ITPro.TV: CompTIA Security+ (SY0-601)
- author:: Nathan Acks
- date:: 2022-04-07
CompTIA Security+ Exam Cram
Today I’ll be reading Chapter 13 of the Security+ Exam Cram, “Cybersecurity Resilience”.
Redundancy
Power redundancy options:
- Dual power supplies (within a PC/server)
- UPSes
- Generators
- Managed power distribution units (PDUs) - basically fancy power strips
Load Balancers
Common algorithms:
- Round-robin
- Random
- Least (open) connections
All of these can also be weighted by various factors.
“Session affinity”, which ties a particular session to a particular server in the cluster, is important here for performance and complexity reasons.
NIC Teaming
The idea here is to “bond” multiple NICs into a single logical unit. This logical unit can then be used for redundancy and load balancing in the same way that a server cluster might be.
RAID
- RAID 0: Drives are logically joined in serial to create a single large drive.
- RAID 1: Drives are logically joined in parallel to create a single redundant drive.
- RAID 5: At least three drives are joined in parallel, with space allocated so that the data on any one drive can be reconstructed should it fail. This creates a logical drive larger than any single drive (though not as large as all three drives), but with similar redundancy characteristics as RAID 1.
- RAID 1+0 / 0+1: Both of these approaches require at least four disks, and are sometimes collectively knowns as “RAID 10”. This is either a RAID 1 array where the component disks are actually RAID 0 arrays (0+1), or a RAID 0 array where the component disks are RAID 1 arrays (1+0)
Backups
- Full
- Differential (difference since the last full backup)
- Incremental (difference since the last full or incremental backup)
At least some operating systems seem to have the concept of an “archive” bit which added whenever a file is changed. Full backups ignore but reset this bit. Differential backups use the archival bit but don’t reset it. Incremental backups use and reset the archival bit.
Defense in Depth
Defense in depth stems from the philosophy that complete security against threats can never be achieved; the components that comprise a layered security strategy only impede threat progress until either the attacker gives up or the organization can respond to the threat.
ITPro.TV: CompTIA Security+ (SY0-601)
Hardware Redundancy
“High availability” is considered 99.999% uptime.
RAID 0 is capped at 32 disks.
Note that all all RAID systems - even RAID 0 - write information sequentially across drives in order to maximize performance (block 1 to drive 1, block 2 to drive 2, and so on). In RAID 5 there is also a parity block that isn’t used for reads but is used to reconstruct data in the event of a disk failure. RAID 5 writes the parity block in a rotating fashion between drives (so that no two parity blocks follow each other and parity blocks are evenly distributed between drives).
Site Redundancy
(General) redundancy types:
- Hot (everything is ready to go, just on standby)
- Warm (key systems are ready to go, but some reconfiguration required)
- Cold (equipment might exist, but no configuration/readiness)