6.3 HP StoreAll 9300/9320 Storage Administrator Guide (AW549-96072, June 2013)

1 Upgrading the StoreAll software to the 6.3 release

This chapter describes how to upgrade to the 6.3 StoreAll software release. You can also use this

procedure for any subsequent 6.3.x patches.

IMPORTANT: Print the following table and check off each step as you complete it.

NOTE: (Upgrades from version 6.0.x) CIFS share permissions are granted on a global basis in

v6.0.X. When upgrading from v6.0.X, confirm that the correct share permissions are in place.

Table 1 Prerequisites checklist for all upgrades

Step

completed?DescriptionStep

Verify that the entire cluster is currently running StoreAll 6.0 or later by entering the following

command:

ibrix_version -l

IMPORTANT: All the StoreAll nodes must be at the same release.

• If you are running a version of StoreAll earlier than 6.0, upgrade the product as

described in “Cascading Upgrades” (page 154).

• If you are running StoreAll 6.0 or later, proceed with the upgrade steps in this section.

Verify that the /local partition contains at least 4 GB for the upgrade by using the

following command:

df -kh /local

The 6.3 release requires that nodes hosting the agile Fusion Manager be registered on the

cluster network. Run the following command to verify that nodes hosting the agile Fusion

Manager have IP addresses on the cluster network:

ibrix_fm -l

If a node is configured on the user network, see “Node is not registered with the cluster

network ” (page 21) for a workaround.

NOTE: The Fusion Manager and all file serving nodes must be upgraded to the new

release at the same time. Do not change the active/passive Fusion Manager configuration

during the upgrade.

Verify that the crash kernel parameter on all nodes has been set to 256M by viewing the

default boot entry in the /etc/grub.conf file, as shown in the following example:

kernel /vmlinuz-2.6.18-194.el5 ro root=/dev/vg1/lv1

crashkernel=256M@16M

The /etc/grub.conf file might contain multiple instances of the crash kernel parameter.

Make sure you modify each instance that appears in the file.

If you must modify the /etc/grub.conf file, follow the steps in this section:

1. Use SSH to access the active Fusion Manager (FM).

2. Do one of the following:

• (Versions 6.2 and later) Place all passive FMs into nofmfailover mode:

ibrix_fm -m nofmfailover -A

• (Versions earlier than 6.2) Place all passive FMs into maintenance mode:

ibrix_fm -m maintenance -A

3. Disable Segment Server Failover on each node in the cluster:

ibrix_server -m -U -h <node>

4. Set the crash kernel to 256M in the /etc/grub.conf file. The /etc/grub.conf

file might contain multiple instances of the crash kernel parameter. Make sure you modify

each instance that appears in the file.