LPIC-3 High Availability and Storage Clusters
More study resources
Cheat sheet
Keep the key ideas within reach.
ExploreKey terms
Get comfortable with the vocabulary.
ExploreStudy plan
Give each week a clear focus.
ExploreCommon mistakes
Understand the traps and how to avoid them.
ExploreCase studies
Put your knowledge to work in context.
ExploreExam-day guide
Prepare for the day ahead.
ExploreAbout this exam
LPIC-3 High Availability and Storage Clusters is one of four separate LPIC-3 specialty certifications offered by the Linux Professional Institute, the vendor-neutral certification body behind the multi-level LPIC program. It targets enterprise-level Linux professionals who design, run and recover highly available systems and clustered storage. The current version is 3.0, exam code 306-300; it results from a split of version 2.0 of exam 304, and the previous 304-200 version was available only until June 20th, 2022. To receive the certification you must hold an active LPIC-2 certification and pass the 306 exam. The exam lasts 90 minutes and contains 60 multiple-choice and fill-in-the-blank questions; the certification is valid for 5 years. Content spans four topics: High Availability Cluster Management (concepts, quorum and fencing, LVS/IPVS with keepalived and ldirectord, HAProxy, Pacemaker 2.x with Corosync), High Availability Cluster Storage (DRBD 9.0.x including LINSTOR awareness, iSCSI, DM-MPIO, clustered LVM with DLM, GFS2 and OCFS2), High Availability Distributed Storage (GlusterFS, Ceph with RBD, CephFS, pools and CRUSH maps), and Single Node High Availability (SMART monitoring, UPS shutdown, monit, advanced RAID, LVM thin pools and RAID, bonding, VLANs and BGP awareness). Prepare hands-on: build a two-node Pacemaker cluster, replicate with DRBD, deploy Ceph and GlusterFS, and practise pcs, crmsh, mdadm and lvm commands. LPI publishes per-objective weights on the objectives page, so allocate study time by weight rather than by topic order. The exam is offered in English and Japanese at VUE test centers and online via OnVUE.
What you'll study
About LPIC-3 High Availability and Storage Clusters preparation
What is the format and length of exam 306-300?
LPI states the exam is 90 minutes long and consists of 60 multiple-choice and fill in the blank questions. Because fill-in-the-blank items require typing exact commands, configuration file names or parameter values, memorising spelling of tools such as pcs, corosync-cfgtool, drbdadm, ceph-volume and mdadm matters as much as understanding the concepts.
Do I need LPIC-2 before taking 306?
Yes. LPI lists an active LPIC-2 certification as the prerequisite, and states that you must be LPIC-2 certified and pass the 306 exam to receive the LPIC-3 High Availability and Storage Clusters certification. You can sit the exam first, but the specialty credential is only granted once the LPIC-2 requirement is satisfied.
Which exam version should I study for?
Study for version 3.0, exam code 306-300. LPI notes that these objectives result from a split of version 2.0 of exam 304, and that the previous version 2.0 (code 304-200) was available only until June 20th, 2022, so it is no longer bookable. Always check the objectives page before scheduling in case a newer version appears.
How long is the certification valid, and what languages are offered?
LPI lists a validity period of 5 years and mentions options for extending beyond that period. The exam is available in English and Japanese both at VUE test centers and online via OnVUE. If you need another language, check LPI's site directly rather than assuming availability from the overview page.
What does the exam cost and what score do I need to pass?
LPI's overview page does not publish a fixed fee or passing score; for cost it links to per-country exam pricing, and no cut score is stated there. Check the LPI site and your test provider account for the current price in your country and any scoring details before you book a seat.
What should I study first?
Follow the published objective weights rather than topic order. Within topic 361, Load Balanced Clusters and Failover Clusters each carry weight 8; Ceph (363.2) also carries 8 and DRBD (362.1) and HA Concepts (361.1) carry 6. Network High Availability (364.4) and GlusterFS (363.1) carry 5. Build the Pacemaker/corrsync cluster early, since DRBD, GFS2 and clustered LVM all integrate with it.
How much hands-on practice is really needed?
Expect to need real practice, because the objectives are written in operational terms: install, configure, maintain, troubleshoot, recover after an outage and restore integrity. Practise fencing with sbd and stonith_admin, split-brain handling in DRBD, replacing failed GlusterFS bricks, and Ceph recovery after an OSD outage. Reading documentation alone rarely covers the recovery scenarios tested.
What are common failure modes on this exam?
Candidates typically lose points on exact syntax: pcs constraint versus crmsh crm syntax, DRBD protocol A/B/C semantics, dm-multipath stanza keywords, and ceph versus rados subcommands. Another frequent problem is weak split-brain and quorum reasoning, especially STONITH policy and Pacemaker resource stickiness. Drill command recall and verify cluster behaviour in a lab, not just concepts.
Is a fill-in-the-blank answer case-sensitive or order-sensitive?
LPI does not publish scoring mechanics for individual items, so treat fill-in answers as exact: use the spelling from the objectives' file, term and utility lists and the real tool's syntax. Learn full paths where they are listed, such as /etc/drbd.d/, /proc/mdstat, /etc/multipath.conf and corosync.conf, so answers match conventional usage.
How does this specialty relate to the other LPIC-3 certifications?
LPI offers four separate LPIC-3 specialty certifications, and passing any one of the four exams grants the LPIC-3 certification for that specialty only. Passing 306 does not confer the Security or Virtualization and Containerization specialties. Each has its own exam code and objectives, so plan which specialty matches your role before committing study time.