You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@lucene.apache.org by ma...@apache.org on 2012/12/15 23:25:29 UTC
svn commit: r1422393 - in /lucene/dev/branches/branch_4x: ./ dev-tools/
lucene/ lucene/analysis/
lucene/analysis/icu/src/java/org/apache/lucene/collation/ lucene/backwards/
lucene/benchmark/ lucene/codecs/ lucene/core/
lucene/core/src/test/org/apache/l...
Author: markrmiller
Date: Sat Dec 15 22:25:26 2012
New Revision: 1422393
URL: http://svn.apache.org/viewvc?rev=1422393&view=rev
Log:
SOLR-4199: When doing zk retries due to connectionloss, rather than just retrying for 2 minutes, retry in proportion to the session timeout.
Modified:
lucene/dev/branches/branch_4x/ (props changed)
lucene/dev/branches/branch_4x/dev-tools/ (props changed)
lucene/dev/branches/branch_4x/lucene/ (props changed)
lucene/dev/branches/branch_4x/lucene/BUILD.txt (props changed)
lucene/dev/branches/branch_4x/lucene/CHANGES.txt (props changed)
lucene/dev/branches/branch_4x/lucene/JRE_VERSION_MIGRATION.txt (props changed)
lucene/dev/branches/branch_4x/lucene/LICENSE.txt (props changed)
lucene/dev/branches/branch_4x/lucene/MIGRATE.txt (props changed)
lucene/dev/branches/branch_4x/lucene/NOTICE.txt (props changed)
lucene/dev/branches/branch_4x/lucene/README.txt (props changed)
lucene/dev/branches/branch_4x/lucene/SYSTEM_REQUIREMENTS.txt (props changed)
lucene/dev/branches/branch_4x/lucene/analysis/ (props changed)
lucene/dev/branches/branch_4x/lucene/analysis/icu/src/java/org/apache/lucene/collation/ICUCollationKeyFilterFactory.java (props changed)
lucene/dev/branches/branch_4x/lucene/backwards/ (props changed)
lucene/dev/branches/branch_4x/lucene/benchmark/ (props changed)
lucene/dev/branches/branch_4x/lucene/build.xml (props changed)
lucene/dev/branches/branch_4x/lucene/codecs/ (props changed)
lucene/dev/branches/branch_4x/lucene/common-build.xml (props changed)
lucene/dev/branches/branch_4x/lucene/core/ (props changed)
lucene/dev/branches/branch_4x/lucene/core/src/test/org/apache/lucene/index/TestBackwardsCompatibility.java (props changed)
lucene/dev/branches/branch_4x/lucene/core/src/test/org/apache/lucene/index/index.40.cfs.zip (props changed)
lucene/dev/branches/branch_4x/lucene/core/src/test/org/apache/lucene/index/index.40.nocfs.zip (props changed)
lucene/dev/branches/branch_4x/lucene/core/src/test/org/apache/lucene/index/index.40.optimized.cfs.zip (props changed)
lucene/dev/branches/branch_4x/lucene/core/src/test/org/apache/lucene/index/index.40.optimized.nocfs.zip (props changed)
lucene/dev/branches/branch_4x/lucene/demo/ (props changed)
lucene/dev/branches/branch_4x/lucene/facet/ (props changed)
lucene/dev/branches/branch_4x/lucene/grouping/ (props changed)
lucene/dev/branches/branch_4x/lucene/highlighter/ (props changed)
lucene/dev/branches/branch_4x/lucene/ivy-settings.xml (props changed)
lucene/dev/branches/branch_4x/lucene/join/ (props changed)
lucene/dev/branches/branch_4x/lucene/licenses/ (props changed)
lucene/dev/branches/branch_4x/lucene/memory/ (props changed)
lucene/dev/branches/branch_4x/lucene/misc/ (props changed)
lucene/dev/branches/branch_4x/lucene/module-build.xml (props changed)
lucene/dev/branches/branch_4x/lucene/queries/ (props changed)
lucene/dev/branches/branch_4x/lucene/queryparser/ (props changed)
