LPIC-2 201-450 study plan
LPIC-2 201-450 · Give each week a clear focus.
8 weeks · About 7 hours per week
Networking Configuration
Configure persistent IPv4/IPv6 addressing, static routes, and a second IP on one interface using ip plus the distribution's native tooling (NetworkManager keyfiles or systemd-networkd .network files).
Create a VLAN-tagged subinterface and an active-backup bond, then prove failover by taking one slave down and reading the bond status.
Attach a namespace or VM to a software bridge and confirm forwarding with ping and bridge link.
Troubleshoot a broken path using ip route get, ss -tulpn, tcpdump on the right interface, and nftables/iptables counters.
Configure name resolution (systemd-resolved or resolv.conf) plus static /etc/hosts entries, and diagnose a deliberate failure with dig and host.
Your checkpoint
A committed lab repo containing the VLAN+bond+bridge config files and pasted output of ip -d link, bridge link, and a timestamped failover log showing traffic surviving slave removal.
System Maintenance
Compile a small program from source (./configure && make) and install it under /usr/local, listing every build dependency you had to add.
Register a custom library directory in /etc/ld.so.conf.d/, run ldconfig, and verify resolution with ldd and ldconfig -p.
Write an rsync backup script using --link-dest, schedule it with a systemd timer, then delete a directory and restore it from the backup.
Route one syslog facility to a dedicated file in rsyslog and make journald persistent (Storage=persistent), proving both survive reboot.
Audit package integrity with rpm -Va or debsums and explain three findings from the output.
Your checkpoint
A restore-verified backup package: diff -r output showing a clean restore, the timer listed in systemctl list-timers, and the restored file's checksum matching the original.
Advanced Storage Device Administration
Build a RAID 1 or RAID 5 array with mdadm, write a marker file, mark one member faulty, remove and re-add it, and watch the resync in /proc/mdstat.
Create a PV/VG/LV stack, format and mount it, then grow the LV and the filesystem online with lvextend -r (or resize2fs / xfs_growfs).
Take an LVM snapshot, corrupt a test file on the origin, roll back from the snapshot, and compare checksums.
Persist the array and volume group across reboot via mdadm.conf, initramfs regeneration, and /etc/fstab UUID entries.
Inspect a disk with hdparm -I and sdparm and justify one tuning decision (readahead, I/O scheduler, or write cache).
Your checkpoint
A RAID+LVM lab report with /proc/mdstat captured in three states (degraded, recovering, clean) and a snapshot rollback proven by matching sha256sum output.
Filesystem and Devices
Create ext4 and XFS filesystems, label and tune them (tune2fs -L/-o, xfs_admin), and mount both from /etc/fstab by UUID.
Break a test filesystem on purpose, then repair it from rescue with fsck and xfs_repair without losing the marker files.
Implement on-demand mounting twice: a systemd .mount/.automount pair and an autofs map.
Enable user quotas on ext4 (or project quotas on XFS) and demonstrate a quota being hit.
Create an encrypted volume with cryptsetup and unlock it at boot through /etc/crypttab with a keyfile.
Your checkpoint
A VM that reboots cleanly with ext4 root, a separate XFS /data under quotas, an automounted share, and an encrypted /srv, captured with blkid and findmnt output.
Linux Kernel
Build a kernel from source using localmodconfig or menuconfig, install modules and an initramfs, and boot it in a VM.
Manage modules with modprobe, depmod, and insmod, blacklist one driver in /etc/modprobe.d, and confirm with lsmod.
Load a module with a non-default parameter, persist it in a modprobe.d options file, and verify after reboot.
Tune kernel runtime parameters with sysctl, persist them in /etc/sysctl.d, and prove the value survives a restart.
Write a udev rule that creates a stable symlink for a device, reload with udevadm control --reload and udevadm trigger, and confirm the link.
Your checkpoint
dmesg output from the self-built kernel showing the new version string, plus a persisted udev symlink and sysctl value demonstrated after reboot.
System Startup
Reinstall and regenerate GRUB2 (grub-install, grub-mkconfig), then recover a deliberately destroyed bootloader from rescue media.
Write a custom systemd service unit plus a target that groups it with a second unit, enable it, and check ordering with systemd-analyze.
Change the default boot target and mask one unit to prove it refuses to start.
Regenerate the initramfs with dracut, mkinitcpio, or update-initramfs and confirm a required module is inside the new image.
Profile boot time with systemd-analyze blame and critical-chain, then disable or delay one slow unit.
Your checkpoint
A one-page boot-recovery runbook validated on a VM whose GRUB was wiped, with exact commands and timestamps for each recovery step.
Capacity Planning
Enable sysstat logging and collect a full week (or a compressed multi-day sample) read back with sar -u, -r, -d, and -n DEV.
Establish baselines for CPU, memory, disk I/O, and network, and rank the host's top three resource consumers.
Reproduce a bottleneck with fio or a memory hog, then prove the diagnosis with iostat, vmstat, and pidstat evidence.
Forecast 6-12 month growth from the collected data and write a sized upgrade recommendation.
Automate a threshold check with a small systemd timer script that parses sar output and logs an alert.
Your checkpoint
A capacity report containing multi-day tables or graphs, a bottleneck verdict naming the constrained resource and its evidence, and a quantified upgrade recommendation.
Networking Configuration & System Maintenance
Re-run every earlier deliverable checklist and close all open items in the review log.
Drill your two lowest-scoring domains in 30-minute timed sets of at least 20 questions each until you clear 85%.
Recall from memory, on a blank VM, the 40 most-used commands across mdadm, LVM, cryptsetup, grub, modprobe, and network setup.
Sit two full-length timed mock exams and log every uncertain answer with the reference that resolves it.
Your checkpoint
A final scorecard showing two full-length mocks at 80% or better overall and 70% or better in every domain, plus a one-page cheat sheet of the commands you personally got wrong.
Milestones
Week 2: Networking and maintenance labs pass a cold-boot test: bond failover, VLAN reachability, backup timer, and the rsyslog rule all still work after a restart, and all configs are committed.
Week 4: A 40-question timed mock covering domains 203-205 scores at least 65%, and every miss is converted into a flashcard or a lab task in the review log.
Week 6: All seven domains have been studied at least once; the self-built kernel boots in a VM and the GRUB recovery runbook has been executed successfully end to end.
Week 8: Two full-length timed mocks score 80% or better overall and 70% or better in every individual domain, with no unclosed items in the review log.