Full Access Nutanix NCSE-Core Tutorials

To address the performance concerns of a prospective customer evaluating a hybrid configuration versus an all-flash array, the following Nutanix features should be discussed:

Data Locality:

Nutanix ensures that data is kept local to the node where the VM is running. This reduces latency and increases performance by minimizing the distance data must travel over the network.

Automatic Data Tiering:

Nutanix can automatically move frequently accessed data to faster storage tiers (e.g., SSDs) and less frequently accessed data to slower storage tiers (e.g., HDDs). This ensures that critical data gets the performance benefits of faster storage without the need for an entirely all-flash array.

References:

Nutanix Bible​

A prospect is looking for a Nutanix solution for a mission-critical application that needs to sustain two simultaneous node outages and prefers the cluster to fully rebuild both data and metadata in the event of hardware issues. Here is the detailed explanation:

Requirement Analysis:

Sustain two simultaneous node outages.

Fully rebuild both data and metadata after hardware issues.

Feature Analysis:

Nutanix clusters use Replication Factor (RF) to ensure data protection and availability. For RF3 (Replication Factor 3), which protects against two simultaneous node failures, a minimum of 5 nodes is required.

To fully rebuild both data and metadata after failures, the system needs an additional node to ensure there is enough redundancy and performance.

Conclusion:To achieve RF3 and sustain two simultaneous node failures while allowing full rebuild of data and metadata, a minimum of 6 nodes is required.

References:

Nutanix documentation on cluster resiliency and RF configurations​

Understanding the Scenario:

The customer has mission-critical applications that need to be backed up to a remote location with a latency of 7 ms.

Nutanix Technology Options:

AsyncDR: Provides asynchronous replication with typically higher RPO (Recovery Point Objective) than NearSync.

NearSync: Offers near-synchronous replication with RPO as low as 1 minute, suitable for low-latency connections like the 7 ms latency mentioned.

Metro Availability: Provides synchronous replication with zero RPO but requires a higher bandwidth and lower latency connection (typically sub-millisecond).

Leap: A comprehensive DR orchestration solution that includes AsyncDR and NearSync but is not a standalone replication technology.

Evaluation of NearSync:

NearSync is designed to provide low RPO (as low as 1 minute) by taking frequent lightweight snapshots.

It is well-suited for environments with moderate latency (like 7 ms) and can handle the backup requirements efficiently without the need for extremely low latency and high bandwidth required by synchronous replication solutions like Metro Availability​​​​.

Conclusion:

Given the latency of 7 ms and the need for short RPO, NearSync is the most suitable Nutanix technology to provide the shortest RPO.

References: Nutanix Bible and Nutanix NDB User Guide details on NearSync and its operational parameters.

To ensure the solution does not exceed 80% CPU utilization when a node is placed in maintenance mode, the SE should:

Access Sizer: Open the Nutanix Sizer tool.

Create New Scenario: Start a new scenario for the project.

Define Workload: Enter the details of the current workload, specifying resource requirements.

Customize CPU Threshold:

In the sizing options, customize the CPU threshold to 80%. This setting ensures that the calculated resources account for keeping CPU utilization at or below 80% when a node is taken out of the pool for maintenance (N+0).

Review Configuration: Verify that the proposed configuration meets the 80% CPU utilization requirement under maintenance conditions.

This approach ensures that the solution is appropriately sized to handle the specified workload without exceeding the desired CPU utilization threshold.

References:

Nutanix Sizer documentation and guides.

Best practices for configuring CPU thresholds and resource allocation.