FAQ
Hello Guys,

I'm starting the setup of my hadoop cluster, the lab has 4 virtual machines
in the virtual box and CM with orchestrator, I think that has stopped
because of a bug (CM4.03 Free) or in the agents of same version.

The Hosts menu showed all of my nodes with integraty ok and green, but if I
attempt to add a cluster and use only currently managed hosts the message
"Server Error: No hosts found" is displayed.

On the google nothing about that is returned, my /etc/hosts and the local
dns zone is set.

CM:

[root@cm ~]# host node1.hdpcore.net
node1.hdpcore.net has address 192.168.137.3

[root@cm ~]# host 192.168.137.3
3.137.168.192.in-addr.arpa domain name pointer node1.hdpcore.net.

Node:

[root@node1 ~]# host cm.hdpcore.net
cm.hdpcore.net has address 192.168.137.2

[root@node1 ~]# host 192.168.137.2
2.137.168.192.in-addr.arpa domain name pointer ns1.hdpcore.net.

Thanks for any help,

Kind regards
Ricardo A. Reis

Search Discussions

  • Manoj Babu at Jul 23, 2012 at 6:43 pm
    Hi Ricardo,

    Ensure that /etc/hosts holds all the newly added nodes details.

    Thanks.
    On Tue, Jul 24, 2012 at 12:03 AM, Ricardo Alves dos Reis wrote:

    Hello Guys,

    I'm starting the setup of my hadoop cluster, the lab has 4 virtual
    machines in the virtual box and CM with orchestrator, I think that has
    stopped because of a bug (CM4.03 Free) or in the agents of same version.

    The Hosts menu showed all of my nodes with integraty ok and green, but if
    I attempt to add a cluster and use only currently managed hosts the message
    "Server Error: No hosts found" is displayed.

    On the google nothing about that is returned, my /etc/hosts and the local
    dns zone is set.

    CM:

    [root@cm ~]# host node1.hdpcore.net
    node1.hdpcore.net has address 192.168.137.3

    [root@cm ~]# host 192.168.137.3
    3.137.168.192.in-addr.arpa domain name pointer node1.hdpcore.net.

    Node:

    [root@node1 ~]# host cm.hdpcore.net
    cm.hdpcore.net has address 192.168.137.2

    [root@node1 ~]# host 192.168.137.2
    2.137.168.192.in-addr.arpa domain name pointer ns1.hdpcore.net.

    Thanks for any help,

    Kind regards
    Ricardo A. Reis
  • Ricardo A. Reis at Jul 23, 2012 at 6:56 pm
    Hi Manoj,

    This is my /etc/hosts "fake fqdn for lab" . You see any problem?

    [root@node1 ~]# cat /etc/hosts
    127.0.0.1 localhost localhost.localdomain localhost4 localhost4.localdomain4
    ::1 localhost localhost.localdomain localhost6 localhost6.localdomain6
    192.168.137.2 cm.hdpcore.net cm
    192.168.137.3 node1.hdpcore.net node1
    192.168.137.4 node2.hdpcore.net node2
    192.168.137.5 node3.hdpcore.net node3

    [root@cm ~]# cat /etc/hosts
    127.0.0.1 localhost localhost.localdomain localhost4 localhost4.localdomain4
    ::1 localhost localhost.localdomain localhost6 localhost6.localdomain6
    192.168.137.2 cm.hdpcore.net cm
    192.168.137.3 node1.hdpcore.net node1
    192.168.137.4 node2.hdpcore.net node2
    192.168.137.5 node3.hdpcore.net node3

    Thanks
    --
    Ricardo A. Reis
    On Mon, Jul 23, 2012 at 3:43 PM, Manoj Babu wrote:

    Hi Ricardo,

    Ensure that /etc/hosts holds all the newly added nodes details.

    Thanks.


    On Tue, Jul 24, 2012 at 12:03 AM, Ricardo Alves dos Reis <
    ricardo.areis@gmail.com> wrote:
    Hello Guys,

    I'm starting the setup of my hadoop cluster, the lab has 4 virtual
    machines in the virtual box and CM with orchestrator, I think that has
    stopped because of a bug (CM4.03 Free) or in the agents of same version.

