Topic
4 replies Latest Post - ‏2013-08-28T14:40:58Z by JayFurmanek
hlucken
hlucken
2 Posts
ACCEPTED ANSWER

Pinned topic Dlpar on Redhat 6.4

‏2013-08-27T14:10:42Z |

I have installed all the latest Dlpar rpms and RSCT rpms    but the HMC is not picking up the rsct connections.  I can telnet from the lpar to the HMC on 657 so the connection is  fine plus we have Suse Linux lpars working just fine on the same network.  I have rebooted both the lpar and the HMC and still not seeing the connection.  It is suppose to work on Redhat 6 but I can't even get the HMC to see the lpar.

  • ArtisWalker
    ArtisWalker
    4 Posts
    ACCEPTED ANSWER

    Re: Dlpar on Redhat 6.4

    ‏2013-08-28T14:22:36Z  in response to hlucken

    Check your VIOS settings for this LPAR.   CHeck your vhost configuration for the LPAR.   Make sure the vhost is connected to a logical volume on your VIOS and make sure it is mapped back to the correct "Client ID".   Run lsmap -vadapter vhost0 on VIOS.

  • hlucken
    hlucken
    2 Posts
    ACCEPTED ANSWER

    Re: Dlpar on Redhat 6.4

    ‏2013-08-28T14:31:06Z  in response to hlucken

    This is all correct. The redhat lpar is up and running. The HMC just does not see it in the lspartition -dlpar which is all rsct based communication. So I am at a loss considering I have suse lpars on the same network running fine.

  • ArtisWalker
    ArtisWalker
    4 Posts
    ACCEPTED ANSWER

    Re: Dlpar on Redhat 6.4

    ‏2013-08-28T14:32:05Z  in response to hlucken

    Another thing.   Pay close attention when you saving your changes to LPAR in HMC.   "Saving Current Configuration" saves the configuration that is already running, not the configuration you modified.   To avoid this, I usually always completely shutdown the LPAR I am changing, change it, save it, then restart it.  

  • JayFurmanek
    JayFurmanek
    59 Posts
    ACCEPTED ANSWER

    Re: Dlpar on Redhat 6.4

    ‏2013-08-28T14:40:58Z  in response to hlucken

    DLPAR does indeed work on RHEL 6, but there can be some things that keep the HMC from being able to index the system

    Is the RHEL system using an IP address that was previously used by a system attached to this HMC? If so, it may be trying to use an old node ID for that system.

     

    The RSCT services should look like:

    # lssrc -a
    Subsystem                 Group             PID     Status
     ctrmc                           rsct                 2933    active
     IBM.DRM                    rsct_rm          3653    active
     IBM.ServiceRM         rsct_rm          3691    active
     IBM.CSMAgentRM   rsct_rm          3727    active
     IBM.AuditRM             rsct_rm          3758    active
     IBM.HostRM              rsct_rm          3837    active
     ctcas                            rsct                              inoperative
     IBM.ERRM                 rsct_rm                       inoperative
     IBM.SensorRM         rsct_rm                        inoperative
     IBM.LPRM                  rsct_rm                      inoperative
     IBM.MgmtDomainRM rsct_rm                   inoperative

    The IBM.HostRM service will pop up when the HMC makes contact and asks for host information. Keep an eye on that. It will go to inoperative when done.

    Do the below to remake the node ID and restart the RSCT deamons to see if that helps:

    /usr/sbin/rsct/bin/rmcctrl -z         (stop)
    /usr/sbin/rsct/bin/mknodeid -f    (force to remake node id)
    /usr/sbin/rsct/bin/rmcctrl -A         (add and start service)
    /usr/sbin/rsct/bin/rmcctrl -p         (enable connections)

    or run this to re-configure the daemons
    /usr/sbin/rsct/install/bin/recfgct

    then reboot.