LPIC-2 Block Devices, Filesystems and Advanced Storage Practice Test 5 — 15 Questions

Question 1

A database server uses LVM with an XFS filesystem on /dev/vg_data/lv_db, mounted at /var/lib/mysql. The LV is at 95% capacity. A new 50GB disk /dev/sdd has been added and initialized as a physical volume, then added to the volume group vg_data. The database is running and cannot be stopped. The administrator must extend the LV and filesystem. What is the correct course of action?

Choose an answer to begin — your selection is scored in the full session.

Accepted Answer

Run `lvextend -L +50G /dev/vg_data/lv_db` followed by `xfs_growfs /var/lib/mysql`.. XFS filesystems must be grown while mounted using `xfs_growfs` with the mount point as the argument, not the block device. Since the database cannot be stopped, the LV can be extended online with `lvextend`, and then `xfs_growfs /var/lib/mysql` resizes the filesystem to fill the expanded LV without requiring unmounting or downtime.

Answer

Run `lvextend -L +50G /dev/vg_data/lv_db` followed by `resize2fs /dev/vg_data/lv_db`.

Answer

Unmount /var/lib/mysql, run `lvextend -L +50G /dev/vg_data/lv_db`, run `resize2fs /dev/vg_data/lv_db`, then remount.

Answer

Run `lvresize -L +50G /dev/vg_data/lv_db`.

Question 2

A system administrator manages a CentOS 7 server with a single XFS filesystem on /dev/vg_data/lv_app mounted at /srv/app. The filesystem is nearly full and the volume group vg_data has 40 GB of free space available. The administrator needs to make the filesystem 20 GB larger without unmounting it and without losing existing data. Which sequence of commands should the administrator run?

Accepted Answer

lvextend -L +20G /dev/vg_data/lv_app && xfs_growfs /srv/app. XFS filesystems are grown online with xfs_growfs after the underlying logical volume has been enlarged. lvextend adds the requested extents to the LV, then xfs_growfs expands the filesystem to use the new space while it remains mounted. No unmount, mkfs, or data migration is needed.

Answer

lvextend -L +20G /dev/vg_data/lv_app && resize2fs /dev/vg_data/lv_app

Answer

xfs_growfs -L +20G /srv/app && lvextend /dev/vg_data/lv_app

Answer

lvresize -L +20G /dev/vg_data/lv_app && xfs_repair /srv/app

Question 3

A storage administrator has an existing software RAID5 array /dev/md0 built from three 2 TB SATA disks on a Linux server. The array is healthy but has no spare device. The administrator adds a fourth 2 TB disk, /dev/sdd, and wants to add it to the array so that it becomes an active member, increasing usable capacity while preserving existing data. Which command should the administrator run first?

Accepted Answer

mdadm --add /dev/md0 /dev/sdd. To expand a RAID5 array with an additional disk, the new device must first be added as a spare using mdadm --add. After the spare is synchronized, mdadm --grow with --raid-devices set to the new count reshapes the array and activates the spare, increasing usable capacity without recreating the array.

Answer

mdadm --grow /dev/md0 --raid-devices=4 --add /dev/sdd

Answer

mdadm --create /dev/md0 --level=5 --raid-devices=4 /dev/sda /dev/sdb /dev/sdc /dev/sdd

Answer

mdadm --assemble /dev/md0 /dev/sda /dev/sdb /dev/sdc /dev/sdd

Question 4

A backup server stores nightly snapshots on an LVM logical volume formatted with ext4. The administrator wants to take a point-in-time copy of the volume while it remains mounted and actively written to, then mount the copy read-only on another host for verification. Which command should the administrator run to create the snapshot?

Accepted Answer

lvcreate --size 20G --snapshot --name lv_backup /dev/vg_data/lv_db. Creating a snapshot of a live, mounted ext4 volume is done with lvcreate using the --snapshot option, specifying the origin LV, a snapshot name, and a COW size. The snapshot preserves the point-in-time image while writes continue to the origin, allowing the copy to be activated and mounted read-only for verification without disrupting the backup server.

Answer

vgsnapshot --name lv_backup --origin /dev/vg_data/lv_db --size 20G

Answer

dd if=/dev/vg_data/lv_db of=/dev/vg_data/lv_backup bs=4M conv=noerror

Answer

lvconvert --snapshot /dev/vg_data/lv_db --name lv_backup --size 20G

Question 5

A junior administrator adds a new 4 TB SATA disk as /dev/sdc and must prepare a single GPT-partitioned layout with one partition spanning the whole disk for a future filesystem. Which command should be used to create the GPT partition table and the partition?

