Storage management organises, protects, and makes data accessible in a computer system. For the CompTIA Server+ SK0-005 exam, understanding RAID, SAN, and NAS is crucial because servers rely on these technologies to keep data safe and available, and the exam will test your ability to pick the right solution for a given scenario.
Jump to a section
A simple way to picture Storage Management (RAID, SAN, NAS)
Because a single bookshelf in a library cannot hold all the books and retrieving a book takes too long, the library installs a set of shelves arranged in a special pattern (RAID) to protect against damage, plus a dedicated conveyor belt system (SAN) that moves books between shelves and reading rooms at very high speed, and a separate reading room (NAS) where multiple librarians can access the same reference books over the network.
RAID is like grouping several cheap bookshelves together. Some shelves mirror each other (RAID 1), so if one loses a book, the copy survives. Others split a thick book across shelves (RAID 0) for faster access, but if a shelf breaks, the whole book is lost. A smart arrangement (RAID 5) adds a summary shelf with parity information, allowing any one lost shelf to be rebuilt. The conveyor belt (SAN) is a private, high-speed system that only storage specialists use, separate from the normal library traffic, moving entire wooden crates (blocks of data) between shelves. The reading room (NAS) is a shared bookshelf accessible via the library's network, easy for anyone to use but slower than the conveyor belt because it must serve many readers at once. This analogy shows how servers use RAID, SAN, and NAS to keep data safe, fast, and accessible.
Storage management covers how servers store and retrieve data reliably and efficiently. Three key technologies are RAID, SAN, and NAS. Each solves a different problem.
RAID (Redundant Array of Independent Disks) combines multiple physical hard drives into a single logical unit to improve performance or provide fault tolerance (the ability to survive a drive failure). A RAID controller (hardware or software) manages this. Different RAID levels offer different trade-offs:
RAID 0 (striping): splits data across all drives. Fast but no fault tolerance. If one drive fails, all data is lost. Requires at least 2 drives.
RAID 1 (mirroring): duplicates data exactly on two drives. Excellent fault tolerance — one drive can fail without data loss. Slower writes but fast reads. Needs exactly 2 drives (or more in pairs).
RAID 5 (striping with parity): stripes data and stores parity information across all drives. Can survive one drive failure. Needs at least 3 drives. Write speed is slower due to parity calculations.
RAID 6 (striping with dual parity): similar to RAID 5 but can survive two simultaneous drive failures. Needs at least 4 drives.
RAID 10 (striping + mirroring): combines RAID 0 and RAID 1. Drives are mirrored in pairs, then striped. Offers high performance and fault tolerance (up to half the drives can fail, depending on which ones). Needs at least 4 drives.
SAN (Storage Area Network) provides block-level storage over a dedicated high-speed network. Think of it as a private highway for storage traffic. A server connects to the SAN using protocols like Fibre Channel (FC) or iSCSI. The SAN presents logical units (LUNs) to the server, which sees them as local hard drives. SAN is ideal for databases and virtualisation because it offers low latency and high throughput. It requires specialised switches and storage arrays.
NAS (Network Attached Storage) provides file-level storage over a standard network (usually Ethernet). A NAS device is a dedicated file server that uses protocols like NFS (Network File System) or SMB/CIFS (Common Internet File System) to share files. Servers and clients access the NAS over the network like any network drive. NAS is simpler to set up than SAN and is great for file sharing, backups, and media libraries. However, it is not as fast as SAN because it has to handle file-system overhead.
Key differences between SAN and NAS:
SAN is block-level; NAS is file-level.
SAN requires a dedicated network; NAS works on existing Ethernet.
SAN is typically faster but more expensive; NAS is slower but cheaper and easier to manage.
How they work together: A data centre might use a SAN for mission-critical databases (block storage, fast), a NAS for shared file storage (file storage, convenient), and RAID inside both the SAN array and the NAS device to protect against disk failures. The server sees the SAN LUN as a local disk, formats it, and installs an application. The NAS is accessed via a network share. RAID ensures that if a physical disk in the storage array fails, the data remains intact.
Why they exist: Before RAID, a single disk failure meant total data loss. Before SAN and NAS, each server had its own directly attached disks (DAS), which wasted capacity and made data sharing difficult. These technologies allow centralised, resilient, and scalable storage.
Assess Storage Requirements
Determine the type of data (database, files, virtual machines), performance needs (IOPS, throughput), fault tolerance requirements, and budget. This decides whether to use RAID, SAN, or NAS.
