[ https://issues.apache.org/jira/browse/CASSANDRA-13265?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15888439#comment-15888439 ]
Christian Esken commented on CASSANDRA-13265: --------------------------------------------- Here is one possibly very important observation. It looks like Coalescing is doing an infinite loop while doing maybeSleep(). I checked 10 Thread dumps, and in each of them the Thread was at the same location. Is it possible that averageGap is 0? This would lead to infinite recursion. {code} private static boolean maybeSleep(int messages, long averageGap, long maxCoalesceWindow, Parker parker) { // only sleep if we can expect to double the number of messages we're sending in the time interval long sleep = messages * averageGap; // TODO can averageGap be 0 ? if (sleep > maxCoalesceWindow) return false; // assume we receive as many messages as we expect; apply the same logic to the future batch: // expect twice as many messages to consider sleeping for "another" interval; this basically translates // to doubling our sleep period until we exceed our max sleep window while (sleep * 2 < maxCoalesceWindow) sleep *= 2; // <<<<<<<<<<<<<<<< CoalescingStrategies:106 parker.park(sleep); return true; } {code} If sum is bigger than MEASURED_INTERVAL, then averageGap() returns 0. I am aware that this is highly unlikely, but I cannot explain the likely hanging in maybeSleep() line 106. {code} private long averageGap() { if (sum == 0) return Integer.MAX_VALUE; return MEASURED_INTERVAL / sum; } {code} > Communication breakdown in OutboundTcpConnection > ------------------------------------------------ > > Key: CASSANDRA-13265 > URL: https://issues.apache.org/jira/browse/CASSANDRA-13265 > Project: Cassandra > Issue Type: Bug > Environment: Cassandra 3.0.9 > Java HotSpot(TM) 64-Bit Server VM version 25.112-b15 (Java version > 1.8.0_112-b15) > Linux 3.16 > Reporter: Christian Esken > Assignee: Christian Esken > Attachments: cassandra.pb-cache4-dus.2017-02-17-19-36-26.chist.xz, > cassandra.pb-cache4-dus.2017-02-17-19-36-26.td.xz > > > I observed that sometimes a single node in a Cassandra cluster fails to > communicate to the other nodes. This can happen at any time, during peak load > or low load. Restarting that single node from the cluster fixes the issue. > Before going in to details, I want to state that I have analyzed the > situation and am already developing a possible fix. Here is the analysis so > far: > - A Threaddump in this situation showed 324 Threads in the > OutboundTcpConnection class that want to lock the backlog queue for doing > expiration. > - A class histogram shows 262508 instances of > OutboundTcpConnection$QueuedMessage. > What is the effect of it? As soon as the Cassandra node has reached a certain > amount of queued messages, it starts thrashing itself to death. Each of the > Thread fully locks the Queue for reading and writing by calling > iterator.next(), making the situation worse and worse. > - Writing: Only after 262508 locking operation it can progress with actually > writing to the Queue. > - Reading: Is also blocked, as 324 Threads try to do iterator.next(), and > fully lock the Queue > This means: Writing blocks the Queue for reading, and readers might even be > starved which makes the situation even worse. > ----- > The setup is: > - 3-node cluster > - replication factor 2 > - Consistency LOCAL_ONE > - No remote DC's > - high write throughput (100000 INSERT statements per second and more during > peak times). > -- This message was sent by Atlassian JIRA (v6.3.15#6346)