Apache
/
lucene
mirror of https://github.com/apache/lucene.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

SOLR-3709: Cache the url list created from the ClusterState in CloudSolrServer on each request. 

SOLR-3708: Add hashcode to ClusterState so that structures built based on the ClusterState can be easily cached.	

git-svn-id: https://svn.apache.org/repos/asf/lucene/dev/trunk@1369465 13f79535-47bb-0310-9956-ffa450edef68
This commit is contained in:
Mark Robert Miller 2012-08-04 21:22:45 +00:00
parent 37153b5c34
commit 238c4274b6
5 changed files with 225 additions and 132 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
solr/CHANGES.txt
View File

@ -109,6 +109,15 @@ New Features
* SOLR-3672: SimplePostTool: Improvements for posting files
Support for auto mode, recursive and wildcards (janhoy)
Optimizations
----------------------
* SOLR-3708: Add hashCode to ClusterState so that structures built based on the
ClusterState can be easily cached. (Mark Miller)
* SOLR-3709: Cache the url list created from the ClusterState in CloudSolrServer on each
request. (Mark Miller, yonik)
Bug Fixes
----------------------

6
solr/core/src/test/org/apache/solr/cloud/ClusterStateTest.java
View File

@ -55,7 +55,7 @@ public class ClusterStateTest extends SolrTestCaseJ4 {
ClusterState clusterState = new ClusterState(liveNodes, collectionStates);
byte[] bytes = ZkStateReader.toJSON(clusterState);
ClusterState loadedClusterState = ClusterState.load(bytes, liveNodes);
ClusterState loadedClusterState = ClusterState.load(null, bytes, liveNodes);
assertEquals("Provided liveNodes not used properly", 2, loadedClusterState
.getLiveNodes().size());
@ -63,13 +63,13 @@ public class ClusterStateTest extends SolrTestCaseJ4 {
assertEquals("Poperties not copied properly", zkNodeProps.get("prop1"), loadedClusterState.getSlice("collection1", "shard1").getShards().get("node1").get("prop1"));
assertEquals("Poperties not copied properly", zkNodeProps.get("prop2"), loadedClusterState.getSlice("collection1", "shard1").getShards().get("node1").get("prop2"));
loadedClusterState = ClusterState.load(new byte[0], liveNodes);
loadedClusterState = ClusterState.load(null, new byte[0], liveNodes);
assertEquals("Provided liveNodes not used properly", 2, loadedClusterState
.getLiveNodes().size());
assertEquals("Should not have collections", 0, loadedClusterState.getCollections().size());
loadedClusterState = ClusterState.load((byte[])null, liveNodes);
loadedClusterState = ClusterState.load(null, (byte[])null, liveNodes);
assertEquals("Provided liveNodes not used properly", 2, loadedClusterState
.getLiveNodes().size());

36
solr/solrj/src/java/org/apache/solr/client/solrj/impl/CloudSolrServer.java
View File

