InfoScale™ 9.0 Storage Foundation Cluster File System High Availability Configuration and Upgrade Guide - Linux

To upgrade the operating system

1. Rename the /etc/llttab file to prevent LLT from starting automatically when the node starts:

   # mv /etc/llttab /etc/llttab.save

2. Create install-db file to prevent VxVM daemons or processes from starting automatically when the node starts.

   # touch /etc/vx/reconfig.d/state.d/install-db 

3. Upgrade the operating system on all nodes in the cluster.

   For instructions, see the operating system documentation.

4. If you upgraded the operating system, restart the nodes if required:

   # shutdown -r now

5. After the system restarts, rename the /etc/llttab file to its original name:

   # mv /etc/llttab.save /etc/llttab

6. Enable VxVM to start after system restarts.

   # rm /etc/vx/reconfig.d/state.d/install-db

To perform the upgrade

1. Log in as superuser.
2. Insert the appropriate media disc per your distribution and architecture into your system's DVD-ROM drive.
3. If volume management software is running on your system, the software disc automatically mounts as /mnt/cdrom.

   If volume management software is not available to mount the disc, you must mount it manually, enter:

   # mount -o ro /dev/cdrom /mnt/cdrom

4. Change to the top-level directory on the disc:

   # cd /mnt/cdrom

5. Verify there are no VxFS file systems mounted on the nodes being upgraded:

   # mount -t vxfs

   If any VxFS file systems are mounted, offline the group on each node of the cluster:

   # hagrp -offline group -sys sys1
   # hagrp -offline group -sys sys2
   # hagrp -offline group -sys sys3
   # hagrp -offline group -sys sys4

   where group is the VCS service group that has the CVMVolDg and CFSMount resource.

   If VxFS are not managed by VCS then unmount them manually:

   # umount mount_point

   Repeat this step for each SFCFSHA service group.

6. If a cache area is online, you must take the cache area offline before you upgrade the VxVM RPM. Use the following command to take the cache area offline:

   # sfcache offline cachename

7. Start the upgrade from any node in the cluster. Enter the following command, and then press y to upgrade the cluster configuration.

   # ./installer -upgrade

8. You are prompted to enter the system names (in the following example, "sys1" and "sys2") on which the software is to be upgraded. Enter the system name or names and then press Return.

   Enter the system names separated by spaces: 
   [q, ?] sys1 sys2

9. During the initial system check, the installer verifies that communication between systems has been set up.

   If the installer hangs or asks for a login password, you have the option to let the installer configure SSH or RSH communications between the systems. If you choose to allow this configuration, select the communication type and provide the root passwords for each system.

10. At the prompt, specify whether you accept the terms of the End User License Agreement (EULA).

    Do you agree with the terms of the End User License Agreement
    as specified in the EULA/en/EULA_ENTERPRISE_Ux_9.0.pdf file 
    present on media? [y,n,q,?] y

11. The installer discovers if any of the systems that you are upgrading have mirrored and encapsulated boot disks. For each system that has a mirrored boot disk, you have the option to create a backup of the system's boot disk group before the upgrade proceeds. If you want to split the boot disk group to create a backup, answer y.
12. The installer then prompts you to name the backup boot disk group. Enter the name for it or press Enter to accept the default.
13. If you are prompted to start the split operation. Press y to continue.

    Note:

    The split operation can take some time to complete.

14. Output shows information that SFCFSHA must be stopped on a running system. Enter y to continue.
15. The installer displays the following question before the installer stops the product processes if the current cluster is secured and version is prior to 6.2.:

    • Do you want to grant read access to everyone? [y,n,q,?]

      • To grant read access to all authenticated users, type y.

      • To grant usergroup specific permissions, type n.

    • Do you want to provide any usergroups that you would like to grant read access?[y,n,q,?]

      • To specify usergroups and grant them read access, type y

      • To grant read access only to root users, type n. The installer grants read access read access to the root users.

    Note:

    Separate the usergroup names with spaces. To grant read access to a usergroup on a specific node, specify usergroup as <usergroup>@<node_name>. You can also specify usergroups here and create them later.

16. Enter y for summary information and reboots if the boot disk is encapsulated before the upgrade.

    Do not remove the log files until the Veritas InfoScale product is working properly on your system. Technical Support will need these log files for debugging purposes.

17. Only perform this step if you have split the mirrored root disk to back it up. After a successful reboot, verify the upgrade and re-join the backup disk group. If the upgrade fails, revert to the backup disk group.

    See Re-joining the backup boot disk group into the current disk group.

    See Reverting to the backup boot disk group after an unsuccessful upgrade.