Did not help finally.

 

So I enabled  logging at debug level.

The log files tell me that the node being added is communicating with the other nodes (that are seed nodes). Still nothing seems to be returning to that node.

The log files on the other nodes are detecting the shadow request, but no other information like being unable to send something back.

And as before, just restarting that node once more does the trick and bootstrap is proceeding.

 

Maybe the problem has something to do about the gossip state?

My test case is : Decommission node 10.164.8.93, restart a clean node 10.164.8.93 and let it bootstrap.

 

In my test case, I see that 7 minutes before the node is added again to the ring the other nodes are detecting the decommission of the node.

 

2014-02-26 10:23:21.443 60000 elapsed, /10.164.8.93 gossip quarantine over

2014-02-26 10:23:21.444 Ignoring state change for dead or unknown endpoint: /10.164.8.93

2014-02-26 10:23:55.636 Forcing conviction of /10.164.8.93

2014-02-26 10:24:00.230 Reseting version for /10.164.8.93

2014-02-26 10:24:00.230 Reseting version for /10.164.8.93

 

 

At the time the node 10.164.8.93 is added the log shows :

--------------------------------------------------------------------

 

2014-02-26 10:30:03.015 Cassandra version: 2.0.5-SNAPSHOT

2014-02-26 10:30:03.016 Thrift API version: 19.39.0

2014-02-26 10:30:03.018 CQL supported versions: 2.0.0,3.1.4 (default: 3.1.4)

2014-02-26 10:30:03.029 Loading persisted ring state

2014-02-26 10:30:03.034 Starting shadow gossip round to check for endpoint collision

2014-02-26 10:30:03.034 Starting Messaging Service on port 9804

2014-02-26 10:30:03.046 attempting to connect to /10.164.8.249

2014-02-26 10:30:03.047 attempting to connect to /10.164.8.250

2014-02-26 10:30:03.048 attempting to connect to /10.164.8.92

2014-02-26 10:30:03.051 Handshaking version with /10.164.8.250

2014-02-26 10:30:03.052 Handshaking version with /10.164.8.249

2014-02-26 10:30:03.052 Handshaking version with /10.164.8.92

2014-02-26 10:30:34.059 Exception encountered during startup

java.lang.RuntimeException: Unable to gossip with any seeds

                at org.apache.cassandra.gms.Gossiper.doShadowRound(Gossiper.java:1173) ~[apache-cassandra-2.0.5-SNAPSHOT.jar:2.0.5-SNAPSHOT]

                at org.apache.cassandra.service.StorageService.checkForEndpointCollision(StorageService.java:424) ~[apache-cassandra-2.0.5-SNAPSHOT.jar:2.0.5-SNAPSHOT]

                at org.apache.cassandra.service.StorageService.joinTokenRing(StorageService.java:615) ~[apache-cassandra-2.0.5-SNAPSHOT.jar:2.0.5-SNAPSHOT]

                at org.apache.cassandra.service.StorageService.initServer(StorageService.java:583) ~[apache-cassandra-2.0.5-SNAPSHOT.jar:2.0.5-SNAPSHOT]

                at org.apache.cassandra.service.StorageService.initServer(StorageService.java:482) ~[apache-cassandra-2.0.5-SNAPSHOT.jar:2.0.5-SNAPSHOT]

                at org.apache.cassandra.service.CassandraDaemon.setup(CassandraDaemon.java:348) ~[apache-cassandra-2.0.5-SNAPSHOT.jar:2.0.5-SNAPSHOT]

                at org.apache.cassandra.service.CassandraDaemon.activate(CassandraDaemon.java:465) ~[apache-cassandra-2.0.5-SNAPSHOT.jar:2.0.5-SNAPSHOT]

                at be.landc.services.search.server.db.baseserver.indexsearch.store.cassandra.CassandraStore$CassThread.startUpCassandra(CassandraStore.java:495) [landc-services-search-1.2.0-SNAPSHOT.jar:1.2.0-SNAPSHOT-92200]

                at be.landc.services.search.server.db.baseserver.indexsearch.store.cassandra.CassandraStore$CassThread.run(CassandraStore.java:461) [landc-services-search-1.2.0-SNAPSHOT.jar:1.2.0-SNAPSHOT-92200]