@ -61,6 +61,10 @@ public class CloudSolrServer extends SolrServer {
private LBHttpSolrServer lbServer;
private HttpClient myClient;
Random rand = new Random();
// since the state shouldn't change often, should be very cheap reads
private volatile List<String> urlList;
private volatile int lastClusterStateHashCode;
/**
* @param zkHost The client endpoint of the zookeeper quorum containing the cloud state,
* in the form HOST:PORT.
@ -168,22 +172,26 @@ public class CloudSolrServer extends SolrServer {
// or shardAddressVersion (which only changes when the shards change)
// to allow caching.
// build a map of unique nodes
// TODO: allow filtering by group, role, etc
Map<String,ZkNodeProps> nodes = new HashMap<String,ZkNodeProps>();
List<String> urlList = new ArrayList<String>();
for (Slice slice : slices.values()) {
for (ZkNodeProps nodeProps : slice.getShards().values()) {
ZkCoreNodeProps coreNodeProps = new ZkCoreNodeProps(nodeProps);
String node = coreNodeProps.getNodeName();
if (!liveNodes.contains(coreNodeProps.getNodeName())
|| !coreNodeProps.getState().equals(
ZkStateReader.ACTIVE)) continue;
if (nodes.put(node, nodeProps) == null) {
String url = coreNodeProps.getCoreUrl();
urlList.add(url);
if (clusterState.hashCode() != this.lastClusterStateHashCode) {
// build a map of unique nodes
// TODO: allow filtering by group, role, etc
Map<String,ZkNodeProps> nodes = new HashMap<String,ZkNodeProps>();
List<String> urlList = new ArrayList<String>();
for (Slice slice : slices.values()) {
for (ZkNodeProps nodeProps : slice.getShards().values()) {
ZkCoreNodeProps coreNodeProps = new ZkCoreNodeProps(nodeProps);
String node = coreNodeProps.getNodeName();
if (!liveNodes.contains(coreNodeProps.getNodeName())
|| !coreNodeProps.getState().equals(ZkStateReader.ACTIVE)) continue;
if (nodes.put(node, nodeProps) == null) {
String url = coreNodeProps.getCoreUrl();
urlList.add(url);
}
}
}
this.urlList = urlList;
this.lastClusterStateHashCode = clusterState.hashCode();
}
Collections.shuffle(urlList, rand);

