Cluster Software Update Essentials (8.3.1 ->)

In the following post I’ll review the methods and essential commands for doing an upgrade from Clustered Data ONTAP 8.3.1. Specifically, we’re going to Clustered Data ONTAP 8.3.2P5 in the examples below. Note: In real world situations consult Upgrade Advisor from AutoSupport.

In the lab, I have a HA-pair running 8.3.1.

Method 1: Automated Non-Disruptive Upgrade via OnCommand System Manager GUI

1) Log into OnCommand System Manager for the Cluster containing the HA-pair(s) that you want to upgrade
2) Expand the Cluster pane
3) Expand the Cluster object
4) Click on “Cluster Update”
5) Follow through the wizard
- first adding the update Software Image
- then Validate
- and finally Update

Image: Automated Non-Disruptive Upgrade via OnCommand System Manager GUI in 8.3.1
Method 2: Cluster Image commands

From the Clustershell::>


cluster image show
cluster image package show-repository
cluster image package get -url http://192.168.0.5:8080/832P5_q_image.tgz


The above triggers a “software get” on each node in the cluster.


cluster image package show-repository
cluster image validate -version 8.3.2P5


Once the validation has completed, review any warnings, and when ready to proceed, continue.


cluster image update -version 8.3.2P5
cluster image show-update-progress
cluster image show


Method 3: Old School Method (still used for Major Upgrades) - System Node Image commands


system node image show
system node image get -node * -package http://10.0.0.1/832P5_q_image.tgz -background true
system node image show-update-progress -node NODENAME
system node image update -node * -package 832P5_q_image.tgz -background true
system node image show-update-progress -node NODENAME
system node image show
system node image modify {-iscurrent false} -isdefault true
system node image show
storage failover show
cluster show
storage failover takeover -ofnode NODE1
storage failover show-takeover


Wait until the node is waiting for giveback.


storage failover giveback -ofnode NODE1
storage failover show-giveback


Wait until each node says “No aggregates to give back”


cluster show
storage failover takeover -ofnode NODE2
storage failover show-takeover


Wait until the node is waiting for giveback.


storage failover giveback -ofnode NODE2
storage failover show-giveback


Wait until each node says “No aggregates to give back”. Verify the cluster version has updated:


cluster show
version


APPENDIX 1: “Cluster Image Update” switches

Image: Cluster Image Update switches

APPENDIX 2: Verification Checks Automated NDU/Cluster Image Update does

Note: The below are taken from an 8.3.2 system.

- Aggregate plex resync status
- CIFS status
- Cluster health status
- Cluster quorum status
- Disk status
- High Availability status
- LIF status
- LIF on home node status
- Manual checks (A warning refers you to Upgrade Advisor)
- MetroCluster configuration status
- NFS netgroup check
- SAN LIF status
- SAN status

Comments