zookeeper-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Martin Gainty <mgai...@hotmail.com>
Subject Re: IllegalStateException: instance must be started before calling this method
Date Tue, 17 Apr 2018 18:12:53 GMT
too many variables so if we can knock this down to:

No Docker (run at known command line and let ZK select container )

JDK 1.8 (im hearing squirrely behaviour by JDK 1.9 with MR jars..lets use 1.9 MR jars after
we verify 1.8 ok)

.hosts file for Host Container Resolver (we dont know if swarm/kubernetes may be causing issues)
e.g:

# .hosts file for the website
10.17.46.142 OutOfReachHost

this error:
2018-02-15T09:58:12.094+0100; ERROR; WSOSTSLXWIT01/MANAGER; P3424/T194; [SPACE/LearnerHandler-/10.17.46.142:49336/LearnerHandler<http://10.17.46.142:49336/LearnerHandler>];
Unexpected exception causing shutdown while sock still open
java.net<http://java.net<http://java.net/>>.SocketTimeoutException: Read timed
out
at java.net.SocketInputStream.socketRead0(Native Method)


smacks of either:
destination unreachable

some intervening node (proxy/reverseproxy) is slowing down socket read
so we will need to set the sessionTimeout and clientTimeout to an insanely high value such
as 60 sec


* @param zkServers
        * The Zookeeper servers
        * @param sessionTimeout
        * The session timeout in milli seconds
        * @param connectionTimeout
        * The connection timeout in milli seconds
        * @param zkSerializer
        * The Zookeeper data serializer
        * @param operationRetryTimeout
        * Most operations done through this {@link org.I0Itec.zkclient.ZkClient} are retried
in cases like
        * connection loss with the Zookeeper servers. During such failures, this
        * <code>operationRetryTimeout</code> decides the maximum amount of time,
in milli seconds, each
        * operation is retried. A value lesser than 0 is considered as
        * "retry forever until a connection has been reestablished".


// Set sessionTimeOut and connectionTimeOut to insanely high value such as 60 sec
//set retry to forever

   final Integer sessionTimeOut= 60000;
   final Integer connectionTimeOut= 60000;
   final Integer forever=0;
   org.I0Itec.zkclient.ZkClient zkClient = new org.I0Itec.zkclient.ZkClient("OutOfReachHost:49336",sessionTimeout,
connectionTimeOut, new org.I0Itec.zkclient.serialize.ZkSerializerSerializableSerializer(),
forever);

if you increment sessionTimeOut and connectionTimeOut to 60sec will ZKClient connect ?

if bumping up sessionTimeOut and/or connectionTimeOut doesnt work then from Client try
>tracert OutOfReachHost


ask your net admin what are the intervening servers between
you
and OutOfReachHost


If one of intervening server hosts is proxy-server
very likely that proxy server that is slowing down your request in which case try using curl
to bypass proxy

curl  -v --noproxy '*' -XGET "OutOfReachHost:4711"


?
Martin

______________________________________________


________________________________
From: VALLASTER Stefan <Stefan.VALLASTER@frequentis.com>
Sent: Tuesday, April 17, 2018 8:59 AM
To: user@zookeeper.apache.org
Subject: RE: IllegalStateException: instance must be started before calling this method

Hello Martin,

did you read the input from Yuval and me. Does this answer your question?

Best regards
Stefan

-----Original Message-----
From: Martin Gainty [mailto:mgainty@hotmail.com]
Sent: Dienstag, 10. April 2018 13:40
To: user@zookeeper.apache.org
Subject: Re: IllegalStateException: instance must be started before calling this method

MG>waiting for Yuval to answer:


in docker  container?
if so how do you resolve hosts (.hosts file, kubernetes, swarm)?


MG>?

________________________________
From: VALLASTER Stefan <Stefan.VALLASTER@frequentis.com>
Sent: Tuesday, April 10, 2018 4:46 AM
To: user@zookeeper.apache.org
Subject: RE: IllegalStateException: instance must be started before calling this method

How can we get this topic forward again? It seems the mail thread is currently stalled for
no obvious reason. Anyone willing to help us out here again?