    The Hosts menu showed all of my nodes with integraty ok and green, but if
    I attempt to add a cluster and use only currently managed hosts the message
    "Server Error: No hosts found" is displayed.

    On the google nothing about that is returned, my /etc/hosts and the local
    dns zone is set.

    CM:

    [root@cm ~]# host node1.hdpcore.net
    node1.hdpcore.net has address 192.168.137.3

    [root@cm ~]# host 192.168.137.3
    3.137.168.192.in-addr.arpa domain name pointer node1.hdpcore.net.

    Node:

    [root@node1 ~]# host cm.hdpcore.net
    cm.hdpcore.net has address 192.168.137.2

    [root@node1 ~]# host 192.168.137.2
    2.137.168.192.in-addr.arpa domain name pointer ns1.hdpcore.net.

    Thanks for any help,

    Kind regards
    Ricardo A. Reis
  • Mark Schnegelberger at Jul 23, 2012 at 7:18 pm
    Hi Ricardo,

    Did you install Cloudera Manager using the automated .bin installer, or did
    you go the manual route by installing the "cloudera-manager-agent" packages
    manually on each node? If you installed the agents manually, is your
    /etc/cloudera-scm-agent/config.ini file configured properly with the
    settings for your environment, which might look something like:

    ----
    [General]
    # Hostname of Cloudera SCM Server
    server_host=cm.hdpcore.net

    # Port that server is listening on
    server_port=7182
    ----

    If all the above are in place and correct, do you have any errors in
    /var/log/cloudera-scm-agent/cloudera-scm-agent.log on one of the nodes? If
    so, perhaps you could provide a brief snippit here or via pastebin?

    Thanks,

    On Mon, Jul 23, 2012 at 2:55 PM, Ricardo A. Reis wrote:

    Hi Manoj,

    This is my /etc/hosts "fake fqdn for lab" . You see any problem?

    [root@node1 ~]# cat /etc/hosts
    127.0.0.1 localhost localhost.localdomain localhost4
    localhost4.localdomain4
    ::1 localhost localhost.localdomain localhost6 localhost6.localdomain6
    192.168.137.2 cm.hdpcore.net cm
    192.168.137.3 node1.hdpcore.net node1
    192.168.137.4 node2.hdpcore.net node2
    192.168.137.5 node3.hdpcore.net node3

    [root@cm ~]# cat /etc/hosts
    127.0.0.1 localhost localhost.localdomain localhost4
    localhost4.localdomain4
    ::1 localhost localhost.localdomain localhost6 localhost6.localdomain6
    192.168.137.2 cm.hdpcore.net cm
    192.168.137.3 node1.hdpcore.net node1
    192.168.137.4 node2.hdpcore.net node2
    192.168.137.5 node3.hdpcore.net node3

    Thanks
    --
    Ricardo A. Reis
    On Mon, Jul 23, 2012 at 3:43 PM, Manoj Babu wrote:

    Hi Ricardo,

    Ensure that /etc/hosts holds all the newly added nodes details.

    Thanks.


    On Tue, Jul 24, 2012 at 12:03 AM, Ricardo Alves dos Reis <
    ricardo.areis@gmail.com> wrote:
    Hello Guys,

    I'm starting the setup of my hadoop cluster, the lab has 4 virtual
    machines in the virtual box and CM with orchestrator, I think that has
    stopped because of a bug (CM4.03 Free) or in the agents of same version.

    The Hosts menu showed all of my nodes with integraty ok and green, but
    if I attempt to add a cluster and use only currently managed hosts the
    message "Server Error: No hosts found" is displayed.

    On the google nothing about that is returned, my /etc/hosts and the
    local dns zone is set.

    CM:

    [root@cm ~]# host node1.hdpcore.net
    node1.hdpcore.net has address 192.168.137.3

    [root@cm ~]# host 192.168.137.3
    3.137.168.192.in-addr.arpa domain name pointer node1.hdpcore.net.