2014-02-26 10:30:34.069 Exception in thread Thread[StorageServiceShutdownHook,5,main]

java.lang.NullPointerException: null

                at org.apache.cassandra.gms.Gossiper.stop(Gossiper.java:1250) ~[apache-cassandra-2.0.5-SNAPSHOT.jar:2.0.5-SNAPSHOT]

                at org.apache.cassandra.service.StorageService$1.runMayThrow(StorageService.java:550) ~[apache-cassandra-2.0.5-SNAPSHOT.jar:2.0.5-SNAPSHOT]

                at org.apache.cassandra.utils.WrappedRunnable.run(WrappedRunnable.java:28) ~[apache-cassandra-2.0.5-SNAPSHOT.jar:2.0.5-SNAPSHOT]

                at java.lang.Thread.run(Thread.java:724) ~[na:1.7.0_40]

2014-02-26 10:31:34.081 ShutDownHook requests shutdown on be.landc.framework.service.ipl.Boot@730e8516

 

 

And at that time 3 other  nodes print log information :

-----------------------------------------------------------------

 

2014-02-26 10:30:03.051 Connection version 7 from /10.164.8.93

2014-02-26 10:30:03.065 Upgrading incoming connection to be compressed

2014-02-26 10:30:03.130 Max version for /10.164.8.93 is 7

2014-02-26 10:30:03.130 Setting version 7 for /10.164.8.93

2014-02-26 10:30:03.131 set version for /10.164.8.93 to 7

2014-02-26 10:30:03.131 Shadow request received, adding all states

 

 

Any more information I can pass?

Regards,

Ignace

 

From: Desimpel, Ignace [mailto:Ignace.Desimpel@nuance.com]
Sent: maandag 24 februari 2014 11:43
To: user@cassandra.apache.org
Subject: FW: Sporadic gossip exception on add node

 

Had a look at the code, and this might be a race-condition like problem at the function StorageService::checkForEndpointCollision and StorageService::prepareReplacementInfo

 

To do a Gossiper.instance.doShadowRound(), the MessagingService.instance().listen(FBUtilities.getLocalAddress()) must be FULLY (accepting connections) running.

However , the listen function is starting SocketThread threads, but is not waiting for these to be started. So I think, at least in theory,  that the doShadowRound function will be sending messages, thus excepting answers, but there is no guarantee that the listeners are actually up and running.

 

As a test I modified the MessagingService::listen code by

SocketThread th = new SocketThread(ss, "ACCEPT-" + localEp);

   synchronized( th ) {

     th.start();

     try { th.wait(); } catch(Throwable tt){}

   }

 

And the SocketThread::run function

public void run()

   {

     synchronized( this ) {

       this.notifyAll();

     }

 

That way there is little chance the socket thread is not running yet (should be blocked in the server.accept call() ).

 

 

Regards,

Ignace Desimpel

 

From: Desimpel, Ignace [mailto:Ignace.Desimpel@nuance.com]
Sent: donderdag 6 februari 2014 12:15
To: user@cassandra.apache.org
Subject: Sporadic gossip exception on add node

 

Environment : linux, cassandra 2.0.4, 3 node, embedded, byte ordered, LCS

 

When I add a node to the existing 3 node cluster I sometimes get the exception ‘Unable to gossip with any seeds ‘ listed below. If I just restart it without any change then mostly it works. Must be some timing issue.

 

The Cassandra at that time is configured using the Cassandra.yaml file

with the auto_bootstrap set true

and the initial_token set to something like : 00f35256, 041e692a, 0562d8b2, 0930274a, 0b16ce96, 0c5b3e1e, 10cac47a, 12b16bc6, 13f5db4e, 186561aa, 1907996e, 1c32b042, 1e19578e ……

 

The two seeds configured in this yaml are 10.164.8.250 and 10.164.8.249 and these are up and running.

The new node to add has ip 10.164.8.93

 

At the time of the exception, I do not get the gossip message ‘Handshaking version with /10.164.8.93’ on the seeds.