lucene/dev/branches/branch_4x/lucene/sandbox/ (props changed)
lucene/dev/branches/branch_4x/lucene/site/ (props changed)
lucene/dev/branches/branch_4x/lucene/spatial/ (props changed)
lucene/dev/branches/branch_4x/lucene/suggest/ (props changed)
lucene/dev/branches/branch_4x/lucene/test-framework/ (props changed)
lucene/dev/branches/branch_4x/lucene/tools/ (props changed)
lucene/dev/branches/branch_4x/solr/ (props changed)
lucene/dev/branches/branch_4x/solr/CHANGES.txt (contents, props changed)
lucene/dev/branches/branch_4x/solr/LICENSE.txt (props changed)
lucene/dev/branches/branch_4x/solr/NOTICE.txt (props changed)
lucene/dev/branches/branch_4x/solr/README.txt (props changed)
lucene/dev/branches/branch_4x/solr/SYSTEM_REQUIREMENTS.txt (props changed)
lucene/dev/branches/branch_4x/solr/build.xml (props changed)
lucene/dev/branches/branch_4x/solr/cloud-dev/ (props changed)
lucene/dev/branches/branch_4x/solr/common-build.xml (props changed)
lucene/dev/branches/branch_4x/solr/contrib/ (props changed)
lucene/dev/branches/branch_4x/solr/core/ (props changed)
lucene/dev/branches/branch_4x/solr/core/src/java/org/apache/solr/cloud/LeaderElector.java
lucene/dev/branches/branch_4x/solr/core/src/java/org/apache/solr/cloud/ZkController.java
lucene/dev/branches/branch_4x/solr/example/ (props changed)
lucene/dev/branches/branch_4x/solr/licenses/ (props changed)
lucene/dev/branches/branch_4x/solr/licenses/httpclient-LICENSE-ASL.txt (props changed)
lucene/dev/branches/branch_4x/solr/licenses/httpclient-NOTICE.txt (props changed)
lucene/dev/branches/branch_4x/solr/licenses/httpcore-LICENSE-ASL.txt (props changed)
lucene/dev/branches/branch_4x/solr/licenses/httpcore-NOTICE.txt (props changed)
lucene/dev/branches/branch_4x/solr/licenses/httpmime-LICENSE-ASL.txt (props changed)
lucene/dev/branches/branch_4x/solr/licenses/httpmime-NOTICE.txt (props changed)
lucene/dev/branches/branch_4x/solr/scripts/ (props changed)
lucene/dev/branches/branch_4x/solr/site/ (props changed)
lucene/dev/branches/branch_4x/solr/solrj/ (props changed)
lucene/dev/branches/branch_4x/solr/solrj/src/java/org/apache/solr/common/cloud/SolrZkClient.java
lucene/dev/branches/branch_4x/solr/solrj/src/java/org/apache/solr/common/cloud/ZkCmdExecutor.java
lucene/dev/branches/branch_4x/solr/solrj/src/java/org/apache/solr/common/cloud/ZkStateReader.java
lucene/dev/branches/branch_4x/solr/test-framework/ (props changed)
lucene/dev/branches/branch_4x/solr/testlogging.properties (props changed)
lucene/dev/branches/branch_4x/solr/webapp/ (props changed)
Modified: lucene/dev/branches/branch_4x/solr/CHANGES.txt
URL: http://svn.apache.org/viewvc/lucene/dev/branches/branch_4x/solr/CHANGES.txt?rev=1422393&r1=1422392&r2=1422393&view=diff
==============================================================================
--- lucene/dev/branches/branch_4x/solr/CHANGES.txt (original)
+++ lucene/dev/branches/branch_4x/solr/CHANGES.txt Sat Dec 15 22:25:26 2012
@@ -175,6 +175,10 @@ Optimizations
* SOLR-4063: Allow CoreContainer to load multiple SolrCores in parallel rather
than just serially. (Mark Miller)
+
+* SOLR-4199: When doing zk retries due to connectionloss, rather than just
+ retrying for 2 minutes, retry in proportion to the session timeout.