    Node:

    [root@node1 ~]# host cm.hdpcore.net
    cm.hdpcore.net has address 192.168.137.2

    [root@node1 ~]# host 192.168.137.2
    2.137.168.192.in-addr.arpa domain name pointer ns1.hdpcore.net.

    Thanks for any help,

    Kind regards
    Ricardo A. Reis

    --
    Mark Schnegelberger
    Customer Operations Engineer
    Cloudera
  • Manoj Babu at Jul 23, 2012 at 7:21 pm
    Hi Ricardo,

    This link may help you.

    https://groups.google.com/a/cloudera.org/forum/?fromgroups#!topic/scm-users/wQfG8RVXH6M


    Cheers!
    Manoj.



    On Tue, Jul 24, 2012 at 12:25 AM, Ricardo A. Reis
    wrote:
    Hi Manoj,

    This is my /etc/hosts "fake fqdn for lab" . You see any problem?

    [root@node1 ~]# cat /etc/hosts
    127.0.0.1 localhost localhost.localdomain localhost4
    localhost4.localdomain4
    ::1 localhost localhost.localdomain localhost6 localhost6.localdomain6
    192.168.137.2 cm.hdpcore.net cm
    192.168.137.3 node1.hdpcore.net node1
    192.168.137.4 node2.hdpcore.net node2
    192.168.137.5 node3.hdpcore.net node3

    [root@cm ~]# cat /etc/hosts
    127.0.0.1 localhost localhost.localdomain localhost4
    localhost4.localdomain4
    ::1 localhost localhost.localdomain localhost6 localhost6.localdomain6
    192.168.137.2 cm.hdpcore.net cm
    192.168.137.3 node1.hdpcore.net node1
    192.168.137.4 node2.hdpcore.net node2
    192.168.137.5 node3.hdpcore.net node3

    Thanks
    --
    Ricardo A. Reis
    On Mon, Jul 23, 2012 at 3:43 PM, Manoj Babu wrote:

    Hi Ricardo,

    Ensure that /etc/hosts holds all the newly added nodes details.

    Thanks.


    On Tue, Jul 24, 2012 at 12:03 AM, Ricardo Alves dos Reis <
    ricardo.areis@gmail.com> wrote:
    Hello Guys,

    I'm starting the setup of my hadoop cluster, the lab has 4 virtual
    machines in the virtual box and CM with orchestrator, I think that has
    stopped because of a bug (CM4.03 Free) or in the agents of same version.

    The Hosts menu showed all of my nodes with integraty ok and green, but
    if I attempt to add a cluster and use only currently managed hosts the
    message "Server Error: No hosts found" is displayed.

    On the google nothing about that is returned, my /etc/hosts and the
    local dns zone is set.

    CM:

    [root@cm ~]# host node1.hdpcore.net
    node1.hdpcore.net has address 192.168.137.3

    [root@cm ~]# host 192.168.137.3
    3.137.168.192.in-addr.arpa domain name pointer node1.hdpcore.net.

    Node:

    [root@node1 ~]# host cm.hdpcore.net
    cm.hdpcore.net has address 192.168.137.2

    [root@node1 ~]# host 192.168.137.2
    2.137.168.192.in-addr.arpa domain name pointer ns1.hdpcore.net.

    Thanks for any help,

    Kind regards
    Ricardo A. Reis
  • Ricardo A. Reis at Jul 23, 2012 at 8:09 pm
    Hi Manoj,

    I used default installer and config.ini seems correct, but on agent have
    this error log.

    The lvm is to facilitate the lab, though the production will not use.

