CompTIA Linux + XK0-006 (V8) Exam Practice Test 3

Runtime configuration applies changes immediately but does not persist across a reload or reboot unless separately saved to permanent configuration. Permanent configuration persists changes but does not apply them until a reload, rich rules and zones are structural concepts rather than persistence scopes, and ports versus services concerns how a specific rule is expressed rather than its persistence.

A load average consistently higher than the number of CPU cores, with processes waiting to be scheduled, is specifically a high load average issue. Swapping concerns disk-based virtual memory usage due to low RAM, packet drops and jitter concern network-layer issues, and low throughput concerns reduced data transfer rate rather than CPU scheduling contention specifically.

An SELinux issue, such as a restrictive policy or incorrect context, can override otherwise correct ACL entries and standard permissions, causing access denial despite seemingly correct configuration. File and directory permission issues generically would typically be resolved by fixing the ACL itself, account access concerns whether the account can log in at all, a certificate issue concerns TLS validation, and a remote access issue concerns SSH or similar connectivity rather than local file access being blocked by policy.

Inconsistent link states between interfaces that are supposed to be working together as a redundant bonded pair is specifically a bonding issue. MTU mismatch concerns packet size handling, MAC spoofing concerns falsified hardware addresses, subnet misconfiguration concerns incorrect network addressing, and link down describes a single interface losing connectivity entirely rather than inconsistency between bonded interfaces specifically.

Two devices assigned the same IP address causing unpredictable connectivity failures is an IP conflict. A DHCP issue concerns automatic address assignment failing rather than a manual duplicate, a DNS issue concerns name resolution, a link negotiation issue concerns speed/duplex mismatch, and a dual stack issue concerns problems specifically arising from running IPv4 and IPv6 simultaneously rather than a duplicate static address.

New files failing to write with an explicit filesystem-full message is OS filesystem full, which can persist even after storage-layer expansion if the filesystem itself was never grown to use the new space. Inode exhaustion produces a similar symptom but relates to running out of file-tracking structures rather than raw space, a partition not writable concerns a mount or permission issue, a package dependency issue concerns software installation, and a kernel panic halts the entire system rather than just blocking writes.

Defining a threshold for a metric and configuring an alert to fire automatically when it is crossed is exactly this threshold-and-alert monitoring configuration pairing. Log aggregation centralizes log data rather than metric thresholds, health checks are periodic probes of service availability, webhooks are one specific delivery mechanism for a notification, and SNMP traps are one specific protocol-based alerting mechanism rather than describing the general threshold-alert configuration concept.

Requiring human review of AI-generated code before merging is specifically the practice of avoiding copy/paste without review and quality assurance. Prompt engineering concerns crafting effective inputs to the AI, data governance generically concerns broader data handling policy, local models generically concerns where the AI runs rather than output review, and adhering to corporate policy generically is a broader category this specific review practice supports rather than the practice itself.

git pull downloads changes from a remote repository and immediately merges them into the current branch in one step. git fetch downloads changes without merging them, allowing separate review first, git clone copies an entire repository for the first time, git log shows commit history, and git diff shows differences rather than downloading and merging remote changes.

A while loop repeatedly executes a block of code for as long as a condition remains true, checking before each iteration. A for loop iterates over a fixed list of items, until loops repeat while a condition is false, case matches specific values, and if performs a single conditional branch rather than repeated iteration.

OpenTofu, like Terraform, maintains a state file tracking the real-world state of deployed resources, comparing it against desired configuration to plan changes. Ansible is generally agentless and does not maintain a persistent state file in this way, Puppet uses a client-server model with its own state tracking approach, and Kickstart and cloud-init handle unattended OS deployment rather than ongoing infrastructure state tracking.

HMAC combines a hash function with a secret key, so that even knowing the hash algorithm, an attacker without the key cannot easily forge or brute-force valid values as feasibly as with a keyless hash. SHA-256 alone is a keyless cryptographic hash function, GPG handles file/message encryption and signing, LUKS2 provides disk encryption, and Argon2 is a password-hashing function without HMAC's shared-key construction specifically.

NOEXEC prevents a command from spawning further subprocesses, limiting what can be done even with elevated privileges, while combining NOPASSWD with a rule granting ALL commands gives unrestricted, passwordless root access. The wheel group is a common administrative group convention, visudo is the safe editor for sudoers files, and sudo -i starts an interactive root shell rather than describing either of these two specific configuration behaviors.

AArch64 refers to the 64-bit ARM architecture increasingly used in cloud instances for cost and power efficiency. x86_64/AMD64 and x86 refer to Intel/AMD architectures instead, and RISC-V is a separate, open instruction set architecture distinct from ARM.

Pruning removes unused, dangling images (and similarly, stopped containers or unused volumes) to reclaim disk space without affecting resources currently in use. Pulling images downloads images from a registry, tags label specific image versions, layers are the filesystem diffs that make up an image, and building an image creates a new one from a Dockerfile rather than cleaning up unused resources.

hostnamectl manages the system hostname, timedatectl manages date, time, and timezone settings, resolvectl manages DNS resolution configuration, and systemd-analyze reports boot performance, each covering a distinct systemd-managed administrative area.

A zombie process has completed execution but remains in the process table because its parent has not yet reaped its exit status. Running describes an actively executing process, blocked and sleeping describe processes waiting on a resource or event, and stopped describes a process paused by a signal rather than one that has already finished executing.

lastlog shows the most recent login time for each user account, including terminal or host information. last shows a broader history of login sessions for all users, who and w show currently logged-in users, and whoami simply displays the current user's own username rather than login history.

A hard link shares the same underlying inode as the original file, so the data remains accessible even if the original directory entry is deleted, as long as at least one link remains. A symbolic link instead points to a path and breaks if the original is deleted, and block, character, and special character devices are device file types in /dev rather than link types.

VirtIO is a paravirtualized driver framework that improves guest VM disk and network I/O performance compared to fully emulated hardware. QEMU is the emulator itself, KVM is the kernel-based hypervisor providing hardware-accelerated virtualization, libvirt is a management API/toolkit, and virt-manager is a graphical management interface rather than the paravirtualized driver framework itself.

xz achieves a higher compression ratio than gzip at the cost of speed, and is commonly used for distributing Linux source tarballs where smaller file size matters more than compression speed.

shred is designed to securely overwrite a file's contents multiple times before deletion, making recovery through normal forensic tools impractical, exactly fitting a secure data destruction procedure. rm removes the directory entry but typically leaves underlying data recoverable, mv relocates a file without touching content, unlink removes a directory entry similarly to rm, and stat displays metadata rather than destroying data.

/etc/nsswitch.conf controls which sources, such as local files or DNS, are consulted and in what order for various lookups including hostname resolution. /etc/hosts provides static mappings, /etc/resolv.conf lists DNS servers to query, /etc/netplan configures interfaces on some distributions, and /etc/fstab defines filesystem mounts rather than lookup source ordering.

du summarizes disk space usage per directory, ideal for identifying which subdirectory consumes the most space. df reports overall filesystem-level space usage rather than per-directory breakdown, fio benchmarks storage I/O performance, stat displays metadata for a single file, and lsblk lists block devices rather than directory-level space consumption.

rmmod removes a currently loaded kernel module from the running kernel. insmod inserts a module, modprobe intelligently loads a module along with its dependencies, depmod generates module dependency information, and modinfo displays information about a module without removing it.