289
solr/solrj/src/java/org/apache/solr/common/cloud/ClusterState.java
View File

@ -32,6 +32,7 @@ import org.apache.solr.common.SolrException;
import org.apache.solr.common.SolrException.ErrorCode;
import org.apache.solr.common.cloud.HashPartitioner.Range;
import org.apache.zookeeper.KeeperException;
import org.apache.zookeeper.data.Stat;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
@ -42,25 +43,49 @@ import org.slf4j.LoggerFactory;
public class ClusterState implements JSONWriter.Writable {
private static Logger log = LoggerFactory.getLogger(ClusterState.class);
private final Map<String, Map<String,Slice>> collectionStates; // Map<collectionName, Map<sliceName,Slice>>
private final Set<String> liveNodes;
private Integer zkClusterStateVersion;
private final Map<String, Map<String,Slice>> collectionStates; // Map<collectionName, Map<sliceName,Slice>>
private final Set<String> liveNodes;
private final HashPartitioner hp = new HashPartitioner();
private final Map<String,RangeInfo> rangeInfos = new HashMap<String,RangeInfo>();
private final Map<String,Map<String,ZkNodeProps>> leaders = new HashMap<String,Map<String,ZkNodeProps>>();
public ClusterState(Set<String> liveNodes,
Map<String, Map<String,Slice>> collectionStates) {
this.liveNodes = new HashSet<String>(liveNodes.size());
this.liveNodes.addAll(liveNodes);
this.collectionStates = new HashMap<String, Map<String,Slice>>(collectionStates.size());
this.collectionStates.putAll(collectionStates);
addRangeInfos(collectionStates.keySet());
getShardLeaders();
}
private void getShardLeaders() {
/**
* Use this constr when ClusterState is meant for publication.
*
* hashCode and equals will only depend on liveNodes and not clusterStateVersion.
*
* @param liveNodes
* @param collectionStates
*/
public ClusterState(Set<String> liveNodes,
Map<String, Map<String,Slice>> collectionStates) {
this(null, liveNodes, collectionStates);
}
/**
* Use this constr when ClusterState is meant for consumption.
*
* @param zkClusterStateVersion
* @param liveNodes
* @param collectionStates
*/
public ClusterState(Integer zkClusterStateVersion, Set<String> liveNodes,
Map<String, Map<String,Slice>> collectionStates) {
this.liveNodes = new HashSet<String>(liveNodes.size());
this.liveNodes.addAll(liveNodes);
this.collectionStates = new HashMap<String, Map<String,Slice>>(collectionStates.size());
this.collectionStates.putAll(collectionStates);
addRangeInfos(collectionStates.keySet());
getShardLeaders();
}
private void getShardLeaders() {
Set<Entry<String,Map<String,Slice>>> collections = collectionStates.entrySet();
for (Entry<String,Map<String,Slice>> collection : collections) {
Map<String,Slice> state = collection.getValue();
@ -85,27 +110,27 @@ public class ClusterState implements JSONWriter.Writable {
}
}
/**
* Get properties of a shard leader for specific collection.
*/
public ZkNodeProps getLeader(String collection, String shard) {
Map<String,ZkNodeProps> collectionLeaders = leaders.get(collection);
if (collectionLeaders == null) return null;
return collectionLeaders.get(shard);
}
/**
* Get shard properties or null if shard is not found.
*/
public ZkNodeProps getShardProps(final String collection, final String coreNodeName) {
Map<String, Slice> slices = getSlices(collection);
for(Slice slice: slices.values()) {
if(slice.getShards().get(coreNodeName)!=null) {
return slice.getShards().get(coreNodeName);
}
}
return null;
}
/**
* Get properties of a shard leader for specific collection.
*/
public ZkNodeProps getLeader(String collection, String shard) {
Map<String,ZkNodeProps> collectionLeaders = leaders.get(collection);
if (collectionLeaders == null) return null;
return collectionLeaders.get(shard);
}
/**
* Get shard properties or null if shard is not found.
*/
public ZkNodeProps getShardProps(final String collection, final String coreNodeName) {
Map<String, Slice> slices = getSlices(collection);
for(Slice slice: slices.values()) {
if(slice.getShards().get(coreNodeName)!=null) {
return slice.getShards().get(coreNodeName);
}
}
return null;
}
private void addRangeInfos(Set<String> collections) {
for (String collection : collections) {
@ -117,72 +142,72 @@ public class ClusterState implements JSONWriter.Writable {
* Get the index Slice for collection.
*/
public Slice getSlice(String collection, String slice) {
if (collectionStates.containsKey(collection)
&& collectionStates.get(collection).containsKey(slice))
return collectionStates.get(collection).get(slice);
return null;
}
if (collectionStates.containsKey(collection)
&& collectionStates.get(collection).containsKey(slice))
return collectionStates.get(collection).get(slice);
return null;
}
/**
* Get all slices for collection.
*/
public Map<String, Slice> getSlices(String collection) {
if(!collectionStates.containsKey(collection))
return null;
return Collections.unmodifiableMap(collectionStates.get(collection));
}
public Map<String, Slice> getSlices(String collection) {
if(!collectionStates.containsKey(collection))
return null;
return Collections.unmodifiableMap(collectionStates.get(collection));
}
/**
* Get collection names.
*/
public Set<String> getCollections() {
return Collections.unmodifiableSet(collectionStates.keySet());
}
/**
* Get collection names.
*/
public Set<String> getCollections() {
return Collections.unmodifiableSet(collectionStates.keySet());
}
/**
* @return Map&lt;collectionName, Map&lt;sliceName,Slice&gt;&gt;
*/
public Map<String, Map<String, Slice>> getCollectionStates() {
