[ceph-users] NFSv3 on Reef

28 Jun 2024

      Hello, 
I’m using NFSGW on a Reef release, I managed to mount cephfs volumes over NFSv4 with no problems.
 Now I need to export a CephFS volume using NFSv3 (I need to mount it from a client that only supports v3), and I tried everything I can imagine to make it work.

I unchecked the flag created an export:

but I got the error: 

that sounds like a wrong configuration.

I created the export using a json file like: 

{
  "access_type": "RW",
  "clients": [
    {
      "access_type": "RW",
      "addresses": [
        “XX.XX.XX.XX"
      ],
      "squash": "no_root_squash"
    },
    {
      "access_type": "RW",
      "addresses": [
        “XX.XX.XX.XX"
      ],
      "squash": "no_root_squash"
    }
  ],
  "cluster_id": “nfsgw",
  "export_id": 1,
  "fsal": {
    "fs_name": “XXX",
    "name": "CEPH",
    "user_id": "nfs.nfsgw.1"
  },
  "path": "/",
  "protocols": [
    3,
    4
  ],
  "pseudo": "***",
  "security_label": false,
  "squash": "no_root_squash",
  "transports": [
    "TCP",
    "UDP"
  ]
}

and applied the conf with ceph nfs export apply nfsgw -i export.yaml

But  /etc/ganesha/ganesha.conf  still contains  

NFS_CORE_PARAM {
        Enable_NLM = false;
        Enable_RQUOTA = false;
        Protocols = 4;
        NFS_Port = 12049;
}

even after pushing a custom config: 

# ceph nfs cluster config get nfsgw
NFS_CORE_PARAM {
        Enable_NLM = false;
        Enable_RQUOTA = false;
        Protocols = 3, 4;
        NFS_Port = 12049;
}

and redeploying everything with ceph orch redeploy.

Any help would be very appreciated, thanks in advance…

Ramon   

-- 
Ramon Orrù
Servizio di Calcolo
Laboratori Nazionali di Frascati
Istituto Nazionale di Fisica Nucleare
Via E. Fermi, 54 - 00044 Frascati (RM) Italy
Tel. +39 06 9403 2345