Best regards
Stefan

-----Original Message-----
From: Martin Gainty [mailto:mgainty@hotmail.com]
Sent: Sonntag, 25. März 2018 22:58
To: user@zookeeper.apache.org
Subject: Re: IllegalStateException: instance must be started before calling this method



in docker  container?

if so how do you resolve hosts (.hosts file, kubernetes, swarm)?

M
________________________________
From: Yuval Dori <yuval.dori@gigaspaces.com>
Sent: Sunday, March 25, 2018 5:34 AM
To: user@zookeeper.apache.org
Cc: Yuval Dori
Subject: IllegalStateException: instance must be started before calling this method

Hi,

Our customer is using ZK 3.4.3 and got IllegalStateException: instance must be started before
calling this method


with the following stack trace:
2018-02-15T09:58:12.094+0100; ERROR; WSOSTSLXWIT01/MANAGER; P3424/T194; [SPACE/LearnerHandler-/10.17.46.142:49336/LearnerHandler<http://10.17.46.142:49336/LearnerHandler>];
Unexpected exception causing shutdown while sock still open
java.net<http://java.net>.SocketTimeoutException: Read timed out
at java.net.SocketInputStream.socketRead0(Native Method)
at java.net.SocketInputStream.socketRead(SocketInputStream.java:116)
at java.net.SocketInputStream.read(SocketInputStream.java:171)
at java.net.SocketInputStream.read(SocketInputStream.java:141)
at java.io.BufferedInputStream.fill(BufferedInputStream.java:246)
at java.io.BufferedInputStream.read(BufferedInputStream.java:265)
at java.io.DataInputStream.readInt(DataInputStream.java:387)
at org.apache.jute.BinaryInputArchive.readInt(BinaryInputArchive.java:63)
at org.apache.zookeeper.server.quorum.QuorumPacket.deserialize(QuorumPacket.java:83)
at org.apache.jute.BinaryInputArchive.readRecord(BinaryInputArchive.java:99)
at org.apache.zookeeper.server.quorum.LearnerHandler.run(LearnerHandler.java:542)


This is the description:

At first ps-main.1 is started as primary on WSOSTSLXBMS01 GSC 3.

24.01.2018 11:31:04,947 +01:00 INFO WSOSTSLXBMS01/GSC/3 P1280/T120 SPACE/Curator-LeaderSelector-0/1
Space instance is Primary

After some relocation of ps-main instances a new backup instance is started on the same GSC.

24.01.2018 11:32:59,886 +01:00 INFO WSOSTSLXBMS01/GSC/3 P1280/T146 SPACE/GS-LRMI Connection-pool-1-thread-6/1
Space instance [ frqMain_container1:frqMain ] has been selected as Backup

Later a failover test is performed by shutting down WSOSTSLXBMS02 and WSOSTSLXMSS02.
We would expect that ps-main.1 backup will be elected as primary because the former primary
was located at WSOSTSLXBMS02 but no election happend. In the end in the whole cluster there
was no primary of ps-main.1 running.

