VCS solaris10 take the bge card offline. - Veritas Cluster Server

This is a discussion on VCS solaris10 take the bge card offline. - Veritas Cluster Server ; Hi, I install VCS on two V240 machines, on one of them when I take the clustergroup offline the machine public nic become unplumb. Any suggestions? Thanks...

+ Reply to Thread
Results 1 to 4 of 4

Thread: VCS solaris10 take the bge card offline.

  1. VCS solaris10 take the bge card offline.


    Hi,
    I install VCS on two V240 machines, on one of them when I take the clustergroup
    offline the
    machine public nic become unplumb.

    Any suggestions?

    Thanks

  2. Re: VCS solaris10 take the bge card offline.

    I guess that

    1. you are using a NIC (bge) in teh clusterservice group
    2. I guess it is the same NIC that gets plumbed up by a hostnaem file
    (at startup)


    Solution :

    1. Take the NIC out. (Just make sure that the interface will get plumbed
    on startup - before VCS)
    2. Show us the /etc/VRTSvcs/conf/config/main.cf and
    /var/VRTSvcs/log/engine_A.log file where the NIC goes offline , and
    /etc/hostname.* files (and perhaps ifconfig -a)

    Erez wrote:
    > Hi,
    > I install VCS on two V240 machines, on one of them when I take the clustergroup
    > offline the
    > machine public nic become unplumb.
    >
    > Any suggestions?
    >
    > Thanks


  3. Re: VCS solaris10 take the bge card offline.


    Attach all the relevant VCS information
    /root# ifconfig -a
    lo0: flags=2001000849 mtu 8232
    index 1
    inet 127.0.0.1 netmask ff000000
    bge0: flags=1000843 mtu 1500 index 2
    inet 10.66.6.18 netmask fffffe00 broadcast 10.66.7.255
    ether 0:14:4f:73:db:64
    lo0: flags=2002000849 mtu 8252
    index 1
    inet6 ::1/128
    bge0: flags=2000841 mtu 1500 index 2
    inet6 fe80::214:4fff:fe73:db64/10
    ether 0:14:4f:73:db:64
    /root# cat /etc/hostname.bge0
    hostname.bge0 hostname6.bge0
    /root# cat /etc/hostname.bge0
    chip
    /root# cat /etc/hosts
    #
    # Internet host table
    #
    ::1 localhost
    127.0.0.1 localhost
    10.66.6.18 chip loghost


    /root# cat /etc/llttab
    set-node chip
    set-cluster 23
    link bge1 /dev/bge:1 - ether - -
    link bge2 /dev/bge:2 - ether - -
    /root#


    include "types.cf"

    cluster vcs10 (
    UserNames = { admin = ********, root = ******* }
    ClusterAddress = "10.66.6.17"
    Administrators = { admin }
    )

    system chip (
    )

    system pooh (
    )

    group ClusterService (
    SystemList = { pooh = 0, chip = 1 }
    AutoStartList = { pooh, chip }
    ClusterFailOverPolicy = Auto
    OnlineRetryLimit = 3
    )

    IP webip (
    Device = bge0
    Address = "10.66.6.17"
    NetMask = "255.255.254.0"
    )

    NIC csgnic (
    Device = bge0
    NetworkType = Ethernet
    NetworkHosts = { "10.66.6.54", "10.66.6.18" }
    )

    VRTSWebApp VCSweb (
    Critical = 0
    AppName = cmc
    InstallDir = "/opt/VRTSweb/VERITAS"
    TimeForOnline = 5
    RestartLimit = 3
    )

    VCSweb requires webip
    webip requires csgnic


    /root# cat /var/VRTSvcs/log/engine_A.log |grep bge
    ifconfig: setifflags: SIOCGLIFFLAGS: bge0:1: no such interface
    2008/03/02 16:13:57 VCS WARNING V-16-10001-7504 (chip) NIC:csgnic:monitor:No
    Broadcast address was found. Resource is OFFLINE. Will not go ONLINE until
    a Broadcast address is assigned to the interface bge0

    Thanks,
    Erez
    Me wrote:
    >I guess that
    >
    >1. you are using a NIC (bge) in teh clusterservice group
    >2. I guess it is the same NIC that gets plumbed up by a hostnaem file
    >(at startup)
    >
    >
    >Solution :
    >
    >1. Take the NIC out. (Just make sure that the interface will get plumbed


    >on startup - before VCS)
    >2. Show us the /etc/VRTSvcs/conf/config/main.cf and
    >/var/VRTSvcs/log/engine_A.log file where the NIC goes offline , and
    >/etc/hostname.* files (and perhaps ifconfig -a)
    >
    >Erez wrote:
    >> Hi,
    >> I install VCS on two V240 machines, on one of them when I take the clustergroup
    >> offline the
    >> machine public nic become unplumb.
    >>
    >> Any suggestions?
    >>
    >> Thanks



  4. Re: VCS solaris10 take the bge card offline.

    OK, most important here is the fact that you have a NIC resource

    > NIC csgnic (
    > Device = bge0
    > NetworkType = Ethernet
    > NetworkHosts = { "10.66.6.54", "10.66.6.18" }
    > )




    and once the ClusterService group goes offline, what do you expect to
    happen ?


