High Availability, RAID, Insurance

A spare tyre for everything

A car with a spare tyre keeps going after a puncture. A plane with two engines keeps flying if one fails. Neither helps if you drive off a cliff.

Fault tolerance lets a system keep running when a part fails. High availability (HA) designs whole services to minimise downtime, using spare parts and fast failover. Backups and insurance cover the cliff.

High availability techniques: clustering, load balancing, replication, redundant parts, power and geography.

Technique	How it works
Clustering	Several servers act as one. Active-active: all of them serve traffic. Active-passive: a standby takes over on failure.
Load balancing	Spreads traffic across servers and stops sending it to failed ones
Replication	Copies data to other systems or sites, synchronously or asynchronously
Redundant components	Dual power supplies, network cards, controllers, network paths
Redundant power	A UPS (batteries) bridges short interruptions and hands over smoothly; a generator covers long outages
Geographic redundancy	Services run in more than one site or cloud region

The weakest link sets availability. Two redundant servers plugged into one switch, or one power feed, still fail together.

RAID

RAID (redundant array of independent disks) combines several disks for speed, protection or both.

Level	How	Survives
RAID 0	Data striped across disks	Nothing. One failure loses everything, despite the name
RAID 1	Data mirrored on two disks	One disk in the pair
RAID 5	Striping with parity spread across all disks	One disk
RAID 6	Striping with two sets of parity	Two disks
RAID 10 (1+0)	Mirrored pairs, then striped	One disk per mirrored pair

RAID protects against a failed disk, and nothing else. It copies deletions, corruption and ransomware encryption instantly, and it’s no help if the whole array or site is lost. RAID is not a backup.

Example. A company relies on RAID 5 storage with nightly snapshots on the same array. Ransomware encrypts the volumes and deletes the snapshots. RAID faithfully protected the encrypted data. Only offline backups would have allowed recovery.

Insurance

Insurance moves some of the financial pain of a disaster to someone else.

Type	Covers
IT equipment and facilities	Physical damage to hardware and premises
Media (software) reconstruction	Rebuilding data and software media
Extra expense	Additional cost of keeping going after damage
Business interruption	Lost profit caused by the disruption (not replacement equipment)
Valuable papers and records	Cash value of records lost or damaged
Errors and omissions	Legal liability for professional mistakes that cause client losses
Fidelity coverage	Losses from employee dishonesty or fraud
Media transportation	Media lost or damaged in transit

Cyber insurance policies often add incident response costs, breach notification, extortion and liability.

What insurance can’t do

Insurance doesn’t bring service back, can’t repair a damaged reputation, and may not cover regulatory fines. It complements a continuity plan; it never replaces one. The auditor checks that cover matches the risks identified, and that the policy’s conditions (such as required controls) are actually being met.

How the exam asks about it

If the question asks…	The answer is usually…
RAID level with no redundancy	RAID 0
RAID level using mirroring	RAID 1
RAID level surviving two disk failures	RAID 6
Does RAID replace backups?	No; it doesn’t protect against deletion, corruption or site loss
Insurance covering lost profit during an outage	Business interruption
Insurance covering employee theft	Fidelity coverage
Device bridging short power outages	UPS

In short

High availability reduces downtime through redundancy: clustering, load balancing, replication and redundant components.
Active-active clusters share the load; active-passive clusters fail over to a standby.
RAID 0 stripes with no redundancy; RAID 1 mirrors; RAID 5 stripes with distributed parity and survives one disk failure; RAID 6 survives two; RAID 10 combines mirroring and striping.
RAID protects against disk failure, not against deletion, corruption, malware or site loss, so it doesn't replace backups.
Power resilience uses UPS for short outages and generators for longer ones, with tested failover.
Insurance transfers financial loss: IT equipment, media reconstruction, extra expense, business interruption, valuable records, errors and omissions, fidelity and media transportation.
Insurance doesn't restore operations; it complements, not replaces, continuity planning.

Exam traps

RAID 0 offers no fault tolerance, despite being RAID.
RAID and replication aren't backups: deleting a file deletes it on every mirror.
A single point of failure anywhere in the chain (power, network switch) defeats redundant servers.
Business interruption insurance covers lost profit, not the cost of replacing equipment.
Fidelity coverage protects against losses from employee dishonesty.

Slides

Click the slide and use the arrow keys. F toggles full screen.