+ (Mark Miller)
Bug Fixes
----------------------
Modified: lucene/dev/branches/branch_4x/solr/core/src/java/org/apache/solr/cloud/LeaderElector.java
URL: http://svn.apache.org/viewvc/lucene/dev/branches/branch_4x/solr/core/src/java/org/apache/solr/cloud/LeaderElector.java?rev=1422393&r1=1422392&r2=1422393&view=diff
==============================================================================
--- lucene/dev/branches/branch_4x/solr/core/src/java/org/apache/solr/cloud/LeaderElector.java (original)
+++ lucene/dev/branches/branch_4x/solr/core/src/java/org/apache/solr/cloud/LeaderElector.java Sat Dec 15 22:25:26 2012
@@ -62,10 +62,11 @@ public class LeaderElector {
protected SolrZkClient zkClient;
- private ZkCmdExecutor zkCmdExecutor = new ZkCmdExecutor();
+ private ZkCmdExecutor zkCmdExecutor;
public LeaderElector(SolrZkClient zkClient) {
this.zkClient = zkClient;
+ zkCmdExecutor = new ZkCmdExecutor((int) (zkClient.getZkClientTimeout()/1000.0 + 3000));
}
/**
Modified: lucene/dev/branches/branch_4x/solr/core/src/java/org/apache/solr/cloud/ZkController.java
URL: http://svn.apache.org/viewvc/lucene/dev/branches/branch_4x/solr/core/src/java/org/apache/solr/cloud/ZkController.java?rev=1422393&r1=1422392&r2=1422393&view=diff
==============================================================================
--- lucene/dev/branches/branch_4x/solr/core/src/java/org/apache/solr/cloud/ZkController.java (original)
+++ lucene/dev/branches/branch_4x/solr/core/src/java/org/apache/solr/cloud/ZkController.java Sat Dec 15 22:25:26 2012
@@ -99,9 +99,7 @@ public final class ZkController {
private final DistributedQueue overseerJobQueue;
private final DistributedQueue overseerCollectionQueue;
- // package private for tests
-
- static final String CONFIGS_ZKNODE = "/configs";
+ public static final String CONFIGS_ZKNODE = "/configs";
public final static String COLLECTION_PARAM_PREFIX="collection.";
public final static String CONFIGNAME_PROP="configName";
@@ -142,6 +140,8 @@ public final class ZkController {
private int clientTimeout;
+ private volatile boolean isClosed;
+
public ZkController(final CoreContainer cc, String zkServerAddress, int zkClientTimeout, int zkClientConnectTimeout, String localHost, String locaHostPort,
String localHostContext, String leaderVoteWait, final CurrentCoreDescriptorProvider registerOnReconnect) throws InterruptedException,
TimeoutException, IOException {
@@ -241,7 +241,7 @@ public final class ZkController {
this.overseerJobQueue = Overseer.getInQueue(zkClient);
this.overseerCollectionQueue = Overseer.getCollectionQueue(zkClient);
- cmdExecutor = new ZkCmdExecutor();
+ cmdExecutor = new ZkCmdExecutor(zkClientTimeout);
leaderElector = new LeaderElector(zkClient);
zkStateReader = new ZkStateReader(zkClient);
@@ -266,6 +266,9 @@ public final class ZkController {
descriptor.getCloudDescriptor().isLeader = false;
publish(descriptor, ZkStateReader.DOWN, updateLastPublished);
} catch (Exception e) {
+ if (isClosed) {
+ return;
+ }
try {
Thread.sleep(1000);
} catch (InterruptedException e1) {
@@ -282,6 +285,9 @@ public final class ZkController {
waitForLeaderToSeeDownState(descriptor, coreZkNodeName);
} catch (Exception e) {
SolrException.log(log, "", e);
+ if (isClosed) {
+ return;
+ }
try {
Thread.sleep(5000);
} catch (InterruptedException e1) {
@@ -307,6 +313,7 @@ public final class ZkController {
* Closes the underlying ZooKeeper client.
