Home Forums HDP on Linux – Installation HMC Add Node Fail==>Node Discovery and Preparation Failed

This topic contains 3 replies, has 3 voices, and was last updated by  Larry Liu 1 year, 9 months ago.

  • Creator
    Topic
  • #13123

    HMC Add Node Fail

    Bug:
    During “Node Discovery and Preparation” stage in the HMC tool,
    “Preparing discovred nodes” failed as the follwling
    (1)Finding reachable nodes: All 4 nodes succeeded
    (2)Obtaining information about reachable nodes: All 4 nodes succeeded
    (3)Verifying and updating node information: All 4 nodes succeeded
    (4)Preparing discovered nodes: 1 / 4 in progress, 3 succeeded

    “Node Discovery and Preparation” stage seems very unstatble,
    I ever tried as many as 5 times to get 1 successful job.
    Failed with no special reason,
    and Success with no special reason.
    It depends on the” good luck”
    I just wonder when can I get a successful or failed cluster deployment

    Environments:
    (1)OS : CentOS6.3
    (2)Hadoop : HDP1.1.1.16 + HDP-UTILS-1.1.0.15
    (3)Cluster : 4 Nodes

    Confirmed Items:
    FQDN ok
    /etc/hosts ok
    iptables stop ok
    SSH keyless ok
    # yum repolist failed

    Clues:
    /var/run/hmc/shark6/cluster/shark6/bootstrapNodes/host002.dmo.com.err
    host001.dmo.com.err 0 byte
    host003.dmo.com.err 0 byte
    host004.dmo.com.err 0 byte
    host005.dmo.com.err 0 byte
    host002.dmo.com.err 47K bytes

    host002.dmo.com.err
    Existing lock /var/run/yum.pid: another copy is running as pid 21564.
    Another app is currently holding the yum lock; waiting for it to exit…
    The other application is: PackageKit
    Memory : 39 M RSS (343 MB VSZ)
    Started: Fri Dec 28 12:50:01 2012 – 2:45:48 ago
    State : Sleeping, pid: 21564
    Another app is currently holding the yum lock; waiting for it to exit…
    The other application is: PackageKit
    Memory : 39 M RSS (343 MB VSZ)
    Started: Fri Dec 28 12:50:01 2012 – 2:45:50 ago
    State : Sleeping, pid: 21564
    …………………………………………………………

    I wonder why” yum lock” happened after another 3 nodes had successfully passed the check , does it mean that it’s too busy for puppet or yum install the HDP package on 4 nodes ? Did Any one has the same problem ? Any help will be GREATLY appreciated..

    Regards,
    Jeff

Viewing 3 replies - 1 through 3 (of 3 total)

You must be logged in to reply to this topic.

  • Author
    Replies
  • #13185

    Larry Liu
    Moderator

    Hi, Jeff

    Thanks for letting us know.

    Have a great 2013.

    Thanks

    Larry

    Collapse
    #13178

    Hi tedr :

    Thanks for your info.

    After “rm -rf /var/run/yum.pid “, it can work now.

    Thanks.

    Jeff

    Collapse
    #13125

    tedr
    Member

    Hi Jeff,

    Thanks for using HDP.

    Actually this is not a bug, the yum lock means that HMC(ambari) was trying to use yum on that particular host and could not because that host thinks that another process was already using it. You can check to see if yum is running by running the following command: “ps -aux | grep yum” If there are any processes in the list other than “grep yum” you can stop them with finding the process id in the list and issuing the command “kill -9 ” Then retry the HDP installer.

    I hope this helps,
    Ted.

    Collapse
Viewing 3 replies - 1 through 3 (of 3 total)