Re: nfs-ganesha HA with arbiter volume

Jiffin Tony Thottan <jthottan@xxxxxxxxxx> · Mon, 21 Sep 2015 15:18:10 +0530



    On 21/09/15 13:56, Tiemen Ruiten wrote:

    
      Hello Soumya, Kaleb, list,
        

        This Friday I created the gluster_shared_storage volume
          manually, I just tried it with the command you supplied, but
          both have the same result:
        

        from etc-glusterfs-glusterd.vol.log on the node where I
          issued the command:
        

          [2015-09-21 07:59:47.756845] I [MSGID: 106474]
            [glusterd-ganesha.c:403:check_host_list] 0-management:
            ganesha host found Hostname is cobalt
          [2015-09-21 07:59:48.071755] I [MSGID: 106474]
            [glusterd-ganesha.c:349:is_ganesha_host] 0-management:
            ganesha host found Hostname is cobalt
          [2015-09-21 07:59:48.653879] E [MSGID: 106470]
            [glusterd-ganesha.c:264:glusterd_op_set_ganesha]
            0-management: Initial NFS-Ganesha set up failed
        
      
    As far as what I understand from the logs, it called
    setup_cluser()[calls `ganesha-ha.sh` script ] but script failed.

    Can u please provide following details :

    -Location of ganesha.sh file??

    -Location of ganesha-ha.conf, ganesha.conf files ?

    
    And also can u cross check whether all the prerequisites before HA
    setup satisfied ?

    
    --

    With Regards,

    Jiffin

    
          [2015-09-21 07:59:48.653912] E [MSGID: 106123]
            [glusterd-syncop.c:1404:gd_commit_op_phase] 0-management:
            Commit of operation 'Volume (null)' failed on localhost :
            Failed to set up HA config for NFS-Ganesha. Please check the
            log file for details
          [2015-09-21 07:59:45.402458] I [MSGID: 106006]
            [glusterd-svc-mgmt.c:323:glusterd_svc_common_rpc_notify]
            0-management: nfs has disconnected from glusterd.
          [2015-09-21 07:59:48.071578] I [MSGID: 106474]
            [glusterd-ganesha.c:403:check_host_list] 0-management:
            ganesha host found Hostname is cobalt
        
        
        from etc-glusterfs-glusterd.vol.log on the other node:
        

          [2015-09-21 08:12:50.111877] E [MSGID: 106062]
            [glusterd-op-sm.c:3698:glusterd_op_ac_unlock] 0-management:
            Unable to acquire volname
          [2015-09-21 08:14:50.548087] E [MSGID: 106062]
            [glusterd-op-sm.c:3635:glusterd_op_ac_lock] 0-management:
            Unable to acquire volname
          [2015-09-21 08:14:50.654746] I [MSGID: 106132]
            [glusterd-proc-mgmt.c:83:glusterd_proc_stop] 0-management:
            nfs already stopped
          [2015-09-21 08:14:50.655095] I [MSGID: 106474]
            [glusterd-ganesha.c:403:check_host_list] 0-management:
            ganesha host found Hostname is cobalt
          [2015-09-21 08:14:51.287156] E [MSGID: 106062]
            [glusterd-op-sm.c:3698:glusterd_op_ac_unlock] 0-management:
            Unable to acquire volname
        
        
        from etc-glusterfs-glusterd.vol.log on the arbiter node:
        

          [2015-09-21 08:18:50.934713] E [MSGID: 101075]
            [common-utils.c:3127:gf_is_local_addr] 0-management: error
            in getaddrinfo: Name or service not known
          [2015-09-21 08:18:51.504694] E [MSGID: 106062]
            [glusterd-op-sm.c:3698:glusterd_op_ac_unlock] 0-management:
            Unable to acquire volname

          
        I have put the hostnames of all servers in my /etc/hosts
          file, including the arbiter node.
        

        On 18 September 2015 at 16:52, Soumya
          Koduri <skoduri@xxxxxxxxxx>
          wrote:

          Hi Tiemen,

            
            One of the pre-requisites before setting up nfs-ganesha HA
            is to create and mount shared_storage volume. Use below CLI
            for that

            
            "gluster volume set all cluster.enable-shared-storage
            enable"

            
            It shall create the volume and mount in all the nodes
            (including the arbiter node). Note this volume shall be
            mounted on all the nodes of the gluster storage pool (though
            in this case it may not be part of nfs-ganesha cluster).