return Collections.unmodifiableMap(collectionStates);
}
/**
* @return Map&lt;collectionName, Map&lt;sliceName,Slice&gt;&gt;
*/
public Map<String, Map<String, Slice>> getCollectionStates() {
return Collections.unmodifiableMap(collectionStates);
}
/**
* Get names of the currently live nodes.
*/
public Set<String> getLiveNodes() {
return Collections.unmodifiableSet(liveNodes);
}
/**
* Get names of the currently live nodes.
*/
public Set<String> getLiveNodes() {
return Collections.unmodifiableSet(liveNodes);
}
/**
* Get shardId for core.
* @param coreNodeName in the form of nodeName_coreName
*/
public String getShardId(String coreNodeName) {
for (Entry<String, Map<String, Slice>> states: collectionStates.entrySet()){
for(Entry<String, Slice> slices: states.getValue().entrySet()) {
for(Entry<String, ZkNodeProps> shards: slices.getValue().getShards().entrySet()){
if(coreNodeName.equals(shards.getKey())) {
return slices.getKey();
}
}
}
}
return null;
}
/**
* Get shardId for core.
* @param coreNodeName in the form of nodeName_coreName
*/
public String getShardId(String coreNodeName) {
for (Entry<String, Map<String, Slice>> states: collectionStates.entrySet()){
for(Entry<String, Slice> slices: states.getValue().entrySet()) {
for(Entry<String, ZkNodeProps> shards: slices.getValue().getShards().entrySet()){
if(coreNodeName.equals(shards.getKey())) {
return slices.getKey();
}
}
}
}
return null;
}
/**
* Check if node is alive.
*/
public boolean liveNodesContain(String name) {
return liveNodes.contains(name);
}
public RangeInfo getRanges(String collection) {
/**
* Check if node is alive.
*/
public boolean liveNodesContain(String name) {
return liveNodes.contains(name);
}
public RangeInfo getRanges(String collection) {
// TODO: store this in zk
RangeInfo rangeInfo = rangeInfos.get(collection);
return rangeInfo;
}
return rangeInfo;
}
private RangeInfo addRangeInfo(String collection) {
List<Range> ranges;
@ -227,29 +252,39 @@ public class ClusterState implements JSONWriter.Writable {
throw new IllegalStateException("The HashPartitioner failed");
}
@Override
public String toString() {
StringBuilder sb = new StringBuilder();
sb.append("live nodes:" + liveNodes);
sb.append(" collections:" + collectionStates);
return sb.toString();
}
@Override
public String toString() {
StringBuilder sb = new StringBuilder();
sb.append("live nodes:" + liveNodes);
sb.append(" collections:" + collectionStates);
return sb.toString();
}
/**
* Create ClusterState by reading the current state from zookeeper.
*/
public static ClusterState load(SolrZkClient zkClient, Set<String> liveNodes) throws KeeperException, InterruptedException {
/**
* Create ClusterState by reading the current state from zookeeper.
*/
public static ClusterState load(SolrZkClient zkClient, Set<String> liveNodes) throws KeeperException, InterruptedException {
Stat stat = new Stat();
byte[] state = zkClient.getData(ZkStateReader.CLUSTER_STATE,
null, null, true);
return load(state, liveNodes);
}
/**
* Create ClusterState from json string that is typically stored in zookeeper.
*/
public static ClusterState load(byte[] bytes, Set<String> liveNodes) {
null, stat, true);
return load(stat.getVersion(), state, liveNodes);
}
/**
* Create ClusterState from json string that is typically stored in zookeeper.
*
* Use {@link ClusterState#load(SolrZkClient, Set)} instead, unless you want to
* do something more when getting the data - such as get the stat, set watch, etc.
*
* @param version zk version of the clusterstate.json file (bytes)
* @param bytes clusterstate.json as a byte array
* @param liveNodes list of live nodes
* @return the ClusterState
*/
public static ClusterState load(Integer version, byte[] bytes, Set<String> liveNodes) {
if (bytes == null || bytes.length == 0) {
return new ClusterState(liveNodes, Collections.<String, Map<String,Slice>>emptyMap());
return new ClusterState(version, liveNodes, Collections.<String, Map<String,Slice>>emptyMap());
}
LinkedHashMap<String, Object> stateMap = (LinkedHashMap<String, Object>) ZkStateReader.fromJSON(bytes);
@ -269,8 +304,8 @@ public class ClusterState implements JSONWriter.Writable {
}
state.put(collectionName, slices);
}
return new ClusterState(liveNodes, state);
}
return new ClusterState(version, liveNodes, state);
}
@Override
public void write(JSONWriter jsonWriter) {
@ -282,5 +317,41 @@ public class ClusterState implements JSONWriter.Writable {
private ArrayList<String> shardList;
}
/**
* The version of clusterstate.json in ZooKeeper.
*
* @return null if ClusterState was created for publication, not consumption
*/
public Integer getZkClusterStateVersion() {
return zkClusterStateVersion;
}
@Override
public int hashCode() {
final int prime = 31;
int result = 1;
result = prime * result
+ ((zkClusterStateVersion == null) ? 0 : zkClusterStateVersion.hashCode());
result = prime * result + ((liveNodes == null) ? 0 : liveNodes.hashCode());
return result;
}
@Override
public boolean equals(Object obj) {
if (this == obj) return true;
if (obj == null) return false;
if (getClass() != obj.getClass()) return false;
ClusterState other = (ClusterState) obj;
if (zkClusterStateVersion == null) {
if (other.zkClusterStateVersion != null) return false;
} else if (!zkClusterStateVersion.equals(other.zkClusterStateVersion)) return false;
if (liveNodes == null) {
if (other.liveNodes != null) return false;
} else if (!liveNodes.equals(other.liveNodes)) return false;
return true;
}
}

