Hello Ruckus Forum,
As a Ruckus Technical Support Engineer, I would like to share my experience here on one the recent issues:
Issue Reported: vSZ-H upgrade failure from the release 6.0 to 6.1.
Issue Details: In this case, one of our customers was trying to upgrade v-SZ-H from 6.0 Release to 6.1 and the upgrade was stuck at 28% /upgrading system.
We could not ping the v-SZ instance after this state and must get into the v-SZ's VM console to find out what is wrong. We noticed the v/SZ had all three interfaces in the same Layer-2 Subnet/VLAN.
In 6.1 version, we have added a check to verify if all the interfaces are in same subnet, and if they are, v/SZ would not allow us creating an instance and an error will be triggered:
However, it does miss a pre-upgrade hook to throw a warning message when hitting such scenarios. This is eventually an improvement we would be adding to the v/SZ.
Following workarounds could help in such cases:
- Assign IP to control, cluster and management interface from separate VLAN/Subnets.(Preferred)
- If allocating separate VLAN to each interface is a challenge, create a fresh instance with single Interface and upload current config backup to the new instance. The config backup do not contain the network configuration and therefore it would help restoring the original config to a single interface instance.