Select the Right RAID Level
Choose a RAID level based on the balance of performance, capacity, and fault tolerance. For example, RAID 10 for databases, RAID 5 for general file storage, RAID 6 for large archival arrays.
Install Physical Disks and Controller
Physically install the required number of drives into the server or storage array. Connect them to a RAID controller (hardware or software). Ensure cabling and power are properly managed.
Configure the RAID Array
Access the RAID controller’s utility (BIOS setup or web interface). Select the disks, choose the RAID level, and create the logical volume. This defines how data is striped, mirrored, or parity-checked.
Create a Logical Unit (LUN) or Share
For SAN, define a LUN on the storage array and map it to the server via iSCSI or Fibre Channel. For NAS, set up a shared folder with appropriate permissions. The server then sees the storage as a local drive or network share.
Format and Mount the Volume
On the server, initialise the new disk, create a file system (e.g., ext4, NTFS), and mount it to a directory. Make the mount persistent across reboots. Verify that the storage is accessible and performance matches expectations.
An IT professional at a medium-sized company is asked to upgrade the storage infrastructure supporting a customer database, a file share for office documents, and a test/dev environment. Here’s how they tackle it.
First, they assess the requirements. The database needs fast, low-latency access and must survive a drive failure. The file share needs to be accessible from many desktops and laptops without special software. The test/dev environment needs flexibility to clone and snapshot volumes quickly.
For the database, they select a SAN using Fibre Channel. They configure a RAID 10 array inside the SAN storage enclosure to balance performance and fault tolerance. They create a LUN of 500 GB and map it to the database server. The server sees the LUN as a local disk, formats it with a file system (ext4 or NTFS), and installs the database software. The SAN ensures dedicated bandwidth so database queries run fast even during backup peaks.
For the file share, they deploy a NAS appliance connected to the office local area network (LAN). They create a shared folder using SMB protocol, set up permissions (using ACLs), and map the share to all employee workstations. The NAS uses RAID 6 internally (six drives) so it can tolerate two disk failures without downtime. The IT pro schedules regular snapshots for quick recovery of accidentally deleted files.
For test/dev, they configure a separate LUN on the SAN with RAID 5 and use the SAN’s snapshot feature to create point-in-time copies. Developers can quickly revert or clone snapshots without affecting production.
Step by step, the IT pro:
Evaluates performance needs (IOPS, throughput).
Decides between SAN and NAS based on protocol and cost.
Orders the hardware (SAN array, Fibre Channel switches, NAS unit, drives).
Physically installs components in racks, cables, and powers on.
Configures RAID levels in the storage array’s management software.
Creates LUNs or shares and maps them to servers.
Tests connectivity, performance, and fault tolerance (e.g., pulling a drive).
Documents the configuration for disaster recovery.
This scenario shows why IT pros need to know the differences: choosing the wrong storage type can cripple performance or cause data loss. The SK0-005 exam expects you to make similar decisions.
The SK0-005 exam tests your ability to implement and differentiate RAID, SAN, and NAS. Expect questions that ask you to select the correct RAID level for a given requirement (e.g., 'Which RAID level offers striping with parity and requires at least three disks?'), or to compare SAN and NAS in terms of performance, cost, and accessibility.
Common question types: - Scenario-based: 'A company needs shared file storage for 50 users. Should they use SAN or NAS?' Answer: NAS (file-level). - RAID selection: 'Which RAID level provides fault tolerance without using dedicated parity disks?' Answer: RAID 1 (mirroring) or RAID 10. - Protocol identification: 'Which protocol is used for block-level storage over IP?' Answer: iSCSI. - Fault tolerance: 'How many disk failures can RAID 5 survive?' Answer: One. - Hardware vs software: 'What is an advantage of hardware RAID over software RAID?' Answer: Less CPU overhead.
Traps they set: - They might ask about 'RAID 0' and 'fault tolerance' in the same question. The trap is that RAID 0 has no fault tolerance — always remember this. - They might describe a scenario where the requirement is 'low cost' and 'easy setup' and the answer is NAS, but a beginner might pick SAN because it sounds more 'enterprise'. - They might test whether you know that iSCSI can run over standard Ethernet, not just Fibre Channel. - They might ask about minimum number of disks for RAID levels. Memorise: RAID 0 = 2, RAID 1 = 2, RAID 5 = 3, RAID 6 = 4, RAID 10 = 4. - They might confuse 'block-level' and 'file-level' — know that SAN = block, NAS = file.
Key concepts to memorise: - RAID levels and their properties (performance, fault tolerance, capacity efficiency). - SAN vs NAS differences. - Protocols: iSCSI, Fibre Channel, NFS, SMB/CIFS. - LUN (Logical Unit Number): a block-level storage volume. - JBOD (Just a Bunch Of Disks): not RAID, no redundancy.
