmin}, {@link CatalogTracker},
* and {@link ZooKeeperWatcher}. HConnection instances can be shared. Sharing
* {@link HBaseAdmin}, and {@link CatalogTracker}. HConnection instances can be shared. Sharing
* is usually what you want because rather than each HConnection instance
* having to do its own discovery of regions out on the cluster, instead, all
* clients get to share the one cache of locations. {@link HConnectionManager} does the
@ -66,17 +64,6 @@ public interface HConnection extends Abortable, Closeable {
*/
public Configuration getConfiguration();
/**
* Retrieve ZooKeeperWatcher used by this connection.
* @return ZooKeeperWatcher handle being used by the connection.
* @throws IOException if a remote or network exception occurs
* @deprecated Removed because it was a mistake exposing zookeeper in this
* interface (ZooKeeper is an implementation detail).
* Deprecated in HBase 0.94
*/
@Deprecated
public ZooKeeperWatcher getZooKeeperWatcher() throws IOException;
/** @return - true if the master server is running */
public boolean isMasterRunning()
throws MasterNotRunningException, ZooKeeperConnectionException;
@ -104,9 +91,9 @@ public interface HConnection extends Abortable, Closeable {
* @throws IOException if a remote or network exception occurs
*/
public boolean isTableAvailable(byte[] tableName) throws IOException;
/**
* Use this api to check if the table has been created with the specified number of
* Use this api to check if the table has been created with the specified number of
* splitkeys which was used while creating the given table.
* Note : If this api is used after a table's region gets splitted, the api may return
* false.
@ -202,7 +189,7 @@ public interface HConnection extends Abortable, Closeable {
*/
public List<HRegionLocation> locateRegions(final byte[] tableName)
throws IOException;
/**
* Gets the locations of all regions in the specified table, <i>tableName</i>.
* @param tableName table to get regions of

18
hbase-client/src/main/java/org/apache/hadoop/hbase/client/HConnectionManager.java
View File

@ -1546,24 +1546,6 @@ public class HConnectionManager {
return serviceName + "@" + rsHostnamePort;
}
@Override
@Deprecated
public ZooKeeperWatcher getZooKeeperWatcher()
throws ZooKeeperConnectionException {
canCloseZKW = false;
try {
return getKeepAliveZooKeeperWatcher();
} catch (ZooKeeperConnectionException e){
throw e;
}catch (IOException e) {
// Encapsulate exception to keep interface
throw new ZooKeeperConnectionException(
"Can't create a zookeeper connection", e);
}
}
private ZooKeeperKeepAliveConnection keepAliveZookeeper;
private int keepAliveZookeeperUserCount;
private boolean canCloseZKW = true;

4
hbase-client/src/main/java/org/apache/hadoop/hbase/client/ZooKeeperKeepAliveConnection.java
View File

@ -45,7 +45,9 @@ class ZooKeeperKeepAliveConnection extends ZooKeeperWatcher{
@Override
public void close() {
((HConnectionManager.HConnectionImplementation)abortable).releaseZooKeeperWatcher(this);
if (this.abortable != null) {
((HConnectionManager.HConnectionImplementation)abortable).releaseZooKeeperWatcher(this);
}
}
void internalClose(){

17
hbase-client/src/main/java/org/apache/hadoop/hbase/zookeeper/ZooKeeperWatcher.java
View File

@ -68,6 +68,8 @@ public class ZooKeeperWatcher implements Watcher, Abortable, Closeable {
// abortable in case of zk failure
protected Abortable abortable;
// Used if abortable is null
private boolean aborted = false;
// listeners to be notified
private final List<ZooKeeperListener> listeners =
@ -128,10 +130,15 @@ public class ZooKeeperWatcher implements Watcher, Abortable, Closeable {
Abortable abortable) throws ZooKeeperConnectionException, IOException {
this(conf, identifier, abortable, false);
}
/**
* Instantiate a ZooKeeper connection and watcher.
* @param identifier string that is passed to RecoverableZookeeper to be used as
* identifier for this instance. Use null for default.
* @param conf
* @param abortable Can be null if there is on error there is no host to abort: e.g. client
* context.
* @param canCreateBaseZNode
* @throws IOException
* @throws ZooKeeperConnectionException
*/
@ -361,8 +368,9 @@ public class ZooKeeperWatcher implements Watcher, Abortable, Closeable {
"ZooKeeper, aborting");
// TODO: One thought is to add call to ZooKeeperListener so say,
// ZooKeeperNodeTracker can zero out its data values.
if (this.abortable != null) this.abortable.abort(msg,
new KeeperException.SessionExpiredException());
if (this.abortable != null) {
this.abortable.abort(msg, new KeeperException.SessionExpiredException());
}
break;
case ConnectedReadOnly:
@ -444,12 +452,13 @@ public class ZooKeeperWatcher implements Watcher, Abortable, Closeable {
@Override
public void abort(String why, Throwable e) {
this.abortable.abort(why, e);
if (this.abortable != null) this.abortable.abort(why, e);
else this.aborted = true;
}
@Override
public boolean isAborted() {
return this.abortable.isAborted();
return this.abortable == null? this.aborted: this.abortable.isAborted();
}
/**

