Power

 View Only
  • 1.  IBM i NFS issues

    Posted 3 days ago

    Hello,

    We need to setup IBM i to use remote optical device for backup using NFS of a backup solution called Netbackup via it's feature called Universal Share.

    We noticed that the IBMi partition can mount the remote nfs, it can create remote images, volume_list but when we try to backup or run command wrkimgclge this error is sent OPT2426. Which seems related to permissions. we gave all permissions on server side already, but the error is still persisting.

    Since this solution is based on Red Hat, we thought of first trying to get the NFS client on IBM i to work from a standard Red Hat server in order to understand what the problem is. However, with this Red Hat server, the NFS mount on the IBM i side doesn't even go through, as error CPDA1B9 is returned. I would like to point out that other Red Hat partitions are able to use NFS servers without any issues, whether from NetBackup or from the standalone Red Hat server.

    Has anyone ever encountered this type of problem? Or does anyone have an idea on how to resolve it?



    ------------------------------
    Assane BA
    ------------------------------


  • 2.  RE: IBM i NFS issues

    Posted 3 days ago
    Edited by Diego KESSELMAN BARRIONUEVO 3 days ago

    Hi Assane,

    I have worked with IBM i , NFS and Remote Virtual Image Catalog in different operating systems, and all have their small details.

    Remember: IBM i Remote Virtual Image Catalog can only works with NFS v3 and UDP, not TCP or other NFS server versions.

    If NetBackup is using NFS server from RedHat we can make it work, but it is important to know the actual RHEL release, because 8.x and 9.x need some adjustments from default values.

    * Run these commands:
    nfsconf --set nfsd udp y
    nfsconf --set nfsd vers3 y
    * Enable specific NFS UDP and TCP ports to work properly with FIREWALLD , or just disable the firewall.
    * Set the required authorizations on the RH directory, following the guide :
    https://www.ibm.com/support/pages/ibm-i-save-and-restore-using-virtual-optical-images-nfs-server
    or this
    https://www.linkedin.com/pulse/ibm-i-backup-your-lpar-nfs-without-using-its-own-diego-kesselman/
    * It is important (still don't know why) to use a second level directory, check documents. ie: /NFS/MyBackups/ 
    * Restart the NFS server to take new values

    Hope these can help you.

    BTW: I don't think this is an efficient way to backup your IBM i. NetBackup will not be able to know media content, so your restore mechanism will be slow, because you need to restore all media set, and complex, because your restore requires some IBM i setup and manual operation.



    ------------------------------
    Diego KESSELMAN BARRIONUEVO
    General Manager
    ESSELWARE Soluciones, SA de CV
    CDMX DIF
    +525540408279
    ------------------------------



  • 3.  RE: IBM i NFS issues

    Posted 3 days ago

    Hello Diego,

    Thank you for your feedback.
    After running the following commands on the Red Hat side:

    nfsconf --set nfsd udp y
    nfsconf --set nfsd vers3 y

    the NFS mount on the IBM i partition now works.

    However, I am back to the same error I had with NetBackup when trying to perform a backup or run the wrkimgclge command on the optical drive (OPT2426 error).
    Yet, I was able to create the remote images without any issue.

    I believe there must be some permission-related problems that I haven't been able to identify yet.

    Here is a view of the file permissions on the Red Hat side:

    [root@srv-syslog test]# ls -altr
    total 102408
    drwxr-xr-x. 22 root root      4096 Aug 25 13:21 ..
    -rwxrwxrwx   1 root root         0 Aug 25 14:19 testTTT
    -rwxrwxrwx   1 root root 104857600 Sep  1 16:34 IMAGE01.ISO
    drwxrwxrwx   2 root root        59 Sep  1 16:36 .
    -rwxrwxrwx   1 root root        14 Sep  1 16:36 VOLUME_LIST

    The firewall service is also stopped:

    [root@srv-syslog test]# systemctl status firewalld
    ● firewalld.service - firewalld - dynamic firewall daemon
       Loaded: loaded (/usr/lib/systemd/system/firewalld.service; enabled; vendor preset: enabled)
       Active: inactive (dead) since Mon 2025-08-25 13:49:51 GMT; 1 weeks 0 days ago
         Docs: man:firewalld(1)
      Process: 1184 ExecStart=/usr/sbin/firewalld --nofork --nopid $FIREWALLD_ARGS (code=exited, status=0/SUCCESS)
    Main PID: 1184 (code=exited, status=0/SUCCESS)

    Red Hat version:

    [root@srv-syslog test]# cat /etc/redhat-release
    Red Hat Enterprise Linux release 8.10 (Ootpa)
    [root@srv-syslog test]#

    On the IBM i side, the error is as follows:

    Message ID . . . . . . : OPT2426
    Sent date . . . . . . : 01/09/25 Sent time . . . . : 16:30:36
    Message . . . . : Not authorized to object. The object is /var/test.

    Cause . . . . . :
    You do not have the authority for the object /var/test or for one of the directories in the path to the object.