If the exception does not occurs, then I do get that gossip message ‘Handshaking version with /10.164.8.93’ on the seed

 

2014-01-31 13:40:36.380 Loading persisted ring state

2014-01-31 13:40:36.386 Starting Messaging Service on port 9804

2014-01-31 13:40:36.408 Handshaking version with /10.164.8.250

2014-01-31 13:40:36.408 Handshaking version with /10.164.8.249

2014-01-31 13:41:07.415 Exception encountered during startup

java.lang.RuntimeException: Unable to gossip with any seeds

                at org.apache.cassandra.gms.Gossiper.doShadowRound(Gossiper.java:1160) ~[apache-cassandra-2.0.4-SNAPSHOT.jar:2.0.4-SNAPSHOT]

                at org.apache.cassandra.service.StorageService.checkForEndpointCollision(StorageService.java:426) ~[apache-cassandra-2.0.4-SNAPSHOT.jar:2.0.4-SNAPSHOT]

                at org.apache.cassandra.service.StorageService.joinTokenRing(StorageService.java:618) ~[apache-cassandra-2.0.4-SNAPSHOT.jar:2.0.4-SNAPSHOT]

                at org.apache.cassandra.service.StorageService.initServer(StorageService.java:586) ~[apache-cassandra-2.0.4-SNAPSHOT.jar:2.0.4-SNAPSHOT]

                at org.apache.cassandra.service.StorageService.initServer(StorageService.java:485) ~[apache-cassandra-2.0.4-SNAPSHOT.jar:2.0.4-SNAPSHOT]

                at org.apache.cassandra.service.CassandraDaemon.setup(CassandraDaemon.java:346) ~[apache-cassandra-2.0.4-SNAPSHOT.jar:2.0.4-SNAPSHOT]

                at org.apache.cassandra.service.CassandraDaemon.activate(CassandraDaemon.java:461) ~[apache-cassandra-2.0.4-SNAPSHOT.jar:2.0.4-SNAPSHOT]

                at be.landc.services.search.server.db.baseserver.indexsearch.store.cassandra.CassandraStore$CassThread.startUpCassandra(CassandraStore.java:469) [landc-services-search-1.2.0-SNAPSHOT.jar:1.2.0-SNAPSHOT-87937]

                at be.landc.services.search.server.db.baseserver.indexsearch.store.cassandra.CassandraStore$CassThread.run(CassandraStore.java:460) [landc-services-search-1.2.0-SNAPSHOT.jar:1.2.0-SNAPSHOT-87937]

java.lang.RuntimeException: Unable to gossip with any seeds

                at org.apache.cassandra.gms.Gossiper.doShadowRound(Gossiper.java:1160)

                at org.apache.cassandra.service.StorageService.checkForEndpointCollision(StorageService.java:426)

                at org.apache.cassandra.service.StorageService.joinTokenRing(StorageService.java:618)

                at org.apache.cassandra.service.StorageService.initServer(StorageService.java:586)

                at org.apache.cassandra.service.StorageService.initServer(StorageService.java:485)

                at org.apache.cassandra.service.CassandraDaemon.setup(CassandraDaemon.java:346)

                at org.apache.cassandra.service.CassandraDaemon.activate(CassandraDaemon.java:461)

                at be.landc.services.search.server.db.baseserver.indexsearch.store.cassandra.CassandraStore$CassThread.startUpCassandra(CassandraStore.java:469)

                at be.landc.services.search.server.db.baseserver.indexsearch.store.cassandra.CassandraStore$CassThread.run(CassandraStore.java:460)

Exception encountered during startup: Unable to gossip with any seeds

2014-01-31 13:41:07.419 Exception in thread Thread[StorageServiceShutdownHook,5,main]

java.lang.NullPointerException: null

                at org.apache.cassandra.service.StorageService.stopNativeTransport(StorageService.java:349) ~[apache-cassandra-2.0.4-SNAPSHOT.jar:2.0.4-SNAPSHOT]

                at org.apache.cassandra.service.StorageService.shutdownClientServers(StorageService.java:364) ~[apache-cassandra-2.0.4-SNAPSHOT.jar:2.0.4-SNAPSHOT]

                at org.apache.cassandra.service.StorageService.access$3(StorageService.java:361) ~[apache-cassandra-2.0.4-SNAPSHOT.jar:2.0.4-SNAPSHOT]

                at org.apache.cassandra.service.StorageService$1.runMayThrow(StorageService.java:551) ~[apache-cassandra-2.0.4-SNAPSHOT.jar:2.0.4-SNAPSHOT]

                at org.apache.cassandra.utils.WrappedRunnable.run(WrappedRunnable.java:28) ~[apache-cassandra-2.0.4-SNAPSHOT.jar:2.0.4-SNAPSHOT]

                at java.lang.Thread.run(Thread.java:724) ~[na:1.7.0_40]

2014-01-31 13:41:07.420 ShutDownHook requests shutdown on be.landc.services.cdi.server.cassandra.CDIServer@7c32d1a3

2014-01-31 13:41:07.421 Shutdown server request