*/
public void close() {
+ this.isClosed = true;
if (cmdDistribExecutor != null) {
try {
Modified: lucene/dev/branches/branch_4x/solr/solrj/src/java/org/apache/solr/common/cloud/SolrZkClient.java
URL: http://svn.apache.org/viewvc/lucene/dev/branches/branch_4x/solr/solrj/src/java/org/apache/solr/common/cloud/SolrZkClient.java?rev=1422393&r1=1422392&r2=1422393&view=diff
==============================================================================
--- lucene/dev/branches/branch_4x/solr/solrj/src/java/org/apache/solr/common/cloud/SolrZkClient.java (original)
+++ lucene/dev/branches/branch_4x/solr/solrj/src/java/org/apache/solr/common/cloud/SolrZkClient.java Sat Dec 15 22:25:26 2012
@@ -71,11 +71,16 @@ public class SolrZkClient {
private volatile SolrZooKeeper keeper;
- private ZkCmdExecutor zkCmdExecutor = new ZkCmdExecutor();
+ private ZkCmdExecutor zkCmdExecutor;
private volatile boolean isClosed = false;
private ZkClientConnectionStrategy zkClientConnectionStrategy;
+ private int zkClientTimeout;
+ public int getZkClientTimeout() {
+ return zkClientTimeout;
+ }
+
public SolrZkClient(String zkServerAddress, int zkClientTimeout) {
this(zkServerAddress, zkClientTimeout, new DefaultConnectionStrategy(), null);
}
@@ -92,6 +97,9 @@ public class SolrZkClient {
public SolrZkClient(String zkServerAddress, int zkClientTimeout,
ZkClientConnectionStrategy strat, final OnReconnect onReconnect, int clientConnectTimeout) {
this.zkClientConnectionStrategy = strat;
+ this.zkClientTimeout = zkClientTimeout;
+ // we must retry at least as long as the session timeout
+ zkCmdExecutor = new ZkCmdExecutor(zkClientTimeout);
connManager = new ConnectionManager("ZooKeeperConnection Watcher:"
+ zkServerAddress, this, zkServerAddress, zkClientTimeout, strat, onReconnect);
try {
Modified: lucene/dev/branches/branch_4x/solr/solrj/src/java/org/apache/solr/common/cloud/ZkCmdExecutor.java
URL: http://svn.apache.org/viewvc/lucene/dev/branches/branch_4x/solr/solrj/src/java/org/apache/solr/common/cloud/ZkCmdExecutor.java?rev=1422393&r1=1422392&r2=1422393&view=diff
==============================================================================
--- lucene/dev/branches/branch_4x/solr/solrj/src/java/org/apache/solr/common/cloud/ZkCmdExecutor.java (original)
+++ lucene/dev/branches/branch_4x/solr/solrj/src/java/org/apache/solr/common/cloud/ZkCmdExecutor.java Sat Dec 15 22:25:26 2012
@@ -27,11 +27,13 @@ import org.apache.zookeeper.data.ACL;
public class ZkCmdExecutor {
- private long retryDelay = 1000L;
- private int retryCount = 15;
+ private long retryDelay = 1300L; // 300 ms over for padding
+ private int retryCount;
private List<ACL> acl = ZooDefs.Ids.OPEN_ACL_UNSAFE;
- public ZkCmdExecutor() {
+ public ZkCmdExecutor(int timeoutms) {
+ double timeouts = timeoutms / 1000.0;
+ this.retryCount = Math.round(0.5f * ((float)Math.sqrt(8.0f * timeouts + 1.0f) - 1.0f));
}
public List<ACL> getAcl() {
Modified: lucene/dev/branches/branch_4x/solr/solrj/src/java/org/apache/solr/common/cloud/ZkStateReader.java
URL: http://svn.apache.org/viewvc/lucene/dev/branches/branch_4x/solr/solrj/src/java/org/apache/solr/common/cloud/ZkStateReader.java?rev=1422393&r1=1422392&r2=1422393&view=diff
==============================================================================
--- lucene/dev/branches/branch_4x/solr/solrj/src/java/org/apache/solr/common/cloud/ZkStateReader.java (original)
+++ lucene/dev/branches/branch_4x/solr/solrj/src/java/org/apache/solr/common/cloud/ZkStateReader.java Sat Dec 15 22:25:26 2012
@@ -127,14 +127,16 @@ public class ZkStateReader {
private boolean closeClient = false;
- private ZkCmdExecutor cmdExecutor = new ZkCmdExecutor();
+ private ZkCmdExecutor cmdExecutor;
public ZkStateReader(SolrZkClient zkClient) {
this.zkClient = zkClient;
+ initZkCmdExecutor(zkClient.getZkClientTimeout());
}
-
+
public ZkStateReader(String zkServerAddress, int zkClientTimeout, int zkClientConnectTimeout) throws InterruptedException, TimeoutException, IOException {
closeClient = true;
+ initZkCmdExecutor(zkClientTimeout);
zkClient = new SolrZkClient(zkServerAddress, zkClientTimeout, zkClientConnectTimeout,
// on reconnect, reload cloud info
new OnReconnect() {
@@ -159,6 +161,11 @@ public class ZkStateReader {
});
}
+ private void initZkCmdExecutor(int zkClientTimeout) {
+ // we must retry at least as long as the session timeout
+ cmdExecutor = new ZkCmdExecutor(zkClientTimeout);
+ }
+
// load and publish a new CollectionInfo
public void updateClusterState(boolean immediate) throws KeeperException, InterruptedException {
updateClusterState(immediate, false);