GCGA Ch. 9 Adding Redundancy and Fault Tolerance (ST)

Question 1

Single point of failure

Answer 1

any component that can cause the entire system to fail if it fails. It normally refers to hardware but can be a person. If one person is the only person who can perform a task, that person can be a single point of failure.

Question 2

RAID disk subsystems

Answer 2

provide fault tolerance and increase availability. RAID-1 (mirroring) uses two disks. RAID-5 uses three or more disks and can survive the failure of one disk. RAID-6 and RAID-10 use four or more disks and can survive the failure of two disks.

Question 3

Load balancers

Answer 3

spread the processing load over multiple servers. In an active/active configuration, all servers are actively processing requests. In an active/passive configuration, at least one server is not active but is instead monitoring activity ready to take over for a failed server. Software-based load balancers use a virtual IP.

Question 4

Affinity scheduling

Answer 4

sends client requests to the same server based on the client’s IP address. This is useful when clients need to access the same server for an entire online session. Round-robin scheduling sends requests to servers using a predefined order.

Question 5

NIC teaming

Answer 5

groups two or more physical network adapters into a single software-based network adapter. It provides load balancing for outgoing traffic and fault tolerance if one of the NICs fails.

Question 6

Power redundancies

Answer 6

include a UPS, a dual power supply, and generators. Managed PDUs monitor the quality of power delivered to devices within a server rack.