3.3 Explain high availability and disaster recovery concepts and summarize which is the best solution.

Which of the following RAID levels does not provide fault tolerance? a) RAID 0 b) RAID 1 c) RAID 5 d) RAID 10

A. Redundant Array of Independent Disks (RAID) is a technology for storing data on multiple hard disk drives, providing fault tolerance, increased performance, or both. The various RAID levels provide different levels of functionality and have different hardware requirements. RAID 0 uses data striping only (blocks written to each disk in turn), which does not provide any form of fault tolerance. RAID 1 provides disk mirroring, RAID 5 combines disk striping with distributed storage of parity information, and RAID 10 creates mirrored stripe sets—these three levels all provide fault tolerance.

Which of the following is not a type of server load balancing mechanism? a) DNS round-robin b) Network Address Translation c) Content switching d) Multilayer switching

B. Load balancing is a method of distributing incoming traffic among multiple servers. Network Address Translation (NAT) is a routing mechanism that enables computers on a private network to share one or more public Internet Protocol (IP) addresses. It is therefore not a load balancing method. Domain Name System (DNS) round-robin, multilayer switching, and content switching are all mechanisms that enable a server cluster to share client traffic.

In a disaster recovery scenario, which of the following terms applies to devices that are not repairable? a) MTBF b) MTTF c) MTTR d) MDT

B. Mean Time to Failure (MTTF) refers to devices that will eventually fail once and then be discarded, rather than repaired. Mean Time Between Failure (MTBF), Mean Time to Repair (MTTR), and Mean Down Time (MDT) all refer to devices that will eventually fail and then be repaired and reused.

When you configure NIC teaming on a server with two network adapters in an active/passive configuration, which of the following services is provided? a) Load balancing b) Fault tolerance c) Server clustering d) Traffic shaping

B. NIC teaming enables you to combine the functionality of two Network Interface Cards (NIC) in one connection. However, when you configure a NIC team to use an active/passive configuration, one of the network adapters remains idle and functions as a fault tolerance mechanism. If the other NIC should fail, the passive NIC becomes active. In this configuration, NIC teaming does not provide load balancing, server clustering, or traffic shaping.

f you back up your network by performing a full backup every Wednesday at 6:00 p.m. and incremental backups in the evening of the other days of the week, how many jobs would be needed to completely restore a computer with a hard drive that failed on a Monday at noon? a) One b) Two c) Five d) Six

C. An incremental backup is a job that backs up all of the files that have changed since the last backup of any kind. Therefore, to restore a system that failed on Monday at noon, you would have to restore the most recent full backup from the previous Wednesday and the incrementals from Thursday, Friday, Saturday, and Sunday.

Which of the following disaster recovery mechanisms can be made operational in the least amount of time? a) A cold site b) A warm site c) A hot site d) All of the options are the same.

C. Cold, warm, and hot backup sites differ in the hardware and software they have installed. A cold site is just a space at a remote location. The hardware and software must be procured and installed before the network can be restored. It is therefore the least expensive and takes the most time. A warm site has hardware in place that must be installed and configured. A hot site has all of the necessary hardware already installed, configured, and ready to go in the event of a disaster. A warm site is more expensive than a cold site, and a hot site is the most expensive and takes the least amount of time to be made operational.

Why does performing incremental backups to a hard drive, rather than a tape drive, make it possible to restore a server with a single job, rather than multiple jobs? a) Because hard drives hold more data than tape drives b) Because hard drives can transfer data faster than tape drives c) Because hard drives are random access devices and tape drives are not d) Because hard drives use a different block size than tape drives

C. Data is stored on tape drives in a linear fashion. Once you write backup data to a tape, you cannot selectively replace individual files. When you perform a restore job, you might have to restore the most recent full backup, followed by incremental backups, which overwrites some of the full backup files with newer ones. Hard disk drives are random access devices, meaning that individual files can be written to and read from any location on the disk. When you perform incremental backup jobs to a hard disk, the software can restore data using any version of each file that is available. Data capacity, transfer speed, and block size are not relevant.

