CompTIA Cloud+ Practice Test (CV0-004)

The correct answer is that a sensor's primary role is to gather data from the physical environment. In an IoT system, sensors are the devices that detect and measure physical properties like temperature, motion, or light and convert them into data. Actuators perform physical actions, the cloud backend processes data, and gateways manage communication; these are different roles within the IoT architecture.

In a synchronous RPC model, the client application makes a request and is blocked from continuing until the server processes the request and returns a response. This pattern is similar to a local function call. Asynchronous RPC would allow the client to continue processing without waiting for an immediate response. Event-driven architecture is a different integration pattern where services communicate by producing and consuming events. Batched calls refer to grouping multiple requests, which does not inherently define the blocking behavior of the call.

With a provider-managed (PaaS) database, the vendor-not the customer-installs and applies operating-system and database-engine patches on the underlying servers. Tasks that require direct access to hardware or the host OS (such as swapping disks, choosing RAID levels, or installing kernel modules) remain outside the customer's control. Administrators gain ease of maintenance but lose low-level hardware and OS access.

Comparing cryptographic sums of backup archives to their reference values is a reliable way to detect if data has been manipulated. A sum is generated from an unaltered baseline, then compared with a sum made of the backup file. If alterations have occurred, the sums will differ. Checking timestamps or file sizes might indicate changes but will not fully confirm content validity. Simply booting a snapshot may confirm the system’s functionality but may not reveal if data was changed in a less obvious way.

By housing environment details in files under version control, changes are easily tracked against a consistent baseline. Relying on ephemeral builds that dynamically generate settings can be tricky to reproduce. Spreadsheets are often disconnected from deployment pipelines, making them prone to errors. A single file housed on a dedicated server without version control can drift away from expected states, causing inconsistent configurations.

Working on a dedicated feature branch, pushing it to the remote, and opening a pull request that triggers automated tests allows code review and verification before integration. After the branch passes CI checks, it is merged into the default branch, protecting existing commits and keeping history clear. Force-pushing, bypassing tests, or deleting the default branch can overwrite colleagues' changes and obscure revision history.

Temporarily blocking the suspicious source addresses at the cloud firewall cuts off the attacker's access (containment) and, because the systems remain online, allows security staff to enable packet capture and collect volatile data and logs. Powering off hosts before containment could disrupt ongoing forensic data collection and violate the 15-minute containment window. Disabling MFA weakens security and does not address the compromise. Raising alert thresholds delays response and leaves the environment exposed.

A combination of local infrastructure for sensitive records and external resources for other data meets government mandates while allowing more freedom in scaling. Placing everything in a remote region violates location rules. Keeping functions in a fully isolated facility limits flexibility for non-sensitive data. Relying on encryption in an external region does not fulfill location requirements regarding sensitive data.

This method employs artificial intelligence to extract insights from multiple data formats. It uses learned algorithms and network structures to determine how to interpret imagery, transform text, and enable spoken responses. In contrast, distributing queries among clusters focuses on database efficiency, container orchestration handles runtime workloads, and a caching layer boosts content delivery without advanced pattern interpretation.

This technique accumulates modifications that have occurred since the last complete copy. Restoration is streamlined because administrators combine a single partial set with the prior complete copy. An incremental method tracks new changes after each smaller copy, which can require multiple sets during restoration. A synthetic full merges existing smaller copies with a prior complete copy, which is not the same approach. Mirroring keeps data synchronized all the time, which is not designed for scheduled collection of changes.

Removing default user accounts and shutting off unneeded services decreases exposure to attacks by limiting entry points. Installing more CPU resources gains performance but does not address security. Opening additional ports makes intrusion more likely. Short password policies weaken protection by simplifying unauthorized access.

Shortening the scope's lease duration causes unused addresses to return to the pool more frequently. This frees addresses for new clients without requiring subnet changes. Disabling dynamic DNS, adding a default-gateway option, or reserving additional addresses do not release any new leases and, in the last case, would further reduce the available pool.

Granular, or file-level, recovery enables administrators to restore exactly the objects that were lost-such as a single file, folder, or mailbox item-without rolling back the entire VM, volume, or application database. This precision reduces downtime and eliminates the risk of overwriting data that is still current. Incremental backup restoration focuses on how data is captured, not on selective restore. A full VM snapshot revert would restore the whole virtual machine to an earlier state, potentially overwriting recent data. Continuous replication failover swaps production to a replica VM for availability and is not intended for item-level recovery.

An environment error referencing a threshold limit often indicates a built-in resource restriction, also referred to as a usage threshold. Requesting a higher limit from the hosting vendor resolves the numeric constraint. The other proposed actions would not affect a quota restriction, which must be adjusted at the provider level.

A perimeter device that monitors incoming flows by checking IP addresses and ports is an example of a firewall solution. It inspects traffic and blocks or allows it based on established rules, stopping port scanning and other attacks. Using static routes or encryption does not enforce rule-based filtering. A content distribution layer addresses performance issues, and it does not inherently stop scanning attempts.

Enabling a legal hold guarantees that logs will stay untouched for the entire duration of the case. Relying on weekly backups guided by administrators or using pre-set removal timelines may create gaps, allowing accidental deletion. Saving logs in a compressed form in a separate archive might still depend on internal policies that allow for early removal. A legal hold is designed to address uncertain schedules by placing records under controlled preservation.

RPM (Red Hat Package Manager) is the native packaging format for Red Hat-based systems like RHEL. It has built-in support for GPG signatures to ensure package authenticity and integrity. RPMs also have a standardized way of embedding version information in their metadata, which tools like YUM or DNF use for version tracking and dependency management. A tar archive with a checksum only verifies file integrity, not authenticity, and lacks automated version tracking. A Debian package is designed for Debian-based distributions (like Ubuntu) and is incompatible with RHEL's native tools. Compiling from source code relies on custom scripts and does not offer a standardized, integrated system for signing and versioning like RPM.

Endpoint protection (sometimes called host-based antivirus, EDR, or HIDS/HIPS) runs a software agent on each machine. The agent scans files and processes locally, compares activity against threat intelligence, and can quarantine or block malware without relying on external network defenses. A perimeter firewall only inspects traffic that reaches the edge, SIEM CPU alerts focus on performance anomalies, and data classification governs information access but does not detect or remove malicious code from a host.

Defining infrastructure in code allows the same configuration to be applied repeatedly with identical results. This repeatability prevents configuration drift that can occur with manual steps. Elasticity refers to scaling resources up or down, multi-tenancy involves hosting multiple customers in the same environment, and vendor lock-in describes dependence on a single provider-none of which specifically address recreating identical environments.

Creating a security group for the marketing department, assigning all necessary permissions to that group, and then adding new employees to the group is the most efficient and scalable method. This approach, known as group-based access control (GBAC), ensures consistency and simplifies administration. Cloning an existing user's account is risky because it can lead to privilege creep, where unnecessary permissions accumulated by the original user are passed on. Assigning permissions individually is time-consuming, error-prone, and does not scale well. Requiring individual requests for each resource places the administrative burden on the end-user and approvers and does not represent an efficient provisioning strategy.