    [23/Jul/2012 16:44:14 +0000] 7169 MainThread throttling_logger ERROR Failed
    to collect disk stats for device 'mapper/vg_ambari-lv_root'
    Traceback (most recent call last):
    File "/usr/lib64/cmf/agent/src/cmf/monitor/host/disk_device.py", line 54,
    in collect_update
    stat_file = open(self._stat_path)
    IOError: [Errno 2] No such file or directory:
    '/sys/block/mapper/vg_ambari-lv_root/stat'
    [23/Jul/2012 16:44:14 +0000] 7169 MainThread disk_devices ERROR Failed to
    add DiskDevice for disk='mapper/vg_ambari-lv_root'

    http://pastebin.com/Vj4xWJMh

    Note: This is an old vm template used to test the ambari project with a
    snapshot rollback to a clean lab.

    Thanks
    --
    Ricardo A. Reis
    On Mon, Jul 23, 2012 at 4:20 PM, Manoj Babu wrote:

    Hi Ricardo,

    This link may help you.


    https://groups.google.com/a/cloudera.org/forum/?fromgroups#!topic/scm-users/wQfG8RVXH6M


    Cheers!
    Manoj.



    On Tue, Jul 24, 2012 at 12:25 AM, Ricardo A. Reis <ricardo.areis@gmail.com
    wrote:
    Hi Manoj,

    This is my /etc/hosts "fake fqdn for lab" . You see any problem?

    [root@node1 ~]# cat /etc/hosts
    127.0.0.1 localhost localhost.localdomain localhost4
    localhost4.localdomain4
    ::1 localhost localhost.localdomain localhost6 localhost6.localdomain6
    192.168.137.2 cm.hdpcore.net cm
    192.168.137.3 node1.hdpcore.net node1
    192.168.137.4 node2.hdpcore.net node2
    192.168.137.5 node3.hdpcore.net node3

    [root@cm ~]# cat /etc/hosts
    127.0.0.1 localhost localhost.localdomain localhost4
    localhost4.localdomain4
    ::1 localhost localhost.localdomain localhost6 localhost6.localdomain6
    192.168.137.2 cm.hdpcore.net cm
    192.168.137.3 node1.hdpcore.net node1
    192.168.137.4 node2.hdpcore.net node2
    192.168.137.5 node3.hdpcore.net node3

    Thanks
    --
    Ricardo A. Reis
    On Mon, Jul 23, 2012 at 3:43 PM, Manoj Babu wrote:

    Hi Ricardo,

    Ensure that /etc/hosts holds all the newly added nodes details.

    Thanks.


    On Tue, Jul 24, 2012 at 12:03 AM, Ricardo Alves dos Reis <
    ricardo.areis@gmail.com> wrote:
    Hello Guys,

    I'm starting the setup of my hadoop cluster, the lab has 4 virtual
    machines in the virtual box and CM with orchestrator, I think that has
    stopped because of a bug (CM4.03 Free) or in the agents of same version.

    The Hosts menu showed all of my nodes with integraty ok and green, but
    if I attempt to add a cluster and use only currently managed hosts the
    message "Server Error: No hosts found" is displayed.

    On the google nothing about that is returned, my /etc/hosts and the
    local dns zone is set.

    CM:

    [root@cm ~]# host node1.hdpcore.net
    node1.hdpcore.net has address 192.168.137.3

    [root@cm ~]# host 192.168.137.3
    3.137.168.192.in-addr.arpa domain name pointer node1.hdpcore.net.

    Node:

    [root@node1 ~]# host cm.hdpcore.net
    cm.hdpcore.net has address 192.168.137.2

    [root@node1 ~]# host 192.168.137.2
    2.137.168.192.in-addr.arpa domain name pointer ns1.hdpcore.net.

    Thanks for any help,

    Kind regards
    Ricardo A. Reis
  • Ricardo A. Reis at Jul 27, 2012 at 10:22 pm
    Philip, Manoj and Mark.

    Thank for answers, I had another repository configured that provided other
    version of hive non-cdh4.

    Kind regards
    -
    Ricardo
    On Wed, Jul 25, 2012 at 8:02 PM, Philip Langdale wrote:

    For hive, yes.

    http://archive.cloudera.com/cdh4/redhat/6/x86_64/cdh/4.0.1/RPMS/noarch/

    As you can see, cdh 4.0.1 comes with hive 0.8.1. So your hive package is
    coming
    from a non-CDH source, so it lacks the necessary metadata for CM to
    identify it.