To build a network with redundant Internet connections that ensures full fault tolerance, it is necessary to use which of the following? a) Different ISPs b) Different WAN connections c) Different routers d) All of the above

D. Access to the Internet can be interrupted by a failure on the Internet Service Provider's (ISP's) network, by a failure on the Wide Area Network (WAN) provider's network, or by a router failure on the local network. Building redundancy into all of these elements is the best way to ensure continuous access to the Internet.

For a complete restore of a computer that failed at noon on Tuesday, how many jobs would be needed if you performed full backups to tape at 6:00 a.m. every Wednesday and Saturday and incremental backups to tape at 6:00 a.m. every other day? a) One b) Two c) Three d) Four

D. An incremental backup is a job that backs up all of the files that have changed since the last backup of any kind. Therefore, to restore a system that failed on Tuesday at noon, you would have to restore the most recent full backup from the previous Saturday and the incrementals from Sunday, Monday, and Tuesday morning.

Which of the following statements about port aggregation is not true? a) All of the aggregated ports use the same MAC address. b) Port aggregation can be a fault tolerance mechanism. c) Aggregating ports increases network throughput. d) Port aggregation provides load balancing

D. Load balancing refers to the distribution of traffic between two or more channels. Port aggregation combines ports into a single logical channel with a single Media Access Control (MAC) address and provides greater throughput. Port aggregation also provides fault tolerance in the event of a port failure.

Which of the following is not a load balancing mechanism? a) NIC teaming b) Server clustering c) DNS round-robin d) RAID 1

D. Redundant Array of Independent Disks (RAID) level 1 is a fault tolerance mechanism that is also known as disk mirroring. A storage subsystem writes data to two or more disks at the same time so that if a disk fails, the data remains available. Because data is written to the disks at the same time, this RAID level does not provide load balancing. NIC teaming balances a network traffic load among two or more Network Interface Cards (NICs), whereas server clustering and Domain Name Service (DNS) round-robin balance a traffic load among multiple servers.

If you have a server with dual power supplies, each of which is connected to a separate Uninterruptible Power Supply (UPS), with each UPS connected to a separate building power circuit connected to a backup generator, which of the following failures can the server survive and keep running indefinitely? (Choose all that apply.) a) Failure of one server power supply b) Failure of one UPS c) Failure of one building power circuit d) Failure of the building backup generator

A, B, C, D. If one of the server's power supplies fails, the other will continue to function. If one of the UPSs fails, the server will continue to run using the other. If one of the building power circuit breakers trips, the server will continue to run using the other one. If the building's backup generator fails, the server will continue to run as long as the building still has outside power.

Which of the following are equivalent terms for the process of combining the bandwidth of two or more network adapters to increase the overall speed of the connection and provide fault tolerance? (Choose all that apply.) a) Bonding b) Link aggregation c) Clustering d) Port aggregation e) NIC teaming

A, B, D, E. Bonding, link aggregation, port aggregation, and Network Interface Card (NIC) teaming are all terms for the same basic technology, in which the bandwidth of multiple network adapter connections is joined to speed up transmissions. The technology also enables the network communication to continue if one of the adapters should be disconnected. Clustering refers to combining servers into a single unit, not network adapters.

Concurrent multipath routing (CMR) is a technique that provides which of the following benefits? (Choose all that apply.) a) Increased bandwidth b) Fault tolerance c) Data encapsulation d) Load balancing

A, B, D. CMR is a technique in which routers transmit data packets using multiple routes to the destination. This provides load balancing and improved performance by spreading the transmission among multiple paths and a measure of fault tolerance because a failure of one route only endangers part of the data. Data encapsulation is not involved in the CMR process.