19
hbase-server/src/main/java/org/apache/hadoop/hbase/mapreduce/Import.java
View File

@ -117,7 +117,7 @@ public class Import {
extends TableMapper<ImmutableBytesWritable, Mutation> {
private Map<byte[], byte[]> cfRenameMap;
private UUID clusterId;
/**
* @param row The current table row key.
* @param value The columns.
@ -175,20 +175,27 @@ public class Import {
Configuration conf = context.getConfiguration();
cfRenameMap = createCfRenameMap(conf);
filter = instantiateFilter(conf);
// TODO: This is kind of ugly doing setup of ZKW just to read the clusterid.
ReplicationZookeeper zkHelper = null;
ZooKeeperWatcher zkw = null;
try {
HConnection connection = HConnectionManager.getConnection(conf);
ZooKeeperWatcher zkw = connection.getZooKeeperWatcher();
ReplicationZookeeper zkHelper = new ReplicationZookeeper(connection, conf, zkw);
clusterId = zkHelper.getUUIDForCluster(zkw);
zkw = new ZooKeeperWatcher(conf, context.getTaskAttemptID().toString(), null);
zkHelper = new ReplicationZookeeper(connection, conf, zkw);
try {
this.clusterId = zkHelper.getUUIDForCluster(zkw);
} finally {
if (zkHelper != null) zkHelper.close();
}
} catch (ZooKeeperConnectionException e) {
LOG.error("Problem connecting to ZooKeper during task setup", e);
} catch (KeeperException e) {
LOG.error("Problem reading ZooKeeper data during task setup", e);
} catch (IOException e) {
LOG.error("Problem setting up task", e);
} finally {
if (zkw != null) zkw.close();
}
}
}

3
hbase-server/src/main/ruby/hbase/admin.rb
View File

@ -32,7 +32,8 @@ module Hbase
@admin = org.apache.hadoop.hbase.client.HBaseAdmin.new(configuration)
connection = @admin.getConnection()
@conf = configuration
@zk_wrapper = connection.getZooKeeperWatcher()
@zk_wrapper = org.apache.hadoop.hbase.zookeeper.ZooKeeperWatcher.new(configuration,
"admin", nil)
zk = @zk_wrapper.getRecoverableZooKeeper().getZooKeeper()
@zk_main = org.apache.zookeeper.ZooKeeperMain.new(zk)
@formatter = formatter

48
hbase-server/src/test/java/org/apache/hadoop/hbase/regionserver/TestSplitTransactionOnCluster.java
View File

