HP StoreAll OS 6.5.1 Upgrade Guide
d. Repeat steps 1 through 3 until you have successfully unmounted the file system.
5. On all servers in the cluster, do the following steps:
a. If you have not already done so, mount the QR ISO:
a. Copy the ISO to the local hard drive, as shown in the following example:
cp X9000-QRDVD-6.x.xxx.x86_64.iso.signed.iso /local
b. Create the /media directory.
c. Mount the ISO image, as shown in the following example:
mount -r -o loop
/local/X9000-QRDVD-6.x.xxx.x86_64.iso.signed.iso /media
b. Run the following script:
/media/ibrix/prepSS.sh
The script performs a number of tasks to prepare the server for the upgrade, such as the
following:
• Verifies that Fusion Manager has been disabled and disables Fusion Manager from
starting on reboot.
• Disables the ibrix_server daemon from starting on reboot.
• Updates the multipath.conf and modprobe.conf files.
• Removes outdated Linux Red Hat packages. The updated packages will be installed
during the upgrade to Red Hat Linux 5.8.
• Verifies that the crash kernel is set properly.
c. If the script indicates that 256 MB have not been reserved for the crash kernel in /etc/
grub.conf, edit the /etc/grub.conf file and verify that all the kernel lines contain
the kernel argument of crashkernel=256M@16M. This helps ensure there is space reserved
to successfully collect crash data if necessary.
The following is an example kernel line from the grub.conf file:
kernel /vmlinuz-2.6.18-194.el5 ro root=/dev/vg1/lv1
crashkernel=256M@16M
IMPORTANT: If Express Query is running on a server, increase the reserved space for
the crash kernel to 384 MB to ensure there is enough space reserved to successfully collect
crash data:
kernel /boot/vmlinuz-2.6.18-194.el5 ro root=LABEL=ROOT2
console=ttyS0,115200n8 console=tty0 selinux=0 crashkernel=384M@16M
elevator=deadline rhgb quiet
d. Upgrade the server to Red Hat Linux 5.8:
yum update --disablerepo=* --enablerepo=media --noplugins
--nogpgcheck
e. Reboot the server:
reboot -f
f. (9300 and 9320 platforms only) After each server has rebooted, verify that Device
Mapper Multipath is working:
multipath -ll
g. Repeat steps 1 through 6 for all servers.
6. Perform the following steps on all servers of the cluster:
Verify that ssh shared keys have been set up. To do this, run the following command on the
node hosting the active instance of the agile Fusion Manager:
ssh <server_name>
62 Manual offline upgrades for StoreAll software 6.1.5 to 6.5.1