    The beeswax package on the other hand should be fine.

    --phil



    On 25 July 2012 15:54, Ricardo A. Reis wrote:

    Thank Phillip,

    I had to remove the hive-0.9.0-1.el6.noarch and
    hue-beeswax-2.0.0+59-1.cdh4.0.1.p0.1.el6.x86_64.

    Now I have non-mixed nodes. This was planned?

    --
    Ricardo
    On Tue, Jul 24, 2012 at 10:16 PM, Philip Langdale wrote:

    Ricardo,

    Look at a single host's page and go to the components tab.
    It will show you what pieces it thinks are from CDH3.

    --phil

    On Tuesday, July 24, 2012, Ricardo A. Reis wrote:

    Hello Philip,

    I think the problem is on the mixed versions, for some reason is not
    only CDH4.

    What do I need remove for correctly CDH4 identification.

    Thanks,
    Ricardo A. Reis

    On Tue, Jul 24, 2012 at 2:04 PM, Philip Langdale wrote:

    Ricardo,

    Can you send me a screenshot of your Hosts page? Also, do you already
    have a cluster? If you went through the default installer and wizard,
    then
    you should. You can't add a new cluster on existing hosts if all the
    existing
    hosts are already assigned to a previous cluster.

    Also, depending on where you get the error, it may be because you
    installed
    one version of CDH (3 vs 4) on the hosts and then said you wanted to
    create
    the other type of cluster.

    --phil




    On 23 July 2012 13:09, Ricardo A. Reis wrote:

    Hi Manoj,

    I used default installer and config.ini seems correct, but on agent
    have this error log.

    The lvm is to facilitate the lab, though the production will not use.

    [23/Jul/2012 16:44:14 +0000] 7169 MainThread throttling_logger ERROR
    Failed to collect disk stats for device 'mapper/vg_ambari-lv_root'
    Traceback (most recent call last):
    File "/usr/lib64/cmf/agent/src/cmf/monitor/host/disk_device.py", line
    54, in collect_update
    stat_file = open(self._stat_path)
    IOError: [Errno 2] No such file or directory:
    '/sys/block/mapper/vg_ambari-lv_root/stat'
    [23/Jul/2012 16:44:14 +0000] 7169 MainThread disk_devices ERROR Failed
    to add DiskDevice for disk='mapper/vg_ambari-lv_root'

    http://pastebin.com/Vj4xWJMh

    Note: This is an old vm template used to test the ambari project with a
    snapshot rollback to a clean lab.

    Thanks
    --
    Ricardo A. Reis

    On Mon, Jul 23, 2012 at 4:20 PM, Manoj Babu wrote:

    Hi Ricardo,

    This link may help you.


    https://groups.google.com/a/cloudera.org/forum/?fromgroups#!topic/scm-users/wQfG8RVXH6M


    Cheers!
    Manoj.



    On Tue, Jul 24, 2012 at 12:25 AM, Ricardo A. Reis <
    ricardo.areis@gmail.com> wrote:

    Hi Manoj,

    This is my /etc/hosts "fake fqdn for lab" . You see any problem?

    [root@node1 ~]# cat /etc/hosts
    127.0.0.1 localhost localhost.localdomain localhost4
    localhost4.localdomain4
    ::1 localhost localhost.localdomain localhost6 localhost6.localdomain6
    192.168.137.2 cm.hdpcore.net cm
    192.168.137.3 node1.hdpcore.net node1
    192.168.137.4 node2.hdpcore.net node2
    --
    --phil

Related Discussions

Discussion Navigation
viewthread | post
Discussion Overview
groupscm-users @
categorieshadoop
postedJul 23, '12 at 6:33p
activeJul 27, '12 at 10:22p
posts7
users4
websitecloudera.com
irc#hadoop

People

Translate

site design / logo © 2022 Grokbase