Question

Alternative to moving config to standard format? Splitting stretch clusters feasible?

AdelMarei
Userlevel 1
Badge +1

Hello,

We have an unsupported configuration on 2016 stretch cluster:

  1. 3 NIC teaming with LACP → should be 1 NIC Teaming with a switch independent

The goal is to upgrade to server 2022, we have the manual option from TAC to upgrade to 2019 than 2022.

Note based on what we have seen we have the following points:-

From the doc. https://portal.nutanix.com/page/documents/details?targetId=Acropolis-Upgrade-Guide-v6_6:upg-cluster-upgrade-recommend-hyperv-r.html

  • Upgrade to Windows Server 2022 Hyper-V from an LACP enabled Hyper-V 2019 cluster is not supported.

  • Enabling Link Aggregation Control Protocol (LACP) for your cluster deployment is supported when upgrading hypervisor hosts from Windows Server 2016 to 2019.

On the other hand in doc. https://portal.nutanix.com/page/documents/kbs/details?targetId=kA00e000000LLI8CAO  the LACP is not supported on 2016 and 2022, since we have seen the following error as seen below (the ISO image was 2019 for the upgrade, not 2022 we can not upgrade to 2022 directly):-

  • Non-standard network settings, cannot proceed with hypervisor upgrade;

 

Q. Do we have an alternative option(s) other than moving the configuration to a standard format? 

Can we split a stretch cluster into 2 standard clusters and then join both sides again as a stretch cluster?  - if my information is correct we can not, since the cluster configuration is created from the foundation step (we have the witness VM network setup we can not disable it and enable it at will).  

This topic has been closed for comments
Terms & Conditions