IBM N Series, Hardware Manual

Page: 307 / 368

Chapter 20. Designing for nondisruptive upgrades
287
When AutoSupport is enabled, failed drives are detected automatically and replacement
drives are shipped for installation at the administrator's convenience. Generally, enable
AutoSupport for all storage systems.
Removing all old core files
Clear /etc/crash/ of old core files before the NDU is performed. Run savecore -l to
determine whether there are any cores in memory and, if so, flush them as required.
Upgrading disk and shelf firmware
Shelf firmware upgrades must be completed before Data ONTAP NDU is performed.
Disk firmware upgrades are automatically performed in the background for all drives starting
with Data ONTAP 8.0.2.
For more information about upgrading shelf and disk firmware, see 20.2, “Shelf firmware
NDU” on page 288, and 20.3, “Disk firmware NDU” on page 290.
Verifying system load
Perform NDUs only when processor and disk activity are as low as possible. The upgrade
process requires one controller to assume the load that is normally handled by both
controllers. By minimizing the system load, you reduce the risk of host I/O requests being
delayed or timed out.
Before a Data ONTAP NDU is started, monitor processor and disk usage for 30 seconds with
the following command at the console of each storage system controller:
sysstat -c 10 -x 3
Avoid having the values in the CPU and Disk Util columns above 50% for all 10
measurements that are reported. Make sure that no other load is added to the storage system
until the upgrade completes.
Synchronizing date and time
Make sure that the date and time are synchronized between the two controllers. Although
synchronized time is not required, it is important in case an issue arises that requires
examining time- and date-based logs from both controllers.
Connecting to the storage controllers
By using serial cables, a console server, and the system's remote LAN module (RLM) or a
baseboard management controller (BMC), open a terminal session to the console port of the
two storage controllers.
Network connections to the controllers are lost during takeover and giveback operations.
Therefore, telnet, SSH, and FilerView sessions do not work for the NDU process.
20.1.6 Steps for major version upgrades NDU in NAS and SAN environments
The procedural documentation for running an NDU is in the product documentation on the
IBM Support site. See the “Upgrade and Revert Guide” of the product documentation for the
destination release of the planned upgrade.
For example, when an NDU from Data ONTAP 7.3.3. to 8.1 is performed, see the Data
ONTAP 8.1 7-Mode Upgrade and Revert/Downgrade Guide that is available at this website:
http://www.ibm.com/support/docview.wss?uid=ssg1S7003776