Forum Discussion

solom's avatar
solom
Level 4
12 years ago
Solved

Resources state offline

Hi

 

After created the resource and trying to going online for resource this command came (

Resource has not been probed on system ) when i probe the resource tell me hares -probe ( name resource )sys (name system) success . but  still offline .  i working from VOM GUI. 

 

Regards

 

 

business continuity
Cluster Server
Linux
  • Amrut_Chougale's avatar
    Amrut_Chougale
    12 years ago

    were you mentioning the problem with service group 'groupy'?

     

    As I can see in main.cf that you have pasted

     

        Mount export3 (
            MountPoint = "/export3"
            BlockDevice = "/dev/vx/dsk/groupydg/vol3"
            FSType = vxfs
            FsckOpt="y"    <<<<<<---- you have given 'y' and not '-y'
            MountOpt = rw
            )

        Mount export4 (
            MountPoint = "/export4"
            BlockDevice = "/dev/vx/dsk/groupydg/vol4"
            FSType = ext2
            FsckOpt="y"  <<<<<<---- you have given 'y' and not '-y'
            MountOpt = rw
            )

     

    can check after chaging these attributes, resources come online?

     

    Thanks and regards,

    Amrut

  • arangari's avatar
    arangari
    Level 5
    12 years ago

    @Solom : you dont need to show the main.cf from two nodes - they must be same, if the nodes are in RUNNING state.

    Which main.cf are you using out of the 4 you have pasted on  your comment above. 

    Please note the following:

    1. A group will go online, if all the resources configured in are probed. If the group is a 'failover' group, it must be probed on all the nodes in its SystemList if the node is in gab-membership. Look at the ProbesPending attribute for the group - if it is non-zero, then some resources are not yet probed.

    2. The dependecies on child should be satisfied.

    3. The failover group can be online on only one node.

    4. The probe operation ('hares -probe' cmd) is asynchronous operation. It is essenitally a monitor operation to find the initial state of resource. It will return immediately - doesn't mean the probe has completed. To verify that a resource is probed, look at the attribute 'Probed' for the resource. If the value is '0' and you beleive that atleast one monitor cycle is over - check the Flag attribute or State attribute for resource. If the flag/state has 'UNKNOWN', you may have configured resource incorrectly.

     

  • solom's avatar
    solom
    Level 4
    12 years ago

    Ok  now i have 2 nodes  node one i have on it   disk groups and volumes but the node 2 not have no volumes no disk groups and i can't create volumes on the same disk on both node so what point from cluster Service Groups move the volumes from node 1 to 2 if Something happened for node1 or what?

     

    Regards

  • solom's avatar
    solom
    Level 4
    12 years ago

    -  If the flag/state has 'UNKNOWN', you may have configured resource incorrectly -

     

    Yes it was  'UNKNOWN', and i configured resource and going online now it was from mount .

     

    Thank you both for help me . 

     

     

    Regards

  • Amrut_Chougale's avatar
    Amrut_Chougale
    12 years ago

    were you mentioning the problem with service group 'groupy'?

     

    As I can see in main.cf that you have pasted

     

        Mount export3 (
            MountPoint = "/export3"
            BlockDevice = "/dev/vx/dsk/groupydg/vol3"
            FSType = vxfs
            FsckOpt="y"    <<<<<<---- you have given 'y' and not '-y'
            MountOpt = rw
            )

        Mount export4 (
            MountPoint = "/export4"
            BlockDevice = "/dev/vx/dsk/groupydg/vol4"
            FSType = ext2
            FsckOpt="y"  <<<<<<---- you have given 'y' and not '-y'
            MountOpt = rw
            )

     

    can check after chaging these attributes, resources come online?

     

    Thanks and regards,

    Amrut

  • stinsong's avatar
    stinsong
    Level 5
    12 years ago

    Hi solom,

    Firstly, I cannot understand why there are 3 configuration within a config file main.cf. It seems about 2 pairs of nodes "vcslx242 and vcslx244" "sysa and sysb". If you are trying to configure 2 clusters you should not put configuration in one main.cf file. If you put configuration in 2 main.cf files, which one is the one cluster with problem? 

    Secondly, there is "AutoStart = 0" in the first configuration which is not right or necessary. And for IPA resource :

        IP IP_172_29_9_100 (

            Device = groupy_eth0

    Should be 

            Device = eth0

    Because there is dependency with the resources.

     

    Generally I cannot understand the strange problem you met, it maybe caused by wrong configuration file ?

  • Venkata_Reddy_C's avatar
    Venkata_Reddy_C
    Level 4
    12 years ago

    And the hostnames suggests that the configuration is copied from the sample main.cf shipped with the product.

    The sample configurations are supposed to be used after making required changes to system names etc. Could you confirm you configured the cluster with node names, sysa, sysb, vcslx242 and vcslx244?


    Regards,

    Venkat