Saturday, 18 March 2017

Upgrade ONTAP 8.3.1 to 9.1 Walkthrough

Back in October I posted Cluster Software Update Essentials which outlined 3 different upgrade methods (really, methods 1 & 2 were the same, just GUI v CLI.) In this post I’m going to run through the method I’d use - “cluster image update” (for anything other than single node clusters.) Upgrading from 8.3.0 to 9.1 is a one step process.


1) Verify your environment is supported for ONTAP 9.1
- and make remediation’s where required.
Note: This includes FC SAN hosts, boot from SAN hosts, OFFTAP products…
2) Verify platform support, and root volume sizes
3) Verify cluster health via AutoSupport
4) Verify cluster configuration with Config Advisor
5) Verify cluster health in OnCommand Unified Manager.
6) Thoroughly go over the Upgrade Advisor outputs and do the appropriate!
7) Download the ONTAP 9.1 image
8) Serve the 91P1_q_image.tgz file on a web server.

Note: 9.1P1 is the recommend release at the time of writing.

Cluster Image Package Get

Use the highlighted commands below to get the new image on the cluster, and perform validation:

cluster1::> cluster image show
Node        Current Version Installation Date
----------- --------------- -------------------
cluster1-01 8.3.1           9/17/2015 16:54:16
cluster1-02 8.3.1           9/17/2015 16:55:04

cluster1::> cluster image package show-repository
There are no packages in the repository.

cluster1::> cluster image package get -url

Software get started on node cluster1-01
Downloading package. This may take up to 10 minutes...

cluster1::> cluster image package show-repository
Package Version Package Build Time
--------------- ------------------
9.1P1           2/14/2017 13:14:46

cluster1::> cluster image validate -version 9.1P1

It can take several minutes to complete validation...

Pre-update Check                  Status
----------------------------      ------
Aggregate plex resync status      OK
Aggregate status                  OK
Autoboot Status                   OK
Broadcast Domain status           OK
CIFS status                       OK
CPU Utilization Status            OK
Cluster health status             OK
Cluster quorum status             OK
Data ONTAP Version Status         OK
Disk status                       OK
High Availability status          OK
Jobs Status                       OK
LIF failover                      OK
LIF load balancing                OK
LIFs not hosted                   OK
LIFs on home node status          OK
Manual checks                     Warning: Manual validation checks need to be performed. Refer to the Upgrade Advisor...
MetroCluster configuration status OK
NDMP status                       OK
NFS netgroup check                OK
Platform status                   OK
Previous Upgrade Status           OK
SAN LIF status                    OK
SAN status                        OK
Security Config SSLv3 check       OK
SnapMirror status                 OK
Snapshot copy count check         OK
Volume move status                OK
Volume status                     OK
Overall Status                    Warning

Update the first HA Pair

Note the following warning from Upgrade Advisor regards SnapMirror:
“To prevent SnapMirror transfers from failing, you must suspend (Quiesce) SnapMirror operations and upgrade destination nodes before upgrading source nodes.
(i) Suspend SnapMirror transfers for a destination volume
(ii) Upgrade the node that contains the destination volume
(iii) Upgrade the node that contains the source volume
(iv) Resume the SnapMirror transfers for the destination volume”

This is a good point at which to send AutoSupports (and verify they have sent)::>

autosupport invoke -node * -type all -message "maint=4h upgrading to ONTAP 9.1P1"
autosupport history show -node cluster1-01
autosupport history show -node cluster1-02

Note: If this is a 4-node or greater cluster, you might like to make sure Epsilon is not on the HA-Pair being upgraded (cluster image update will handle this.)

set adv
cluster show
cluster modify -node EPSILON_NODE -epsilon false
cluster modify -node NEW_EPSILON_NODE -epsilon true

Update the HA-Pair:

cluster1::> cluster image update -version 9.1P1 -nodes cluster1-02,cluster1-01

Starting validation for this update. Please wait...
It can take several minutes to complete validation...
Warning: Validation has reported warnings. Do you want to continue? {y|n}: y
Starting update...

Note: You don’t need to do 1 HA-pair at a time, “cluster image update” is designed to update the entire cluster.

Verifying the Update Process

You should be connected to Service-Processors to observe the update process.
Note: If you have aggregates serving CIFS, you may need to do a “storage failover giveback -ofnode NODE -override-vetoes true”

cluster image show-update-progress
storage failover show-takeover
storage failover show
storage failover show-giveback

Update Subsequent HA Pairs

Follow the above to update subsequent HA Pairs as required.


Verify the cluster version has increased (the cluster version only updates when all nodes are upgraded).
If you suspended SnapMirrors re-enable them.
Send AutoSupports::>

snapmirror resume -source-path {SRC_PATH} -destination-path {DEST_PATH}
autosupport invoke -node * -type all -message "maint=END finished upgrade to ONTAP 9.1P1"

And test your environment.

Appendix: (Some) Supported things with ONTAP 9.1

Just a personal reference to save some IMT results (correct at 2017.03.18):

OnCommand Unified Manager 7.0
OnCommand Unified Manager 7.1
OnCommand Performance Manager 7.0
OnCommand Performance Manager 7.1
SnapDrive 7.1.3 for Windows
SnapDrive 7.1.4 for Windows
SnapCenter Host Plug-in 1.1 for Microsoft Windows
SnapCenter Host Plug-in 1.1 for UNIX
SnapCenter Host Plug-in 2.0 for Microsoft Windows
SnapCenter Host Plug-in 2.0 for UNIX
SnapCenter Host Plug-in 2.0 for VMware vSphere
Virtual Storage Console 6.1
Virtual Storage Console 6.2
Virtual Storage Console 6.2.1
SnapDrive 5.2.2 for Unix*
SnapDrive 5.3 for Unix*
SnapDrive 5.3.1 for Unix*

SnapCenter Application Plugin has dependencies with:
- SnapCenter Host Plugin (if 1.1 use 1.1/if 2.0 use 2.0)
- SnapCenter Server
- Host Application

SnapCenter Server has dependencies with:
- SnapCenter Host Plugin (if 1.1 use 1.1/if 2.0 use 2.0)
- SnapCenter Application Plugin
- Virtual Storage Console (VMware)

SnapManager for Exchange has dependencies with:
- SnapDrive (for Windows version)
- Host OS (Windows Server version)
- Host Application (Exchange Server version)

CN1610 Cluster Switch FASTPATH and RCF 1.2
Cisco NX3132V Cluster Switch NX-OS 7.0(3)I4(1) and RCF 1.1
Cisco NX5596 Cluster Switch NX-OS 7.1(1)N1(1) and RCF 1.3
Cisco NX5020 Cluster Switch NX-OS 5.2(1)N1(8b) and RCF 1.3
Cisco NX5010 Cluster Switch NX-OS 5.2(1)N1(8b) and RCF 1.3

No comments:

Post a Comment