    Yep, the NIC will be taken offline (basically unplumbed)

    so, remove the NIC resource





    Erez wrote:
    > Attach all the relevant VCS information
    > /root# ifconfig -a
    > lo0: flags=2001000849 mtu 8232
    > index 1
    > inet 127.0.0.1 netmask ff000000
    > bge0: flags=1000843 mtu 1500 index 2
    > inet 10.66.6.18 netmask fffffe00 broadcast 10.66.7.255
    > ether 0:14:4f:73:db:64
    > lo0: flags=2002000849 mtu 8252
    > index 1
    > inet6 ::1/128
    > bge0: flags=2000841 mtu 1500 index 2
    > inet6 fe80::214:4fff:fe73:db64/10
    > ether 0:14:4f:73:db:64
    > /root# cat /etc/hostname.bge0
    > hostname.bge0 hostname6.bge0
    > /root# cat /etc/hostname.bge0
    > chip
    > /root# cat /etc/hosts
    > #
    > # Internet host table
    > #
    > ::1 localhost
    > 127.0.0.1 localhost
    > 10.66.6.18 chip loghost
    >
    >
    > /root# cat /etc/llttab
    > set-node chip
    > set-cluster 23
    > link bge1 /dev/bge:1 - ether - -
    > link bge2 /dev/bge:2 - ether - -
    > /root#
    >
    >
    > include "types.cf"
    >
    > cluster vcs10 (
    > UserNames = { admin = ********, root = ******* }
    > ClusterAddress = "10.66.6.17"
    > Administrators = { admin }
    > )
    >
    > system chip (
    > )
    >
    > system pooh (
    > )
    >
    > group ClusterService (
    > SystemList = { pooh = 0, chip = 1 }
    > AutoStartList = { pooh, chip }
    > ClusterFailOverPolicy = Auto
    > OnlineRetryLimit = 3
    > )
    >
    > IP webip (
    > Device = bge0
    > Address = "10.66.6.17"
    > NetMask = "255.255.254.0"
    > )
    >
    > NIC csgnic (
    > Device = bge0
    > NetworkType = Ethernet
    > NetworkHosts = { "10.66.6.54", "10.66.6.18" }
    > )
    >
    > VRTSWebApp VCSweb (
    > Critical = 0
    > AppName = cmc
    > InstallDir = "/opt/VRTSweb/VERITAS"
    > TimeForOnline = 5
    > RestartLimit = 3
    > )
    >
    > VCSweb requires webip
    > webip requires csgnic
    >
    >
    > /root# cat /var/VRTSvcs/log/engine_A.log |grep bge
    > ifconfig: setifflags: SIOCGLIFFLAGS: bge0:1: no such interface
    > 2008/03/02 16:13:57 VCS WARNING V-16-10001-7504 (chip) NIC:csgnic:monitor:No
    > Broadcast address was found. Resource is OFFLINE. Will not go ONLINE until
    > a Broadcast address is assigned to the interface bge0
    >
    > Thanks,
    > Erez
    > Me wrote:
    >
    >>I guess that
    >>
    >>1. you are using a NIC (bge) in teh clusterservice group
    >>2. I guess it is the same NIC that gets plumbed up by a hostnaem file
    >>(at startup)
    >>
    >>
    >>Solution :
    >>
    >>1. Take the NIC out. (Just make sure that the interface will get plumbed

    >
    >
    >>on startup - before VCS)
    >>2. Show us the /etc/VRTSvcs/conf/config/main.cf and
    >>/var/VRTSvcs/log/engine_A.log file where the NIC goes offline , and
    >>/etc/hostname.* files (and perhaps ifconfig -a)
    >>
    >>Erez wrote:
    >>
    >>>Hi,
    >>>I install VCS on two V240 machines, on one of them when I take the clustergroup
    >>>offline the
    >>>machine public nic become unplumb.
    >>>
    >>>Any suggestions?
    >>>
    >>>Thanks

    >
    >


+ Reply to Thread