If you have a server with dual power supplies, one of which is plugged into a single Uninterruptible Power Supply (UPS) and the other into a wall socket with a surge protector, and the building's power circuit is connected to a backup generator, which of the following failures can the server survive and keep running indefinitely? (Choose all that apply.) a) Failure of one server power supply b) Failure of the UPS c) Failure of the building power circuit d) Failure of the building backup generator

A, B, D. If one of the server's power supplies fails, the other will continue to function. If the UPS fails, the server will continue to use the power supply plugged into the wall socket. If the building's backup generator fails, the server will continue to run as long as the building still has outside power. If the breaker for the building power circuit trips, the server will run only as long as the UPS battery holds out.

Which of the following networking concepts frequently use virtual Internet Protocol (IP) addresses to provide high availability? (Choose all that apply.) a) Clustering b) Load balancing c) Network Address Translation (NAT) d) Network Interface Card (NIC) teaming

A, B. A high availability virtual IP address implementation is when multiple servers are identified by a single address, enabling all of the servers to receive incoming client traffic. In the case of server clustering and network load balancing arrangements, the cluster itself has a unique name and IP address, separate from those of the individual servers. Clients address themselves to the cluster, not to one of the servers in the cluster. NAT is not a high availability technology, and NIC teaming does not use virtual IP addresses.

Which of the following types of backup jobs are supported by the Windows Server Backup program? (Choose all that apply.) a) Incremental b) Differential c) Full d) Supplemental

A, C. Windows Server Backup can perform full backups and incremental backups. It does not support differential backups, and there is no backup job called a supplemental.

Which of the following specifications would you most want to examine when comparing hard disk models for your new Redundant Array of Independent Disks (RAID) array? a) MTBF b) SLA c) AUP d) MTTR

A. Mean Time Between Failure (MTBF) specifies how long you can expect a device to run before it malfunctions. For a hard disk, this specification indicates the life expectancy of the device. A Service Level Agreement (SLA) and an Acceptable Use Policy (AUP) are not specifications associated with hard disk drives. Mean Time to Repair (MTTR) can conceivably be specified for a hard disk, but hard disk drives in a RAID array are typically replaced, not repaired.

Which of the following statements about backing up a firewall's state and backing up its configuration are true? (Choose all that apply.) a) A backup of a firewall's state contains more data than a backup of its configuration. b) A backup of a firewall's configuration contains more data than a backup of its state. c) A backup of a firewall's state contains the exact same data as a backup of its configuration. d) A backup of a firewall's state includes its configuration. e) A backup of a firewall's configuration includes its state.

A, D. A backup of a firewall's state includes its configuration as well as other elements, such as templates and policies. The state therefore contains more data than the configuration.

If you have a server with dual power supplies, both of which are connected to a single Uninterruptible Power Supply (UPS), with a building power circuit connected to a backup generator, which of the following failures can the server survive and keep running indefinitely? (Choose all that apply.) a) Failure of one server power supply b) Failure of the UPS c) Failure of the building power circuit d) Failure of the building backup generator

A, D. If one of the server's power supplies fails, the other will continue to function. If the building's backup generator fails, the server will continue to run as long as the building still has outside power. If the UPS fails, the server will go down. If the breaker for the building power circuit trips, the server will run only as long as the UPS battery holds out.

Redundant servers running in an active-active configuration provide which of the following advantages that the same servers in an active-passive configuration do not? (Choose all that apply.) a) Increased performance b) Fault tolerance c) Data encapsulation d) Load balancing

A, D. In an active-active configuration, servers can balance the incoming client load between them. Because the active servers are all servicing clients, the overall performance of the cluster is increased. Both active-active and active-passive configurations provide fault tolerance. Data encapsulation is not a factor in either configuration.

Which of the following mechanisms for load balancing web servers is able to read the incoming HTTP and HTTPS requests and perform advanced functions based on the information they contain? a) Content switches b) Multilayer switches c) Failover clustering d) DNS round-robin

A. A content switch is an application layer device, which is what renders it capable of reading the incoming Hypertext Transfer Protocol (HTTP) and Hypertext Transfer Protocol Secure (HTTPS) messages. HTTP is an application layer protocol. Multilayer switches do not operate above the transport layer. Failover clustering and (DNS) round-robin are both techniques for distributing incoming traffic without actually processing it.

