Updated Branches: refs/heads/cassandra-2.0 91f976ce1 -> b45715665 refs/heads/trunk fa86ce88c -> 0619da2ae
Improve Gossip tolerance of abnormally slow tasks patch by jbellis; reviewed by jasobrown for CASSANDRA-6338 Project: http://git-wip-us.apache.org/repos/asf/cassandra/repo Commit: http://git-wip-us.apache.org/repos/asf/cassandra/commit/b4571566 Tree: http://git-wip-us.apache.org/repos/asf/cassandra/tree/b4571566 Diff: http://git-wip-us.apache.org/repos/asf/cassandra/diff/b4571566 Branch: refs/heads/cassandra-2.0 Commit: b45715665845f7d2b597adfc0fc9c055b271e8d0 Parents: 91f976c Author: Jonathan Ellis <[email protected]> Authored: Wed Nov 13 09:37:45 2013 -0600 Committer: Jonathan Ellis <[email protected]> Committed: Wed Nov 13 09:37:45 2013 -0600 ---------------------------------------------------------------------- CHANGES.txt | 1 + src/java/org/apache/cassandra/gms/Gossiper.java | 29 ++++++++++++++++++-- .../cassandra/net/MessageDeliveryTask.java | 10 +++++++ 3 files changed, 38 insertions(+), 2 deletions(-) ---------------------------------------------------------------------- http://git-wip-us.apache.org/repos/asf/cassandra/blob/b4571566/CHANGES.txt ---------------------------------------------------------------------- diff --git a/CHANGES.txt b/CHANGES.txt index 3305739..e4f1862 100644 --- a/CHANGES.txt +++ b/CHANGES.txt @@ -1,4 +1,5 @@ 2.0.3 + * Improve Gossip tolerance of abnormally slow tasks (CASSANDRA-6338) * Fix trying to hint timed out counter writes (CASSANDRA-6322) * Allow restoring specific columnfamilies from archived CL (CASSANDRA-4809) * Avoid flushing compaction_history after each operation (CASSANDRA-6287) http://git-wip-us.apache.org/repos/asf/cassandra/blob/b4571566/src/java/org/apache/cassandra/gms/Gossiper.java ---------------------------------------------------------------------- diff --git a/src/java/org/apache/cassandra/gms/Gossiper.java b/src/java/org/apache/cassandra/gms/Gossiper.java index 8237fa3..e3079c0 100644 --- a/src/java/org/apache/cassandra/gms/Gossiper.java +++ b/src/java/org/apache/cassandra/gms/Gossiper.java @@ -34,6 +34,9 @@ import org.slf4j.Logger; import org.slf4j.LoggerFactory; import org.apache.cassandra.concurrent.DebuggableScheduledThreadPoolExecutor; +import org.apache.cassandra.concurrent.JMXEnabledThreadPoolExecutor; +import org.apache.cassandra.concurrent.Stage; +import org.apache.cassandra.concurrent.StageManager; import org.apache.cassandra.config.DatabaseDescriptor; import org.apache.cassandra.dht.Token; import org.apache.cassandra.net.IAsyncCallback; @@ -113,6 +116,8 @@ public class Gossiper implements IFailureDetectionEventListener, GossiperMBean private boolean inShadowRound = false; + private volatile long lastProcessedMessageAt = System.currentTimeMillis(); + private class GossipTask implements Runnable { public void run() @@ -162,8 +167,6 @@ public class Gossiper implements IFailureDetectionEventListener, GossiperMBean if (!gossipedToSeed || liveEndpoints.size() < seeds.size()) doGossipToSeed(message); - if (logger.isTraceEnabled()) - logger.trace("Performing status check ..."); doStatusCheck(); } } @@ -193,6 +196,11 @@ public class Gossiper implements IFailureDetectionEventListener, GossiperMBean } } + public void setLastProcessedMessageAt(long timeInMillis) + { + this.lastProcessedMessageAt = timeInMillis; + } + protected void checkSeedContact(InetAddress ep) { if (!seedContacted && seeds.contains(ep)) @@ -586,9 +594,26 @@ public class Gossiper implements IFailureDetectionEventListener, GossiperMBean private void doStatusCheck() { + if (logger.isTraceEnabled()) + logger.trace("Performing status check ..."); + long now = System.currentTimeMillis(); long nowNano = System.nanoTime(); + long pending = ((JMXEnabledThreadPoolExecutor) StageManager.getStage(Stage.GOSSIP)).getPendingTasks(); + if (pending > 0 && lastProcessedMessageAt < now - 1000) + { + // if some new messages just arrived, give the executor some time to work on them + Uninterruptibles.sleepUninterruptibly(100, TimeUnit.MILLISECONDS); + + // still behind? something's broke + if (lastProcessedMessageAt < now - 1000) + { + logger.warn("Gossip stage has {} pending tasks; skipping status check (no nodes will be marked down)", pending); + return; + } + } + Set<InetAddress> eps = endpointStateMap.keySet(); for (InetAddress endpoint : eps) { http://git-wip-us.apache.org/repos/asf/cassandra/blob/b4571566/src/java/org/apache/cassandra/net/MessageDeliveryTask.java ---------------------------------------------------------------------- diff --git a/src/java/org/apache/cassandra/net/MessageDeliveryTask.java b/src/java/org/apache/cassandra/net/MessageDeliveryTask.java index d7645b6..e49b93c 100644 --- a/src/java/org/apache/cassandra/net/MessageDeliveryTask.java +++ b/src/java/org/apache/cassandra/net/MessageDeliveryTask.java @@ -17,9 +17,13 @@ */ package org.apache.cassandra.net; +import java.util.EnumSet; + import org.slf4j.Logger; import org.slf4j.LoggerFactory; +import org.apache.cassandra.gms.Gossiper; + public class MessageDeliveryTask implements Runnable { private static final Logger logger = LoggerFactory.getLogger(MessageDeliveryTask.class); @@ -54,5 +58,11 @@ public class MessageDeliveryTask implements Runnable } verbHandler.doVerb(message, id); + if (GOSSIP_VERBS.contains(message.verb)) + Gossiper.instance.setLastProcessedMessageAt(constructionTime); } + + EnumSet<MessagingService.Verb> GOSSIP_VERBS = EnumSet.of(MessagingService.Verb.GOSSIP_DIGEST_ACK, + MessagingService.Verb.GOSSIP_DIGEST_ACK2, + MessagingService.Verb.GOSSIP_DIGEST_SYN); }