Accepted Answer

gdisk /dev/sdc. GPT partitioning on a large disk is performed with gdisk, which writes GPT structures and supports partitions above the 2 TB MBR limit. The administrator can create one partition spanning the disk with the interactive new-partition command and then write the table. partprobe is only used afterward to make the kernel aware of the new layout.

Answer

partprobe /dev/sdc

Answer

mkfs.gpt /dev/sdc

Answer

fdisk /dev/sdc

Question 6

A server has a software RAID5 array /dev/md0 built from four SATA disks. After a power loss, the array assembles but shows one member as faulty. The administrator replaces the failed disk with an identical model and partitions it to match the others. Which command should be used to add the new partition to the array and begin reconstruction?

Accepted Answer

mdadm --add /dev/md0 /dev/sde1. Replacing a failed member of an active RAID array requires hot-adding the new device as a spare. The mdadm --add command performs this and triggers a rebuild on a degraded RAID5 array. The grow option reshapes geometry, re-add expects prior membership, and assemble activates a stopped array, so none of them fit the replacement scenario.

Answer

mdadm --assemble /dev/md0 /dev/sde1

Answer

mdadm --re-add /dev/md0 /dev/sde1

Answer

mdadm --grow /dev/md0 --raid-devices=4

Question 7

A Linux server has an LVM volume group named vg_data containing a 200 GB logical volume lv_archive formatted with XFS and mounted at /srv/archive. The volume group has 300 GB of unallocated space on a single physical volume. The administrator must expand /srv/archive by 100 GB while the filesystem remains mounted and accessible. Which command sequence accomplishes this?

Accepted Answer

lvextend -L +100G /dev/vg_data/lv_archive, then xfs_growfs /srv/archive. Growing an XFS filesystem on LVM requires two distinct operations: first extend the logical volume with lvextend so more extents back it, then grow the filesystem into that space with xfs_growfs using the mount point. XFS supports online growth, so neither unmounting nor a repair step is needed. Because vg_data already has unallocated extents, no vgextend or pvresize is required beforehand.

Answer

vgextend vg_data /dev/sdb1, then xfs_growfs /srv/archive

Answer

lvextend -L 100G /dev/vg_data/lv_archive, then xfs_growfs /srv/archive

Answer

lvextend -L +100G /dev/vg_data/lv_archive, then resize2fs /srv/archive

Question 8

A Linux server has a 500 GB disk /dev/sdb that will be used exclusively as a single XFS filesystem mounted at /data. The administrator wants to create the filesystem directly on the whole disk without a partition table. Which command should be used?

Accepted Answer

mkfs.xfs /dev/sdb. Creating an XFS filesystem directly on a whole disk is done with mkfs.xfs followed by the block device path. Since no partition table exists, /dev/sdb1 is invalid, and xfs_growfs only expands existing filesystems. Adding a label with -L is optional and does not change where the filesystem is created.

Answer

xfs_growfs /dev/sdb

Answer

mkfs.xfs /dev/sdb1

Answer

mkfs -t xfs -L /data /dev/sdb

Question 9

A technician is preparing a new 4 TB external USB disk on a Linux server. The disk currently has an old MBR partition table and must be repartitioned to use a single partition spanning the entire disk. The technician wants to use parted in non-interactive mode. Which command should the technician run to create a GPT label and a single partition covering the whole disk on /dev/sdc?

Accepted Answer

parted -s /dev/sdc mklabel gpt mkpart primary ext4 0% 100%. The non-interactive parted syntax uses -s followed by subcommands in order. mklabel gpt replaces the MBR table with GPT, which is required to address the full 4 TB. mkpart primary ext4 0% 100% then creates one partition spanning from the beginning to the end of the disk using percentages.

Answer

parted -s /dev/sdc mklabel msdos mkpart primary ext4 1MiB 4TB

Answer

parted -s /dev/sdc unit TB mklabel gpt mkpart primary ext4 0TB 4TB

Answer

parted /dev/sdc print | grep -i gpt && parted /dev/sdc mkpart primary ext4 0 100

Question 10

An administrator manages an LVM volume group vg_app with several logical volumes. A monitoring alert shows the volume group is out of free extents, and a new logical volume must be created for a staging database. Which command should the administrator run first to determine how much space can be reclaimed or added?

Accepted Answer

vgs vg_app. The volume group summary command vgs shows total and free space for a named group in one line, making it the fastest way to confirm whether extents are available. lvs and pvs provide logical-volume and physical-volume detail respectively, and pvdisplay requires a device path rather than a group name, so none of them answer the capacity question as directly.