Which of the following is a term for a read-only copy of a data set made at a specific moment in time? a) Snapshot b) Incremental c) Hot site d) Differential

A. A snapshot is a read-only copy of a data set taken at a specific moment in time. By creating a snapshot and then backing it up, you can be sure that no data corruption has occurred due to version skew. A hot site is an alternative network location in which all hardware and software is installed and ready. Incrementals and differentials are types of backup jobs.

Which of the following disaster recovery mechanisms is the least expensive to implement? a) A cold site b) A warm site c) A hot site d) All of the options cost the same.

A. Cold, warm, and hot backup sites differ in the hardware and software they have installed. A cold site is just a space at a remote location. The hardware and software must be procured and installed before the network can be restored. It is therefore the least expensive. A warm site has hardware in place that must be installed and configured. A hot site has all of the necessary hardware installed and configured. A warm site is more expensive than a cold site, and a hot site is the most expensive.

Which of the following statements about the differences between online and standby Uninterruptible Power Supplies (UPSs) are correct? (Choose all that apply.) a) A standby UPS runs devices using battery power all the time. b) An online UPS provides no gap in the power supplied to the devices during a main power failure. c) An online UPS switches devices to battery power only during a main power failure. d) A standby UPS provides only enough power for an orderly shutdown of the devices.

B, D. It is an online UPS that runs devices using battery power all the time so that there is no gap to the power supplied to devices during a failure. It is a standby UPS that switches devices to battery power during a main power failure. Both online and standby UPSs provide only enough power for an orderly shutdown of the devices.

Which of the following RAID levels provide fault tolerance without using parity data? (Choose all that apply.) a) RAID 0 b) RAID 1 c) RAID 5 d) RAID 10

B, D. Redundant Array of Independent Disks (RAID) is a technology for storing data on multiple hard disk drives, providing fault tolerance, increased performance, or both. The various RAID levels provide different levels of functionality and have different hardware requirements. RAID 1 and RAID 10 both use disk mirroring to provide fault tolerance, which does not require parity data. RAID 0 uses data striping only (blocks written to each disk in turn), which does not provide any form of fault tolerance. RAID 5 combines disk striping with distributed storage of parity information.

Which of the following RAID levels provides fault tolerance with the smallest amount of usable disk space? (Choose all that apply.) a) RAID 0 b) RAID 1 c) RAID 5 d) RAID 10

B, D. Redundant Array of Independent Disks (RAID) is a technology for storing data on multiple hard disk drives, providing fault tolerance, increased performance, or both. The various RAID levels provide different levels of functionality and have different hardware requirements. RAID 1 provides disk mirroring, and RAID 10 creates mirrored stripe sets. Both provide fault tolerance by maintaining two copies of every stored file, for a usable disk space percentage of 50 percent. Some mirroring configurations store more than two copies of each file, for even less usable space. RAID 0 provides data striping only, with no fault tolerance. RAID 5 combines disk striping (blocks written to each disk in turn) with distributed storage of parity information, for fault tolerance with a usable disk space percentage of at least 66 percent.

Which of the following is the primary difference between the Power Distribution Units (PDUs) used in datacenters and the standard power strips used in offices and homes? a) More outlets b) Larger power input c) Larger power output d) All of the above

B. A PDU for a datacenter performs the same basic function as an office power strip, but it typically has a larger power input. It does not necessarily have more outlets or a larger power output.

If you back up your network by performing a full backup every Wednesday at 6:00 p.m. and differential backups in the evening on the other six days of the week, how many jobs would be needed to completely restore a computer with a hard drive that failed on a Tuesday at noon? a) One b) Two c) Six d) Seven

B. A differential backup is a job that backs up all the files that have changed since the last full backup. Therefore, to restore a system that failed on Tuesday at noon, you would have to restore the most recent full backup from the previous Wednesday and the most recent differential from Monday.

