Q: 2
An administrator has two identical clusters managed by separate Prism Central instances. The guest
VMs have pass-through GPUs. A scheduled maintenance is set for one of the clusters.
Which option would migrate VMs minimizing downtime?
Options
Discussion
B , saw a similar question in a practice set where Cross-Cluster Live Migration was called out for keeping downtime super low, even with pass-through GPUs. The async protection domains need failover which means more outage. I think B is the safest pick here, but wouldn't mind if someone disagrees.
C/D? D is a classic trap since async PDs aren't live, but B is probably what they're looking for.
D tbh
Nah, D's a DR option and will force downtime when you fail over. Similar exam questions point to B because Cross-Cluster Live Migration actually keeps VMs running with just seconds of interruption. Agree?
Its D for me. Asynchronous Protection Domains do support VM migration across clusters, and you can schedule the failover to line up with the maintenance window. Might not be zero downtime but the interruption is pretty minimal, especially if set up right. Not 100% on this since GPU passthrough might complicate things, open to corrections.
C, not B. Nutanix Move migration plan sounds like it could handle movement between clusters, and I thought it was designed for easy migrations. Maybe not perfect for GPU VMs, but I figured it would minimize downtime? Disagree if I'm off here.
I actually think D works here. Asynchronous Protection Domains let you migrate VMs to another cluster, and you can use them for planned moves during maintenance. Might not be zero downtime, but it should minimize interruption. Pretty sure that's how it's meant to be used, but correct me if I got that wrong.
D imo
I get why people think D, but async protection domains mean you have to fail over-so there’s downtime. For live or near-zero downtime, it’s B. Cross-Cluster Live Migration is designed for this situation (even with GPUs). Pretty sure that’s what Nutanix wants here, but happy to hear other views.
Option D since Asynchronous Protection Domains are meant for cross-cluster moves, but tricky since failover can cause some downtime.
Be respectful. No spam.