            
            So instead of manually creating those directory paths,
            please use above CLI and try re-configuring the setup.

            
            Thanks,

            Soumya

              
              On 09/18/2015 07:29 PM, Tiemen Ruiten wrote:

            
                Hello Kaleb,

                
                I don't:

                
                # Name of the HA cluster created.

                # must be unique within the subnet

                HA_NAME="rd-ganesha-ha"

                #

                # The gluster server from which to mount the shared data
                volume.

                HA_VOL_SERVER="iron"

                #

                # N.B. you may use short names or long names; you may
                not use IP addrs.

                # Once you select one, stay with it as it will be mildly
                unpleasant to

                # clean up if you switch later on. Ensure that all names
                - short and/or

                # long - are in DNS or /etc/hosts on all machines in the
                cluster.

                #

                # The subset of nodes of the Gluster Trusted Pool that
                form the ganesha

                # HA cluster. Hostname is specified.

                HA_CLUSTER_NODES="cobalt,iron"

                #HA_CLUSTER_NODES="server1.lab.redhat.com

              
              <http://server1.lab.redhat.com>,server2.lab.redhat.com

              <http://server2.lab.redhat.com>,..."

                #

                # Virtual IPs for each of the nodes specified above.

                VIP_server1="10.100.30.101"

                VIP_server2="10.100.30.102"

                #VIP_server1_lab_redhat_com="10.0.2.1"

                #VIP_server2_lab_redhat_com="10.0.2.2"

                
                hosts cobalt & iron are the data nodes, the arbiter
                ip/hostname (neon)

                isn't mentioned anywhere in this config file.

                
                On 18 September 2015 at 15:56, Kaleb S. KEITHLEY <kkeithle@xxxxxxxxxx

              
                  <mailto:kkeithle@xxxxxxxxxx>>
                  wrote:

                  
                      On 09/18/2015 09:46 AM, Tiemen Ruiten wrote:

                      > Hello,

                      >

                      > I have a Gluster cluster with a single
                  replica 3, arbiter 1 volume (so

                      > two nodes with actual data, one arbiter
                  node). I would like to setup

                      > NFS-Ganesha HA for this volume but I'm having
                  some difficulties.

                      >

                      > - I needed to create a directory
                  /var/run/gluster/shared_storage

                      > manually on all nodes, or the command
                  'gluster nfs-ganesha enable would

                      > fail with the following error:

                      > [2015-09-18 13:13:34.690416] E [MSGID:
                  106032]

                      > [glusterd-ganesha.c:708:pre_setup]
                  0-THIS->name: mkdir() failed on path

                      > /var/run/gluster/shared_storage/nfs-ganesha,
                  [No such file or directory]

                      >

                      > - Then I found out that the command connects
                  to the arbiter node as

                      > well, but obviously I don't want to set up
                  NFS-Ganesha there. Is it

                      > actually possible to setup NFS-Ganesha HA
                  with an arbiter node? If it's

                      > possible, is there any documentation on how
                  to do that?

                      >

                  
                      Please send the /etc/ganesha/ganesha-ha.conf file
                  you're using.

                  
                      Probably you have included the arbiter in your HA
                  config; that would be

                      a mistake.

                  
                      --

                  
                      Kaleb

                  
                  --

                  Tiemen Ruiten

                  Systems Engineer

                  R&D Media

                  
              _______________________________________________

              Gluster-users mailing list

              Gluster-users@xxxxxxxxxxx

              http://www.gluster.org/mailman/listinfo/gluster-users

              
        -- 

        
          Tiemen Ruiten

            Systems Engineer

            R&D Media

          
      _______________________________________________
Gluster-users mailing list
Gluster-users@xxxxxxxxxxx
http://www.gluster.org/mailman/listinfo/gluster-users
    
    
_______________________________________________
Gluster-users mailing list
Gluster-users@xxxxxxxxxxx
http://www.gluster.org/mailman/listinfo/gluster-users