Arg, I'm used to seeing the username first... same issue though 4190 instead of 501...

You can follow along on our dev list for full details on the 3.3.0 release. It's still slated for 3/10 (tomorrow). There are some open issues but those are non-blockers. We're having trouble with hudson which is slowing us down. We are def. in the endgame for 3.3.0 and I'm pushing everyone to close it out.

Patrick

Ted Yu wrote:
501 is userid, not pid.
The old pid file was deleted after the unsuccessful kill attempt. Maybe we should keep the pid file if we are unable to kill. Alternatively, we can use ps command to find org.apache.zookeeper.server.quorum.QuorumPeerMain instance.

We use Tanuki wrapper around zk.
zookeeper is started automatically after reboot last night.
See /etc/rc3.d/S20zk-wrapper.sh in attachment

I will report the issue if it happens again.

BTW
There are several unresolved issues for zk 3.3.0
When can we expect zk 3.3.0 to be released ?

Thanks

On Tue, Mar 9, 2010 at 9:56 AM, Patrick Hunt <ph...@apache.org <mailto:ph...@apache.org>> wrote:

    That does seem unusual. Notice the pid of the current jvm is 501,
    while the killed jvm is 6826. Is this 501 jvm created before 6826 or
    has the pid "rolled over". Hard for me to tell... Are you running
    the std zkServer.sh unmodified?

    A shot in the dark: Permission issue? (i.e. zkStart failed to
    overwrite an old/existing pid file due to file permissions). Perhaps
    something started the server twice before stopping?

    Can you reproduce this? If you do please report the issue here:
    https://issues.apache.org/jira/browse/ZOOKEEPER

    Patrick


    Ted Yu wrote:

        Hi,
        We use zookeeper-3.2.1

        When I tested ruok :
        2010-03-09 08:56:36,711 INFO  [NIOServerCxn.Factory:2181]
        server.NIOServerCnxn Processing ruok command from
        /10.10.31.135:57084 <http://10.10.31.135:57084>
        2010-03-09 08:56:36,711 WARN  [NIOServerCxn.Factory:2181]
        server.NIOServerCnxn Exception causing close of session 0x0 due to
        java.io.IOException: Responded to info probe
        2010-03-09 08:56:36,711 INFO  [NIOServerCxn.Factory:2181]
        server.NIOServerCnxn closing session:0x0 NIOServerCnxn:
        java.nio.channels.SocketChannel[connected
        local=/10.10.31.135:2181 <http://10.10.31.135:2181> remote=/
        10.10.31.135:57084 <http://10.10.31.135:57084>]

        [r...@snv-it-lin-010 software]# bin/zkServer.sh stop
        JMX enabled by default
        Using config: /opt/kindsight/zookeeper/software/bin/../conf/zoo.cfg
        Stopping zookeeper ...
        bin/zkServer.sh: line 80: kill: (6826) - No such process
        STOPPED

        I wonder why pidfile held an invalid pid:
        501       4190  0.5  4.9 1523960 814644 ?      Sl   Mar08   3:48
        /usr/local/jdk1.6.0_14/bin/java -Dcom.sun.management.jmxremote
        -Xms100m
        -Xmx1000m -Djava.library.path=lib -classpath
        
lib/wrapper.jar:lib/zabbix_common.jar:zookeeper-3.2.1.jar:lib/log4j-1.2.15.jar:lib/3rdparty/slf4j/jcl104-over-slf4j-1.5.0.jar:lib/3rdparty/slf4j/slf4j-api-1.5.0.jar:lib/3rdparty/slf4j/slf4j-log4j12-1.5.0.jar:lib/3rdparty/ognl/ognl-2.6.9.jar:conf
        -Dwrapper.key=fcnqS87eXW_Tp5xC -Dwrapper.port=32005
        -Dwrapper.jvm.port.min=31000 -Dwrapper.jvm.port.max=31999
        -Dwrapper.pid=4160
        -Dwrapper.version=3.2.3 -Dwrapper.native_library=wrapper
        -Dwrapper.service=TRUE -Dwrapper.cpu.timeout=10 -Dwrapper.jvmid=1
        org.tanukisoftware.wrapper.WrapperSimpleApp
        com.rialto.common.zabbix.ZabbixWrapper
        org.apache.zookeeper.server.quorum.QuorumPeerMain conf/zoo.cfg

        Can someone comment ?

        On Mon, Mar 8, 2010 at 10:25 AM, Patrick Hunt <ph...@apache.org
        <mailto:ph...@apache.org>> wrote:

            Donno if you noticed but ZK 3.3.0 (releasing this month)
            changes that
            message to an info. I'm hoping it reduces these types of
            false positives.
            :-) Not sure why we had it as warn, but that was wrong.

            Patrick


            Jean-Daniel Cryans wrote:

                HBase relies on exception like those when creating a
                znode eg "create
                znode x, if NodeExists then update its value".

                J-D

                On Sun, Mar 7, 2010 at 3:49 AM, Ahmed Suhail Manzoor
                <suhail...@gmail.com <mailto:suhail...@gmail.com>> wrote:

                    Hi,

                    I am hoping someone can help me with a zookeeper
                    setup. I am running
                    Zookeeper wih export HBASE_MANAGES_ZK=false and
                    start it using bin/hbase
                    zookeeper. It starts up fine. However when I start
                    up Hbase I get the
                    following in my log file:

                    10/03/07 11:41:35 WARN server.PrepRequestProcessor:
                    Got exception when
                    processing sessionid:0x127386d00840000 type:create
                    cxid:0x12
                    zxid:0xfffffffffffffffe txntype:unknown n/a
                    org.apache.zookeeper.KeeperException$NodeExistsException:
                    KeeperErrorCode
                    =
                    NodeExists
                         at

                    
org.apache.zookeeper.server.PrepRequestProcessor.pRequest(PrepRequestProcessor.java:245)
                         at

                    
org.apache.zookeeper.server.PrepRequestProcessor.run(PrepRequestProcessor.java:114)

                    I know its only a warning but I would like someone
                    to shed some light on
                    it
                    so I can be confident that my installation is
                    correct. Thank you kindly
                    for
                    your help.

                    Cheers
                    su./hail





Reply via email to