Answer

pvdisplay vg_app

Answer

pvs vg_app

Answer

lvs vg_app

Question 11

A Linux server has an ext4 filesystem on /dev/mapper/vg_data-lv_home mounted at /home. The administrator notices that the filesystem is full and that 5% of blocks are reserved for root, which is wasteful on this large data volume. The administrator wants to reduce the reserved block percentage to 1% without unmounting the filesystem. Which command should the administrator run?

Accepted Answer

tune2fs -m 1 /dev/mapper/vg_data-lv_home. The tune2fs utility adjusts ext4 filesystem parameters on a mounted filesystem. The -m option sets the percentage of blocks reserved for the root user. Changing it from 5% to 1% frees additional space for regular users while retaining a small reserve, and the change takes effect without unmounting.

Answer

resize2fs -m 1 /dev/mapper/vg_data-lv_home

Answer

e2fsck -m 1 /dev/mapper/vg_data-lv_home

Answer

tune2fs -r 1 /dev/mapper/vg_data-lv_home

Question 12

A technician is preparing a new 500 GB SATA disk at /dev/sdc. The disk must contain a single GPT partition table with one partition spanning the entire disk, and the partition type must be set for Linux filesystems. Which command creates the GPT label on the disk?

Accepted Answer

parted /dev/sdc mklabel gpt. Creating a GPT partition table is a distinct step performed before any partitions exist. The parted mklabel gpt command writes the GUID Partition Table to the disk non-interactively, which is why it fits the requirement. Commands such as mkpart, interactive fdisk sequences, or nonexistent mkfs helpers either operate on an existing label or address the wrong layer entirely.

Answer

parted /dev/sdc mkpart primary 0% 100%

Answer

mkfs.gpt /dev/sdc

Answer

fdisk /dev/sdc then press g to create a new GPT disklabel

Question 13

A server has two paths to a SAN LUN, /dev/sdb and /dev/sdc, and multipath -ll reports a single mpath device with one active and one standby path. The administrator wants I/O to use both paths simultaneously for higher throughput. Which action is required?

Accepted Answer

Set path_grouping_policy to multibus in the multipath configuration and reload the maps.. Aggregate throughput across multiple paths requires the paths to share one priority group, which the multibus grouping policy accomplishes. With all paths in a single group, the path selector can distribute I/O across them. Standby grouping, by contrast, deliberately keeps one group active and the rest passive for failover only.

Answer

Increase the queue_if_no_path timeout so both paths remain queued and usable.

Answer

Enable the retain_attached_hw_handler option and restart the multipathd service.

Answer

Change the path_selector to round-robin and run multipath -r without altering grouping.

Question 14

A backup server has /dev/md0 as a RAID1 array assembled from /dev/sdd1 and /dev/sde1. During routine maintenance the administrator must remove /dev/sde1 from the array so the disk can be replaced, while keeping /dev/md0 mounted and serving data. Which command marks /dev/sde1 as failed and removes it from the active array?

Accepted Answer

mdadm /dev/md0 --fail /dev/sde1 --remove /dev/sde1. Removing a member from a live RAID1 array is done in two logical steps: mark the device faulty so the kernel stops issuing I/O to it, then detach it from the array. mdadm accepts both --fail and --remove in one invocation, and because a healthy mirror remains, /dev/md0 keeps operating in degraded mode with no unmount required. Zeroing superblocks or stopping the array are either premature or disruptive.

Answer

mdadm --stop /dev/md0 then mdadm --assemble /dev/md0 /dev/sdd1

Answer

mdadm --zero-superblock /dev/sde1 then mdadm /dev/md0 --remove /dev/sde1

Answer

mdadm --manage /dev/md0 --add /dev/sde1 --re-add /dev/sde1

Question 15

A technician is asked to create a software RAID1 array using two new disks, /dev/sdb and /dev/sdc, and make it persistent across reboots. The array must be usable immediately. Which command creates the array with persistent metadata?

Accepted Answer

mdadm --create /dev/md0 --level=1 --raid-devices=2 /dev/sdb /dev/sdc. Creating a new software RAID array requires mdadm in --create mode with the correct level and member devices. The --create action writes metadata superblocks to each member, enabling automatic assembly at boot. Assemble, manage, and the wrong RAID level do not produce a new persistent RAID1 array from blank disks.

Answer

mdadm --assemble /dev/md0 /dev/sdb /dev/sdc

Answer

mdadm --create /dev/md0 --level=0 --raid-devices=2 /dev/sdb /dev/sdc

Answer

mdadm --manage /dev/md0 --add /dev/sdb /dev/sdc