    Recovery . . . :
    Contact your security officer, the object owner, or another authorized person to obtain the appropriate authority to use the object /var/test or the directories in the object path, and then resubmit the request.
    If the object name is *N, it is not possible to determine which of the object authorities is insufficient.

    Note: Long names may be truncated.

    Technical description . . . . . . :
    Extended status for the error is 0D1A61A5819961A385A2A34040404040



    ------------------------------
    Assane BA
    ------------------------------



  • 4.  RE: IBM i NFS issues

    Posted 3 days ago

    Can you share the PATH to the files, the authorities on the directories and the /etc/exports content from your RedHat?

    FYI: In my experience, I do not suggest to use an exported directory under /var file system.



    ------------------------------
    Diego E. KESSELMAN
    IBM Champion
    ESSELWARE Soluciones
    ------------------------------



  • 5.  RE: IBM i NFS issues

    Posted yesterday
    Edited by Assane BA yesterday

    Hello Diego,

    We finally created the share on new filesystem separated from root filesystem to avoid permission issues on /var. The option insecure was added after testing unsuccefully with secure option

    [root@srv-syslog data]# exportfs -v

    /data           10.1.0.2(sync,wdelay,hide,no_subtree_check,sec=sys,rw,insecure,no_root_squash,no_all_squash)

    The authorities on directories:

    [root@srv-syslog data]# pwd

    /data

    [root@srv-syslog data]# ls -altr

    total 102604

    drwxrwxrwx   2 root root     16384  2 sept. 13:31 lost+found

    dr-xr-xr-x. 18 root root       256  2 sept. 13:58 ..

    -rwxrwxrwx   1 root root 104857600  2 sept. 14:08 IMAGE01.ISO

    -rwxrwxrwx   1 root root        14  2 sept. 14:09 VOLUME_LIST

    -rw-r--r--   1 root root    167689  2 sept. 14:38 logs

    drwxrwxrwx   3 root root      4096  2 sept. 14:40 .

    -rw-r--r--   1 root root     13665  2 sept. 14:40 log2.txt

    [root@srv-syslog data]# ls -altr /

    total 32

    drwxr-xr-x.   2 root root    6 21 juin   2021 srv

    lrwxrwxrwx.   1 root root    8 21 juin   2021 sbin -> usr/sbin

    drwxr-xr-x.   2 root root    6 21 juin   2021 opt

    drwxr-xr-x.   2 root root    6 21 juin   2021 mnt

    drwxr-xr-x.   2 root root    6 21 juin   2021 media

    lrwxrwxrwx.   1 root root    9 21 juin   2021 lib64 -> usr/lib64

    lrwxrwxrwx.   1 root root    7 21 juin   2021 lib -> usr/lib

    drwxr-xr-x.   3 root root   16 21 juin   2021 home

    lrwxrwxrwx.   1 root root    7 21 juin   2021 bin -> usr/bin

    drwxr-xr-x.  12 root root  144 24 avril 10:46 usr

    dr-xr-xr-x.   5 root root 4096 25 avril 09:55 boot

    -rw-r--r--    1 root root    0 25 avril 11:04 .autorelabel

    dr-xr-xr-x  385 root root    0  3 juil. 02:16 proc

    dr-xr-xr-x   13 root root    0  3 juil. 02:16 sys

    drwxr-xr-x.  22 root root 4096 25 août  13:21 var

    dr-xr-x---.  16 root root 4096  1 sept. 16:17 root

    drwxr-xr-x   47 root root 1320  2 sept. 06:09 run

    drwxr-xr-x   20 root root 3220  2 sept. 13:31 dev

    dr-xr-xr-x.  18 root root  256  2 sept. 13:58 ..

    dr-xr-xr-x.  18 root root  256  2 sept. 13:58 .

    drwxr-xr-x. 145 root root 8192  2 sept. 14:20 etc

    drwxrwxrwx    3 root root 4096  2 sept. 14:40 data

    drwxrwxrwt.   9 root root 4096  2 sept. 15:35 tmp


    but still having OPT2426 on IBM i side.
    ------------------------------
    Assane BA
    ------------------------------



  • 6.  RE: IBM i NFS issues

    Posted yesterday

    Hi Assane,

    Please, try these:

    • Create a 2nd level directory, let say /data/images/ , and export the /data directory. I know it sounds absurd, but sometimes fixes issues with IBM i. 
    • Fix the path on the optical device
    • Is the IP address on your /etc/exports pointing to your IBM I interface or your Service Tools Server LAN Adapter address? Try adding both, please.
    • Try using this values in your /etc/exports
      /data 10.1.0.2,10.1.0.102(rw,sync,no_subtree_check)
    • Add both IP addresses to the /etc/hosts file with any name. THIS IS REALLY IMPORTANT!

    Please, let me know if this helps



    ------------------------------
    Diego E. KESSELMAN
    IBM Champion
    ESSELWARE Soluciones
    ------------------------------