We found suspicious exception in the logs:
24.01.2018 11:32:44,987 +01:00 INFO WSOSTSLXBMS01/GSC/3 P1280/T120 SPACE/Curator-LeaderSelector-0/frqMain_container1:frqMain
Lost leadership [sessionId=0x36128253b820001]
24.01.2018 11:32:44,987 +01:00 INFO WSOSTSLXBMS01/GSC/3 P1280/T98 SPACE/Curator-Framework-0/CuratorFrameworkImpl
backgroundOperationsLoop exiting
24.01.2018 11:32:44,990 +01:00 INFO WSOSTSLXBMS01/GSC/3 P1280/T120 SPACE/Curator-LeaderSelector-0/1
Quiesce state set to SUSPENDED
24.01.2018 11:32:44,991 +01:00 INFO WSOSTSLXBMS01/GSC/3 P1280/T24 SPACE/GS-LRMI Connection-pool-1-thread-1/ZooKeeper
Session: 0x36128253b820001 closed
24.01.2018 11:32:44,991 +01:00 ERROR WSOSTSLXBMS01/GSC/3 P1280/T120 SPACE/Curator-LeaderSelector-0/LeaderSelector
The leader threw an exception
24.01.2018 11:32:44,991 +01:00 ERROR WSOSTSLXBMS01/GSC/3 P1280/T120 SPACE/Curator-LeaderSelector-0/LeaderSelector
java.lang.IllegalStateException: instance must be started before calling this method
24.01.2018 11:32:44,991 +01:00 ERROR WSOSTSLXBMS01/GSC/3 P1280/T120 SPACE/Curator-LeaderSelector-0/LeaderSelector
at org.apache.curator.shaded.com<http://org.apache.curator.shaded.com>.google.common.base.Preconditions.checkState(Preconditions.java:176)
24.01.2018 11:32:44,991 +01:00 ERROR WSOSTSLXBMS01/GSC/3 P1280/T120 SPACE/Curator-LeaderSelector-0/LeaderSelector
at org.apache.curator.framework.imps.CuratorFrameworkImpl.delete(CuratorFrameworkImpl.java:359)
24.01.2018 11:32:44,991 +01:00 ERROR WSOSTSLXBMS01/GSC/3 P1280/T120 SPACE/Curator-LeaderSelector-0/LeaderSelector
at org.apache.curator.framework.recipes.locks.LockInternals.deleteOurPath(LockInternals.java:339)
24.01.2018 11:32:44,991 +01:00 ERROR WSOSTSLXBMS01/GSC/3 P1280/T120 SPACE/Curator-LeaderSelector-0/LeaderSelector
at org.apache.curator.framework.recipes.locks.LockInternals.releaseLock(LockInternals.java:123)
24.01.2018 11:32:44,991 +01:00 ERROR WSOSTSLXBMS01/GSC/3 P1280/T120 SPACE/Curator-LeaderSelector-0/LeaderSelector
at org.apache.curator.framework.recipes.locks.InterProcessMutex.release(InterProcessMutex.java:154)
24.01.2018 11:32:44,991 +01:00 ERROR WSOSTSLXBMS01/GSC/3 P1280/T120 SPACE/Curator-LeaderSelector-0/LeaderSelector
at org.apache.curator.framework.recipes.leader.LeaderSelector.doWork(LeaderSelector.java:449)
24.01.2018 11:32:44,991 +01:00 ERROR WSOSTSLXBMS01/GSC/3 P1280/T120 SPACE/Curator-LeaderSelector-0/LeaderSelector
at org.apache.curator.framework.recipes.leader.LeaderSelector.doWorkLoop(LeaderSelector.java:466)
24.01.2018 11:32:44,991 +01:00 ERROR WSOSTSLXBMS01/GSC/3 P1280/T120 SPACE/Curator-LeaderSelector-0/LeaderSelector
at org.apache.curator.framework.recipes.leader.LeaderSelector.access$100(LeaderSelector.java:65)
24.01.2018 11:32:44,991 +01:00 ERROR WSOSTSLXBMS01/GSC/3 P1280/T120 SPACE/Curator-LeaderSelector-0/LeaderSelector
at org.apache.curator.framework.recipes.leader.LeaderSelector$2.call(LeaderSelector.java:246)
24.01.2018 11:32:44,991 +01:00 ERROR WSOSTSLXBMS01/GSC/3 P1280/T120 SPACE/Curator-LeaderSelector-0/LeaderSelector
at org.apache.curator.framework.recipes.leader.LeaderSelector$2.call(LeaderSelector.java:240)
24.01.2018 11:32:44,991 +01:00 ERROR WSOSTSLXBMS01/GSC/3 P1280/T120 SPACE/Curator-LeaderSelector-0/LeaderSelector
at java.util.concurrent.FutureTask.run(FutureTask.java:266)
24.01.2018 11:32:44,991 +01:00 ERROR WSOSTSLXBMS01/GSC/3 P1280/T120 SPACE/Curator-LeaderSelector-0/LeaderSelector
at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
24.01.2018 11:32:44,991 +01:00 ERROR WSOSTSLXBMS01/GSC/3 P1280/T120 SPACE/Curator-LeaderSelector-0/LeaderSelector
at java.util.concurrent.FutureTask.run(FutureTask.java:266)
24.01.2018 11:32:44,991 +01:00 ERROR WSOSTSLXBMS01/GSC/3 P1280/T120 SPACE/Curator-LeaderSelector-0/LeaderSelector
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
24.01.2018 11:32:44,991 +01:00 ERROR WSOSTSLXBMS01/GSC/3 P1280/T120 SPACE/Curator-LeaderSelector-0/LeaderSelector
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
24.01.2018 11:32:44,991 +01:00 ERROR WSOSTSLXBMS01/GSC/3 P1280/T120 SPACE/Curator-LeaderSelector-0/LeaderSelector
at java.lang.Thread.run(Thread.java:748)
24.01.2018 11:32:44,991 +01:00 ERROR WSOSTSLXBMS01/GSC/3 P1280/T120 SPACE/Curator-LeaderSelector-0/LeaderSelector
24.01.2018 11:32:44,991 +01:00 INFO WSOSTSLXBMS01/GSC/3 P1280/T97 SPACE/GS-LRMI Connection-pool-1-thread-1-EventThread/ClientCnxn
EventThread shut down for session: 0x36128253b820001
24.01.2018 11:32:44,997 +01:00 INFO WSOSTSLXBMS01/GSC/3 P1280/T24 SPACE/GS-LRMI Connection-pool-1-thread-1/frqMain1_1
Channel is closed [target=frqMain1_1, target url=jini://*/frqMain_container1_1/frqMain?groups=POLBAY&ignoreValidation=true&schema=persistent&id=1&total_members=4,1&versioned=true&mirror=true&cluster_schema=partitioned&locators=WSOSTSLXMSS01,WSOSTSLXMSS02&state=started&timeout=5000,
target machine connection url=NIO://WSOSTSLXBMS02:11002/pid[3756]/24795468496_3_2538032931939568226_details[class
com.gigaspaces.internal.cluster.node.impl.router.AbstractConnectionProxyBasedReplicationRouter$ConnectionEndpoint(frqMain_container1_1:frqMain)]]
24.01.2018 11:32:44,998 +01:00 INFO WSOSTSLXBMS01/GSC/3 P1280/T24 SPACE/GS-LRMI Connection-pool-1-thread-1/ps-main-mirror-service
Channel is closed [target=ps-main-mirror-service, target url=jini://*/ps-main-mirror-service_container/ps-main-mirror-service?schema=persistent&id=1&total_members=4,1&versioned=true&mirror=true&groups=POLBAY&cluster_schema=partitioned&locators=WSOSTSLXMSS01,WSOSTSLXMSS02&state=started&timeout=5000,
target machine connection url=NIO://WSOSTSLXBMS01:11002/pid[3220]/25906827215_3_7196393537085178059_details[class
com.gigaspaces.internal.cluster.node.impl.router.AbstractConnectionProxyBasedReplicationRouter$ConnectionEndpoint(ps-main-mirror-service_container:ps-main-mirror-service)]]
24.01.2018 11:32:45,767 +01:00 INFO WSOSTSLXBMS01/GSC/3 P1280/T24 SPACE/GS-LRMI Connection-pool-1-thread-1/1
Shutdown complete
24.01.2018 11:32:45,768 +01:00 INFO WSOSTSLXBMS01/GSC/3 P1280/T24 SPACE/GS-LRMI Connection-pool-1-thread-1/container
Container <frqMain_container1> shutdown completed

