aurora-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Kevin Sweeney <kevi...@apache.org>
Subject Re: Aurora stuck in LEADER_AWAITING_REGISTRATION
Date Thu, 16 Oct 2014 00:02:59 GMT
The scheduler uses that zookeeper path (or whatever url you pass to
-mesos_msater_address) to find the mesos master is should register with.

Is there a mesos master running with a command-line that looks something
like

exec /usr/local/sbin/mesos-master \
    --zk=zk://$ZK_HOST:2181/mesos/master \
...?

On Wed, Oct 15, 2014 at 4:53 PM, Dobromir Montauk <dobromir@tellapart.com>
wrote:

> Next problem: framework is failing to register.
>
> 2014-10-15 22:15:48,024:28619(0x7f852d259700):ZOO_INFO@check_events@1750:
> session establishment complete on server [127.0.0.1:2181],
> sessionId=0x349110433970e3b, negotiated timeout=10000
> I1015 22:15:48.024787 28763 group.cpp:313] Group process (group(3)@
> 10.88.26.227:34348) connected to ZooKeeper
> I1015 22:15:48.024837 28763 group.cpp:787] Syncing group operations: queue
> size (joins, cancels, datas) = (0, 0, 0)
> I1015 22:15:48.024868 28763 group.cpp:385] Trying to create path
> '/mesos/master' in ZooKeeper
> E1015 22:16:48.013 THREAD132
> org.apache.aurora.scheduler.SchedulerLifecycle$6$3.run: Framework has not
> been registered within the tolerated delay.
> I1015 22:16:48.014 THREAD132
> com.twitter.common.util.StateMachine$Builder$1.execute: SchedulerLifecycle
> state machine transition LEADER_AWAITING_REGISTRATION -> DEAD
>
> My Zookeeper logs have this:
>
> 2014-10-15 22:15:47,831 - INFO  [NIOServerCxn.Factory:
> 0.0.0.0/0.0.0.0:2181:ZooKeeperServer@839] - Client attempting to establish
> new session at /127.0.0.1:41116
> 2014-10-15 22:15:47,836 - INFO  [CommitProcessor:3:ZooKeeperServer@595] -
> Established session 0x349110433970e3a with negotiated timeout 4000 for
> client /127.0.0.1:41116
> 2014-10-15 22:15:48,017 - INFO  [NIOServerCxn.Factory:
> 0.0.0.0/0.0.0.0:2181:NIOServerCnxnFactory@197] - Accepted socket
> connection
> from /127.0.0.1:41117
> 2014-10-15 22:15:48,017 - WARN  [NIOServerCxn.Factory:
> 0.0.0.0/0.0.0.0:2181:ZooKeeperServer@793] - Connection request from old
> client /127.0.0.1:41117; will be dropped if server is in r-o mode
> 2014-10-15 22:15:48,018 - INFO  [NIOServerCxn.Factory:
> 0.0.0.0/0.0.0.0:2181:ZooKeeperServer@839] - Client attempting to establish
> new session at /127.0.0.1:41117
> 2014-10-15 22:15:48,024 - INFO  [CommitProcessor:3:ZooKeeperServer@595] -
> Established session 0x349110433970e3b with negotiated timeout 10000 for
> client /127.0.0.1:41117
> 2014-10-15 22:16:48,435 - WARN  [NIOServerCxn.Factory:
> 0.0.0.0/0.0.0.0:2181:NIOServerCnxn@349] - caught end of stream exception
> EndOfStreamException: Unable to read additional data from client sessionid
> 0x349110433970e39, likely client has closed socket
> at org.apache.zookeeper.server.NIOServerCnxn.doIO(NIOServerCnxn.java:220)
> at
>
> org.apache.zookeeper.server.NIOServerCnxnFactory.run(NIOServerCnxnFactory.java:208)
> at java.lang.Thread.run(Thread.java:745)
> 2014-10-15 22:16:48,436 - INFO  [NIOServerCxn.Factory:
> 0.0.0.0/0.0.0.0:2181:NIOServerCnxn@1001] - Closed socket connection for
> client /127.0.0.1:41114 which had sessionid 0x349110433970e39
>
> I've removed /mesos/master and verified that it gets recreated:
>
> [zk: localhost:2181(CONNECTED) 4] rmr /mesos/master
> [zk: localhost:2181(CONNECTED) 5] ls /mesos/master
> Node does not exist: /mesos/master