17
solr/solrj/src/java/org/apache/solr/common/cloud/ZkStateReader.java
View File

@ -42,6 +42,7 @@ import org.apache.zookeeper.KeeperException;
import org.apache.zookeeper.WatchedEvent;
import org.apache.zookeeper.Watcher;
import org.apache.zookeeper.Watcher.Event.EventType;
import org.apache.zookeeper.data.Stat;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
@ -193,10 +194,11 @@ public class ZkStateReader {
synchronized (ZkStateReader.this.getUpdateLock()) {
// remake watch
final Watcher thisWatch = this;
byte[] data = zkClient.getData(CLUSTER_STATE, thisWatch, null,
Stat stat = new Stat();
byte[] data = zkClient.getData(CLUSTER_STATE, thisWatch, stat ,
true);
ClusterState clusterState = ClusterState.load(data,
ClusterState clusterState = ClusterState.load(stat.getVersion(), data,
ZkStateReader.this.clusterState.getLiveNodes());
// update volatile
ZkStateReader.this.clusterState = clusterState;
@ -242,8 +244,10 @@ public class ZkStateReader {
LIVE_NODES_ZKNODE, this, true);
Set<String> liveNodesSet = new HashSet<String>();
liveNodesSet.addAll(liveNodes);
ClusterState clusterState = new ClusterState(liveNodesSet,
ZkStateReader.this.clusterState.getCollectionStates());
ClusterState clusterState = new ClusterState(
ZkStateReader.this.clusterState.getZkClusterStateVersion(),
liveNodesSet, ZkStateReader.this.clusterState
.getCollectionStates());
ZkStateReader.this.clusterState = clusterState;
}
} catch (KeeperException e) {
@ -293,7 +297,8 @@ public class ZkStateReader {
clusterState = ClusterState.load(zkClient, liveNodesSet);
} else {
log.info("Updating live nodes from ZooKeeper... ");
clusterState = new ClusterState(liveNodesSet,
clusterState = new ClusterState(
ZkStateReader.this.clusterState.getZkClusterStateVersion(), liveNodesSet,
ZkStateReader.this.clusterState.getCollectionStates());
}
}
@ -325,7 +330,7 @@ public class ZkStateReader {
clusterState = ClusterState.load(zkClient, liveNodesSet);
} else {
log.info("Updating live nodes from ZooKeeper... ");
clusterState = new ClusterState(liveNodesSet, ZkStateReader.this.clusterState.getCollectionStates());
clusterState = new ClusterState(ZkStateReader.this.clusterState .getZkClusterStateVersion(), liveNodesSet, ZkStateReader.this.clusterState.getCollectionStates());
}
ZkStateReader.this.clusterState = clusterState;