24.01.2018 11:32:46,037 +01:00 WARN WSOSTSLXBMS01/GSC/3 P1280/T24 SPACE/GS-LRMI Connection-pool-1-thread-1/PUServiceBeanImpl
Failed to delete deployed processing unit from [C:\Program Files\Frequentis\Frequentis LifeX
Platform\gigaspaces\work\processing-units\ps-main_1_2008048111]
24.01.2018 11:32:46,037 +01:00 INFO WSOSTSLXBMS01/GSC/3 P1280/T24 SPACE/GS-LRMI Connection-pool-1-thread-1/PUServiceBeanImpl
Stopped

Attached thread dump. Failover happend around 24.01.2018 11:40:20,893 +01:00
WARN WSOSTSLXBMS01/GSC/3 P1280/T59 SPACE/GS-Watchdog/watchdog The ServerEndPoint [WSOSTSLXMSS02/10.17.46.142:11001<http://10.17.46.142:11001/>]
is not reachable or is reachable but with no matching invocation in progress at the server
peer, closing invalid connection with local address [/10.17.46.136:49864<http://10.17.46.136:49864/>]
[ remote invocation of: net.jini.lookup.DiscoveryAdmin.getMemberGroups] [java.net<http://java.net>.SocketTimeoutException]


Could it be ZK bug?

Thanks,

Yuval



Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message