> [zk: localhost:2181(CONNECTED) 6] get /mesos/master
> cZxid = 0x10000b717
> ctime = Wed Oct 15 23:34:17 UTC 2014
> mZxid = 0x10000b717
> mtime = Wed Oct 15 23:34:17 UTC 2014
> pZxid = 0x10000b717
> cversion = 0
> dataVersion = 0
> aclVersion = 0
> ephemeralOwner = 0x0
> dataLength = 0
> numChildren = 0
>
> Anyone know what's going wrong?
>
>
> On Wed, Oct 15, 2014 at 2:47 PM, Dobromir Montauk <dobromir@tellapart.com>
> wrote:
>
> > I found instructions that fixed the current issue here:
> >
> > http://wilderness.apache.org/channels/?f=aurora/2014-05-27
> >
> > Would be nice to have a more friendly error message :)
> >
> > On Wed, Oct 15, 2014 at 2:22 PM, Dobromir Montauk <
> dobromir@tellapart.com>
> > wrote:
> >
> >> Hi,
> >>
> >> I've brought up Aurora on my Mesos master node with the following
> command:
> >>
> >> ubuntu@ec2-54-82-17-37:~/$
> >>
> >> GLOG_v=2
> >> LIBPROCESS_PORT=5050
> >> LIBPROCESS_IP=127.0.0.1
> >> AURORA_HOME=/usr/local/aurora-scheduler
> >> DIST_DIR=/home/ubuntu/aurora-scheduler/dist
> >> AURORA_HOME=/usr/local/aurora-scheduler
> >>
> >> sudo /usr/local/aurora-scheduler/bin/aurora-scheduler \
> >>   -cluster_name=tellapart \
> >>   -http_port=8081 \
> >>   -native_log_quorum_size=1 \
> >>   -zk_endpoints=localhost:2181 \
> >>   -mesos_master_address=54.166.50.69:5050,54.160.61.169:5050
> ,localhost:5050
> >> \
> >>   -serverset_path=/aurora/scheduler \
> >>   -native_log_zk_group_path=/aurora/replicated-log \
> >>   -native_log_file_path=$AURORA_HOME/scheduler/db \
> >>   -backup_dir=$AURORA_HOME/scheduler/backups \
> >>   -thermos_executor_path=/dev/null \
> >>   -gc_executor_path=$DIST_DIR/gc_executor.pex \
> >>   -enable_beta_updater=true \
> >>   -vlog=INFO \
> >>   -logtostderr
> >>
> >> Attached is the entire log, but basically I'm seeing this:
> >>
> >> I1015 21:18:05.315263 27634 group.cpp:313] Group process (group(1)@
> >> 10.88.26.227:40393) connected to ZooKeeper
> >> I1015 21:18:05.315322 27634 group.cpp:787] Syncing group operations:
> >> queue size (joins, cancels, datas) = (0, 0, 0)
> >> I1015 21:18:05.315348 27634 group.cpp:385] Trying to create path
> >> '/aurora/replicated-log' in ZooKeeper
> >> I1015 21:18:05.316 THREAD1
> >> com.twitter.common.zookeeper.CandidateImpl$4.onGroupChange: Candidate
> >> /aurora/scheduler/singleton_candidate_0000000008 is now leader of group:
> >> [singleton_candidate_0000000008]
> >> I1015 21:18:05.317 THREAD1
> >> com.twitter.common.util.StateMachine$Builder$1.execute:
> SchedulerLifecycle
> >> state machine transition STORAGE_PREPARED ->
> LEADER_AWAITING_REGISTRATION
> >> I1015 21:18:05.317 THREAD1
> >> org.apache.aurora.scheduler.SchedulerLifecycle$6.execute: Elected as
> >> leading scheduler!
> >> I1015 21:18:05.330394 27634 network.hpp:423] ZooKeeper group memberships
> >> changed
> >> I1015 21:18:05.330660 27639 group.cpp:658] Trying to get
> >> '/aurora/replicated-log/0000000008' in ZooKeeper
> >> I1015 21:18:05.331550 27635 network.hpp:461] ZooKeeper group PIDs: {
> >> log-replica(1)@10.88.26.227:40393 }
> >> I1015 21:18:06.027016 27634 replica.cpp:638] Replica in EMPTY status
> >> received a broadcasted recover request
> >> I1015 21:18:06.027216 27634 recover.cpp:188] Received a recover response
> >> from a replica in EMPTY status
> >> <repeat last 2 message ad nauseum>
> >>
> >> How can I debug what's going on?
> >>
> >> Thanks,
> >> Dobromir
> >>
> >
> >
>

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