The ‘Inline Compression’ feature with a ‘Performance’ setting is designed for this scenario. It provides compression (saving space) but prioritizes low latency by compressing data as it is written, avoiding the performance overhead of post-process deduplication/compression. RF2, all-flash, and enabling both dedupe/compression add cost or latency and don’t optimally balance the requirements.

The Controller VM (CVM) is responsible for serving storage I/O to VMs on its host. A high ‘IO Wait’ on the CVM specifically indicates it is waiting on underlying physical storage devices (SSDs/HDDs in the disk group). This points directly to a storage media bottleneck as the primary issue to investigate.

For cross-site migration over Layer 3 with minimal downtime, using a Protection Domain for replication and a planned failover is the standard Nutanix approach. This replicates data asynchronously and allows for a controlled switchover. Leap is for DR, not general migration. ‘Move’ in PC is for within a cluster. vMotion is not applicable to native AHV.

The statement is false. RF3 (three copies of data) provides higher fault tolerance than RF2 (two copies) but consumes more raw storage for the same amount of logical data. Therefore, changing from RF2 to RF3 for an existing container would decrease, not increase, its usable capacity.

To provide true physical isolation and prevent contention at the hardware level, you should create separate virtual switches in AHV, each with their own dedicated physical NICs (uplinks). You would then attach the VM’s vNICs to the appropriate virtual switch. VLANs (Option A) provide logical separation but share the same physical uplinks.

The ‘Hypervisor Memory’ reservation in the CVM configuration is memory reserved from the physical host for the hypervisor (AHV) itself. Setting this too high unnecessarily reduces the total host memory pool available for allocation to guest VMs, impacting VM density and performance.

Nutanix requires a successful cluster health check to be run immediately before initiating a software update. This ensures the cluster is in a stable state and can safely proceed with the upgrade. If a health check has failed (or was never run), the update option will be disabled until the issue is resolved and a health check passes.

A snapshot is a point-in-time, read-only reference to a VM’s disk state that uses redirect-on-write technology for space efficiency. A clone is a complete, independent, and writable copy of a VM or disk that consumes storage equal to its full size at creation. Both can be local or remote, scheduled/manual, and have consistency options.

Nutanix clusters are designed to handle disk failures automatically using RF or Erasure Coding. The immediate action is not to touch the hardware but to open a support case. Prism will then provide a guided procedure (often via LCM) to safely remove the failed disk’s metadata and prepare for the physical replacement, ensuring no data loss or service disruption.

The ‘Inline Compression’ feature with a ‘Performance’ setting is designed for this scenario. It provides compression (saving space) but prioritizes low latency by compressing data as it is written, avoiding the performance overhead of post-process deduplication/compression. RF2, all-flash, and enabling both dedupe/compression add cost or latency and don’t optimally balance the requirements.0