@ -37,6 +37,7 @@ import org.apache.commons.logging.LogFactory;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.hbase.Abortable;
import org.apache.hadoop.hbase.HBaseIOException;
import org.apache.hadoop.hbase.HBaseTestingUtility;
import org.apache.hadoop.hbase.HConstants;
@ -426,6 +427,23 @@ public class TestSplitTransactionOnCluster {
}
}
/**
* Noop Abortable implementation used below in tests.
*/
static class UselessTestAbortable implements Abortable {
boolean aborted = false;
@Override
public void abort(String why, Throwable e) {
LOG.warn("ABORTED (But nothing to abort): why=" + why, e);
aborted = true;
}
@Override
public boolean isAborted() {
return this.aborted;
}
}
/**
* Verifies HBASE-5806. When splitting is partially done and the master goes down
* when the SPLIT node is in either SPLIT or SPLITTING state.
@ -453,6 +471,8 @@ public class TestSplitTransactionOnCluster {
this.admin.setBalancerRunning(false, true);
// Turn off the meta scanner so it don't remove parent on us.
cluster.getMaster().setCatalogJanitorEnabled(false);
ZooKeeperWatcher zkw = new ZooKeeperWatcher(t.getConfiguration(),
"testMasterRestartWhenSplittingIsPartial", new UselessTestAbortable());
try {
// Add a bit of load up into the table so splittable.
TESTING_UTIL.loadTable(t, HConstants.CATALOG_FAMILY);
@ -467,14 +487,11 @@ public class TestSplitTransactionOnCluster {
this.admin.split(hri.getRegionNameAsString());
checkAndGetDaughters(tableName);
// Assert the ephemeral node is up in zk.
String path = ZKAssign.getNodeName(t.getConnection()
.getZooKeeperWatcher(), hri.getEncodedName());
Stat stats = t.getConnection().getZooKeeperWatcher()
.getRecoverableZooKeeper().exists(path, false);
String path = ZKAssign.getNodeName(zkw, hri.getEncodedName());
Stat stats = zkw.getRecoverableZooKeeper().exists(path, false);
LOG.info("EPHEMERAL NODE BEFORE SERVER ABORT, path=" + path + ", stats="
+ stats);
byte[] bytes = ZKAssign.getData(t.getConnection()
.getZooKeeperWatcher(), hri.getEncodedName());
byte[] bytes = ZKAssign.getData(zkw, hri.getEncodedName());
RegionTransition rtd = RegionTransition.parseFrom(bytes);
// State could be SPLIT or SPLITTING.
assertTrue(rtd.getEventType().equals(EventType.RS_ZK_REGION_SPLIT)
@ -498,6 +515,7 @@ public class TestSplitTransactionOnCluster {
admin.setBalancerRunning(true, false);
cluster.getMaster().setCatalogJanitorEnabled(true);
t.close();
zkw.close();
}
}
@ -526,6 +544,8 @@ public class TestSplitTransactionOnCluster {
this.admin.setBalancerRunning(false, true);
// Turn off the meta scanner so it don't remove parent on us.
cluster.getMaster().setCatalogJanitorEnabled(false);
ZooKeeperWatcher zkw = new ZooKeeperWatcher(t.getConfiguration(),
"testMasterRestartAtRegionSplitPendingCatalogJanitor", new UselessTestAbortable());
try {
// Add a bit of load up into the table so splittable.
TESTING_UTIL.loadTable(t, HConstants.CATALOG_FAMILY);
@ -536,22 +556,17 @@ public class TestSplitTransactionOnCluster {
this.admin.split(hri.getRegionNameAsString());
checkAndGetDaughters(tableName);
// Assert the ephemeral node is up in zk.
String path = ZKAssign.getNodeName(t.getConnection()
.getZooKeeperWatcher(), hri.getEncodedName());
Stat stats = t.getConnection().getZooKeeperWatcher()
.getRecoverableZooKeeper().exists(path, false);
String path = ZKAssign.getNodeName(zkw, hri.getEncodedName());
Stat stats = zkw.getRecoverableZooKeeper().exists(path, false);
LOG.info("EPHEMERAL NODE BEFORE SERVER ABORT, path=" + path + ", stats="
+ stats);
String node = ZKAssign.getNodeName(t.getConnection()
.getZooKeeperWatcher(), hri.getEncodedName());
String node = ZKAssign.getNodeName(zkw, hri.getEncodedName());
Stat stat = new Stat();
byte[] data = ZKUtil.getDataNoWatch(t.getConnection()
.getZooKeeperWatcher(), node, stat);
byte[] data = ZKUtil.getDataNoWatch(zkw, node, stat);
// ZKUtil.create
for (int i=0; data != null && i<60; i++) {
Thread.sleep(1000);
data = ZKUtil.getDataNoWatch(t.getConnection().getZooKeeperWatcher(),
node, stat);
data = ZKUtil.getDataNoWatch(zkw, node, stat);
}
assertNull("Waited too long for ZK node to be removed: "+node, data);
@ -571,6 +586,7 @@ public class TestSplitTransactionOnCluster {
this.admin.setBalancerRunning(true, false);
cluster.getMaster().setCatalogJanitorEnabled(true);
t.close();
zkw.close();
}
}