You can update the entire system in a coordinated process with no user intervention after
the update is initiated.
Follow these steps to update to version 8.1.0 or later from version 7.7.0 or later.
Note: If you are updating from version 8.3.0.0 or earlier to
any version before 8.3.1.2, configuring IP by using the cfgportip command with
gateway 0.0.0.0 is not supported. If you already have any configured IP using
cfgportip with gateway as 0.0.0.0, then either change the gateway address or
update directly to 8.3.1.2 or later.
Note: The update might stall if you have
hot-spare nodes that are connected via a Fibre Channel adapter and update software level 7.8.1.1 or
earlier to software level 7.8.1.2 or later. To avoid this issue, remove each Fibre Channel adapter
from the node that is being added to the system and check for the presence of a secure jumper. If
the secure jumper is present, remove the secure jumper to allow the firmware update to complete
successfully. After the update completes, return the secure jumper to the Fibre Channel adapter. For
more information, search for "Upgrade to Spectrum Virtualize software-only v7.8.1.2 requires
hardware check" at
www.ibm.com/support.
For the most recent information about
restrictions before you update, see the following website:
http://www.ibm.com/support/docview.wss?uid=ssg1S1001707
Before you update your system, review all of the topics in
Updating to understand how the process works. Allow
adequate time, such as up to a week in some cases to look for potential problems or known bugs.
If you encounter a memory DIMM failure to any node during the
update process, stop immediately and follow directions in Updating the system.
Additional information on updating the system is available at the
following website:
www.ibm.com/support
If
you want to update without host I/O, shut down all hosts before you start the update.
The
system update process can be done automatically or manually. Complete the following steps to update
the system automatically:
- In the management GUI, select .
- Click Test & Update.
- Select the test utility and the code package that you downloaded from the support site. The test
utility verifies that the system is ready to be updated.
- Click Next. Select Automatic update.
- Select whether you want to create intermittent pauses in the update to verify the process.
Select one of the following options:
- Fully automatic update without pauses
- Pause update after half of the nodes are updated
- Pause update before each node updates
- Click Finish. As the nodes on
the system are updated, the management GUI displays the progress for each node.
If you have hot-spare nodes that are configured on your system, the
hot-spare node assumes I/O operations from each node as it is updated. After the configuration node
is updated, all hot-spare nodes are updated.
Monitor the update information in the management GUI to determine when the process is
complete.
If
you selected a fully automated update without pauses and the process stalls during the update, click
Resume to continue with the update or click Cancel to
abandon the update and restore the previous level of code. You can also use the CLI command
applysoftware -resume to resume the update.
If you selected
to pause the update, click Continue to resume with the update. You can also
use the CLI command applysoftware -continue to resume the update.
Important: If you update your
system software to version 8.1.1 or later from a version earlier than 8.1.0, on a system where
you have already installed more than 64 GB of RAM, all nodes return from the update with an
error code of 841. Versions 8.1.0 and later allocate memory in a different way than previous
versions, so the RAM must be "accepted" again. To resolve the error, complete the following steps:
- On a single node, run the svctask chnodehw command. Do not run the command on more than one node at a
time.
- Wait for the node to restart and return without the error.
- Wait an additional 30 minutes for multipath drives to recover on the host.
- Repeat this process for each node individually until you clear the error on all nodes.