RF3 (Replication Factor 3) stores three copies of data across the cluster. This provides the highest level of availability and fault tolerance, as the cluster can withstand the simultaneous failure of two copies (e.g., two nodes or disks) without data loss. This is crucial for the stated requirement of no data loss during a rebuild, as a rebuild in RF2 could be vulnerable if another failure occurs.

The ‘Table’ view within a VM’s ‘Monitor’ page in Prism Central allows you to add and view multiple performance metrics (like IOPS, latency, throughput) for the VM, its attached disks, and the related storage containers side-by-side in a single, customizable table. This direct correlation is the most efficient way to identify if high VM latency aligns with high latency on a specific disk or container.

Curator is Nutanix’s distributed map-reduce job that manages storage optimization. After adding nodes, running Curator (which happens automatically on a schedule but can be triggered manually) is crucial. It rebalances data (I/O and capacity) across all nodes in the cluster, including the new ones, ensuring even distribution and maintaining performance efficiency.

For replication to function, the target/remote storage container must have an SSD tier configured and enabled, even if the primary is all-SSD and the remote is hybrid. The Nutanix Distributed Storage Fabric (DSF) uses the SSD tier to ingest replicated writes (acting as a landing zone) before later tiering data to HDD if applicable. Without an SSD tier, replication cannot be configured.

Projects are the core Nutanix capability for self-service IT and resource governance. By creating a Project, you can define resource quotas (compute, storage), network configurations, and most importantly, publish standardized application Blueprints. You can then assign users to the project with a ‘Developer’ or ‘Consumer’ role, allowing them to deploy from published blueprints within the defined quotas without accessing broader infrastructure settings.

High ‘Memory Usage’ with low ‘Memory Pressure’ is a classic sign of efficient memory management in a virtualized environment. The Nutanix CVM uses intelligent ballooning and memory sharing techniques to actively use most of the available physical RAM to cache data (improving storage performance). ‘Memory Pressure’ is the key metric; if it is low, the system is effectively managing its resources, and the high usage is by design for optimization.

Nutanix provides storage-level encryption at the container granularity. Using the ‘one-click encryption’ feature on the specific storage container (via Prism Element or Central) encrypts all data within that container, including existing data and all future writes. This is the most direct, manageable, and performance-efficient method native to the Nutanix platform, satisfying the at-rest encryption requirement for the application.

While a single Prism Central (PC) instance can manage many clusters, Nutanix best practices recommend considering multiple PC instances for very large-scale deployments (often cited as around 50 clusters or more). This is for reasons of scalability, manageability, fault isolation, and to avoid a single point of management failure. PC should be deployed on a dedicated management cluster, not on a production cluster it manages.

This sequence follows a logical troubleshooting path from the hypervisor host inward. First, check if AHV host firewall rules are blocking traffic. Then, verify the upstream connectivity by checking the state of the network controller’s (OVS) uplink to the physical network. Finally, inspect the virtual switch port specific to the VM for errors or misconfiguration. This methodically rules out host, uplink, and port-specific issues.

RF3 (Replication Factor 3) stores three copies of data across the cluster. This provides the highest level of availability and fault tolerance, as the cluster can withstand the simultaneous failure of two copies (e.g., two nodes or disks) without data loss. This is crucial for the stated requirement of no data loss during a rebuild, as a rebuild in RF2 could be vulnerable if another failure occurs.0