Practice questions often have a 'best fit' style. Read the scenario carefully: if it says 'multiple servers need to access the same files concurrently', that points to NAS. If it says 'high-performance database', point to SAN. The exam also tests your ability to calculate usable capacity given parity overhead – but that’s rare for SK0-005.
RAID 0 offers high performance but zero fault tolerance; never use it for critical data.
RAID 1 uses mirroring and can survive one disk failure, but capacity is halved.
RAID 5 uses distributed parity and can survive one disk failure; it requires at least three disks.
RAID 6 can survive two disk failures and requires at least four disks, offering higher resilience at the cost of write performance.
SAN provides block-level storage over a dedicated network, ideal for databases and virtualisation.
NAS provides file-level storage over a standard network, ideal for file sharing and backups.
iSCSI is a protocol that allows block-level storage over IP networks, making SAN more affordable.
These come up on the exam all the time. Here's how to tell them apart.
RAID 5
Can survive one disk failure.
Uses single parity.
Minimum 3 disks required.
RAID 6
Can survive two disk failures.
Uses dual parity.
Minimum 4 disks required.
SAN
Provides block-level storage.
Requires dedicated network (or iSCSI can share).
Higher performance, lower latency.
NAS
Provides file-level storage.
Works on existing Ethernet network.
Lower performance, easier setup.
Fibre Channel SAN
Requires specialised FC switches and HBAs.
Higher performance and reliability.
More expensive.
iSCSI SAN
Uses standard Ethernet and NICs.
Lower cost, easier to manage.
Can share network with other traffic.
Hardware RAID
Uses dedicated RAID controller card.
Offloads RAID processing from CPU.
Often has battery backup and cache.
Software RAID
Uses OS kernel modules.
Consumes CPU cycles for RAID operations.
No extra hardware cost, but may be slower.
Mistake
RAID 0 provides fault tolerance because it uses multiple disks.
Correct
RAID 0 provides no fault tolerance; it actually increases risk because any single disk failure destroys all data.
Beginners assume that using multiple disks must add safety, but striping spreads data without duplication.
Mistake
NAS and SAN are essentially the same thing – both are network storage.
Correct
NAS provides file-level access over a standard network, while SAN provides block-level access over a dedicated high-speed network.
Both are network-based and often used together, so their differences in protocol and performance are easily overlooked.
Mistake
RAID 5 can survive two disk failures because it has parity.
Correct
RAID 5 can survive only one disk failure. RAID 6 is needed for two simultaneous failures.
The term 'parity' sounds like it covers everything, but RAID 5 uses only one set of parity data.
Mistake
Using more disks in a RAID array always improves performance.
Correct
More disks can improve read speeds (especially in RAID 0 or 10), but write performance may degrade in parity-based RAIDs (5 or 6) due to computation overhead.
People think 'more is better' without considering the type of RAID and the overhead of parity calculations.
Mistake
A SAN must use Fibre Channel cables; it cannot run over ordinary Ethernet.
Correct
SAN can use either Fibre Channel or iSCSI. iSCSI transports SCSI commands over IP networks, including standard Ethernet.
The term 'storage area network' often evokes images of dedicated Fibre Channel, but iSCSI is a common and cheaper alternative.
Reveal each answer, then mark whether you got it right. Score 60%+ to unlock the next chapter.
RAID 5 can survive one disk failure and uses single parity; RAID 6 can survive two disk failures and uses dual parity. RAID 6 requires at least four disks, while RAID 5 requires at least three.
Technically yes, but the array will be limited to the capacity of the smallest disk. It is recommended to use identical disks for optimal performance and capacity.
A LUN (Logical Unit Number) is a logical volume presented by a SAN storage array to a server. The server sees it as a local hard drive and can format and use it.
Not necessarily. Software RAID (e.g., built into the OS) uses the CPU to handle RAID operations. Hardware RAID uses a dedicated controller card, which offloads the work and often provides better performance and features.
iSCSI (Internet Small Computer System Interface) is a protocol that allows sending SCSI commands over IP networks. It enables a server to connect to a SAN using standard Ethernet instead of dedicated Fibre Channel.
Block-level storage treats storage as raw blocks (like a hard drive) and requires the server to manage the file system. File-level storage already has a file system and shares files over the network, so the client sees a shared folder.
You've finished Storage Management (RAID, SAN, NAS). Continue through the SK0-005 study guide to build a complete picture of the exam.
Done with this chapter?