NDB Multi-Cluster is a feature that allows multiple Nutanix clusters to share a common NDB service for database management and automation. NDB service consists of two components: NDB Server and NDB Agent. NDB Server is the central component that provides the web UI, REST API, and database orchestration logic. NDB Agent is thecomponent that runs on each Nutanix cluster and communicates with the NDB Server to perform database operations.
When NDB Multi-Cluster is enabled on Cluster A, it becomes the primary cluster that hosts the NDB Server VM. Cluster B, which is registered with NDB, becomes a secondary cluster that hosts only the NDB Agent VM. Therefore, Cluster A has both NDB Server and NDB Agent, while Cluster B has only NDB Agent. This configuration allows Cluster B to leverage the NDB service running on Cluster A for database management and automation.
References:
* Nutanix Database Automation (NCP-DB) Course Details, Section 3.1: NDB Multi-Cluster Overview
* Nutanix Database Automation (NCP-DB) Certification Details, Objective 3.1: Configure NDB Multi-Cluster
* Nutanix Database Automation (NCP-DB) YouTube Playlist, Video 3.1: NDB Multi-Cluster Overview

To create an NDB Software Profile Version for PostgreSQL, it is required to have an installer package for the database software. This package forms the basis of the software profile,enabling the deployment and management of PostgreSQL instances.References:Nutanix Database Automation documentation, particularly in the area of software profile creation and management for PostgreSQL.

The correct answer is C because the NDB Database Agent service is responsible for scheduling and monitoring the NDB-related operations, such as provisioning, patching, cloning, and restoring databases. The NDB Database Agent service is deployed on the NDB server and on each of the database server VMs provisioned by NDB. It contains the tools required to perform each task and communicates with the NDB server through the NDB API. The other options are not correct because they describe different components or functions of the NDB solution. Option A is not correct because the transaction logs are not stored by the NDB Database Agent service, but by the NDB Time Machine service, which is a distributed file system that stores the database snapshots and logs. Option B is not correct because the source databases are not hosted by the NDB Database Agent service, but by the database server VMs, which are virtual machines that contain an instance of the database engine software and the database files. Option D is not correct because the NDB Database Agent service does not contain an instance of the database engine software, but only the tools required to perform the NDB-related operations. The database engine software is installed on the database server VMs, which are provisioned by NDB using the profiles defined by the administrator. References: Nutanix Database Management & Automation (NDMA) course, Nutanix Certified Professional - Database Automation (NCP-DB) certification, Nutanix NCP-DB Certification Exam Syllabus and Study Guide, Nutanix Support & Insights

The error message in the exhibit indicates that the operation failed during the "Create and Register Database Server VMs" step because "Provisioning of all the observers simultaneously took more than two hours." This type of error is associated with MySQL, as it involves observers which are a part of MySQL Group Replication, used for ensuring high availability1. The other options are not related to the error message, as they do not use observers or Group Replication for clustering. References:
* 1: Nutanix Database Automation (NCP-DB) course, Module 5: Database High Availability, Lesson 5.2:
MySQL Group Replication, slide 7

The administrator should perform a rolling upgrade, applying the patch to the secondary members first, followed by the primary member, to accomplish the task of patching a MongoDB server cluster within an NDB environment. A rolling upgrade is a method of applying patches or updates to a cluster without downtime or interruption of service. The administrator can use the NDB patching feature to perform a rolling upgrade on a MongoDB server cluster, which consists of a primary member and one or more secondary members that form a replica set. The NDB patching feature allows the administrator to select the software profile version, the database parameters profile, and the network profile for the patching operation. The NDB patching feature also automates the steps of the rolling upgrade, such as:
* Step 1: The administrator initiates the patching operation on the NDB instance, and selects the MongoDB server cluster to be patched.
* Step 2: The NDB instance verifies the prerequisites and compatibility of the patching operation, and creates a pre-patch snapshot of the MongoDB server cluster.
* Step 3: The NDB instance applies the patch to the first secondary member of the MongoDB server cluster, and waits for the patching to complete successfully.
* Step 4: The NDB instance verifies the status and functionality of the patched secondary member, and repeats the patching process for the remaining secondary members of the MongoDB server cluster, one at a time.
* Step 5: The NDB instance performs a failover of the primary member to one of the patched secondary members, and applies the patch to the original primary member.
* Step 6: The NDB instance verifies the status and functionality of the patched primary member, and performs a failback of the primary member to the original primary member, if desired.
* Step 7: The NDB instance creates a post-patch snapshot of the MongoDB server cluster, and completes the patching operation.
Performing a rolling upgrade, applying the patch to the secondary members first, followed by the primary member, is the recommended and best practice method of patching a MongoDB server cluster within an NDB environment, as it ensures the high availability, consistency, and performance of the MongoDB server cluster and the databases.
Performing a rolling upgrade, applying the patch to the primary member first, followed by the secondary members, is not a valid or feasible method of patching a MongoDB server cluster within an NDB environment, as it would cause downtime, data loss, and inconsistency of the MongoDB server cluster and the databases. Applying the patch to the primary member first would disrupt the replication and synchronization of the MongoDB server cluster, and would require manual intervention and recovery steps to restore the MongoDB server cluster to a functional state.
Applying the patch to all nodes at once is not a valid or feasible method of patching a MongoDB server cluster within an NDB environment, as it would cause downtime, data loss, and inconsistency of the MongoDB server cluster and the databases. Applying the patch to all nodes at once would require shutting down the entire MongoDB server cluster, and would expose the MongoDB server cluster and the databases to potential errors, failures, and corruption during the patching process.
Disabling the replica set while patching is not a valid or feasible method of patching a MongoDB server cluster within an NDB environment, as it would cause downtime, data loss, and inconsistency of the MongoDB server cluster and the databases. Disabling the replica set while patching would break the replication and synchronization of the MongoDB server cluster, and would require manual intervention and recovery steps to re-enable the replica set and restore the MongoDB server cluster to a functional state.
References:
* Nutanix Database Management & Automation Training Course, Module 5: Nutanix Era Operations, Lesson 5.1: Nutanix Era Operations, slides 11-12, 15-16.
* Nutanix Database Management & Automation Training Course, Module 5: Nutanix Era Operations, Lesson 5.3: Nutanix Era Patching, slides 5-9.
* Nutanix Database Management & Automation Training Course, Module 5: Nutanix Era Operations, Lesson 5.4: Nutanix Era Patching Lab, slides 5-10.
* Nutanix Database Management & Automation Training Course, Module 7: Nutanix Era Troubleshooting, Lesson 7.1: Nutanix Era Troubleshooting, slide 6.