How does an autochanger increase the overall storage capacity of a backup solution? a) By compressing data before it is stored on the medium b) By automatically inserting media into and removing it from a drive c) By running a tape drive at half its normal speed d) By writing two tracks at once onto a magnetic tape

B. An autochanger is a robotic device containing one or more removable media drives, such as magnetic tape or optical disk drives. The robotic mechanism inserts and removes media cartridges automatically so that a backup job can span multiple cartridges, increasing its overall capacity.

A network load balancing cluster is made up of multiple computers that function as a single entity. Which of the following terms is used to describe an individual computer in a load balancing cluster? a) Node b) Host c) Server d) Box

B. In a network load balancing cluster, each computer is referred to as a host. Other types of clusters use other terms. For example, in a failover cluster, each computer is called a node. The terms _server_ and _box_ are not used in clustering.

Which of the following are valid reasons why online Uninterruptible Power Supplies (UPSs) are more expensive than standby UPSs? a) Online UPSs enable devices to run longer when a main power failure occurs. b) Online UPSs enable devices to run continuously when a main power failure occurs. c) Online UPSs are managed devices that can generate alerts. d) Online UPSs provide greater protection against power spikes and sags.

B. Online UPSs run devices from the battery all the time, while simultaneously keeping the battery charged. There is therefore no switchover gap when a power failure occurs. Online UPSs do not necessarily run longer than standby UPSs, nor do they provide more protection again power spikes and sags. Both online and standby UPSs can be managed devices.

Installing an electrical generator for your datacenter is an example of which of the following fault tolerance concepts? a) Uninterruptible Power Supply (UPS) b) Power redundancy c) Dual power supplies d) Redundant circuits

B. Power redundancy is a general term describing any fault tolerance mechanism that enables equipment to continue functioning when one source of power fails. A UPS is a device that uses battery power, not a generator. The term _dual power supplies_ refers to the power supply units inside a computer, not a separate generator. The term _redundant circuits_ refers to multiple connections to the building's main power, not to a generator.

You are installing a new Windows server with two hard disk drives in it, and you want to use RAID to create a fault-tolerant storage system. Which of the following RAID levels can you configure the server to use? a) RAID 0 b) RAID 1 c) RAID 5 d) RAID 10

B. Redundant Array of Independent Disks (RAID) is a technology for storing data on multiple hard disk drives, providing fault tolerance, increased performance, or both. The various RAID levels provide different levels of functionality and have different hardware requirements. RAID 1 provides disk mirroring for fault tolerance and requires two or more disk drives. RAID 0 provides data striping only, with no fault tolerance. RAID 5 combines disk striping (blocks written to each disk in turn) with distributed storage of parity information for fault tolerance, but it requires a minimum of three disk drives. RAID 10 creates mirrored stripe sets and requires at least four disk drives.

Which of the following terms defines how long it will take to restore a server from backups if a complete system failure occurs? a) RPO b) RTO c) BCP d) MIB

B. The Recovery Time Objective (RTO) specifies the amount of time needed to restore a server from the most recent backup if it should fail. This time interval depends on the amount of data involved and the speed of the backup medium. A Recovery Point Objective (RPO) specifies how much data is likely to be lost if a restore from backups should be necessary. This figure is based on the frequency of the backups and the amount of new data generated by the system. Business contingency planning (BCP) is an umbrella term for procedures enacted to keep the organization functioning in the event of a disaster. A Management Information Base (MIB) is a database used by Simple Network Management Protocol (SNMP) systems.

Which of the following is the primary function of an Uninterruptible Power Supply (UPS) during a two-hour building power failure. a) To keep servers running until building power is restored b) To keep servers running until they can be powered down safely c) To keep servers from being damaged by power spikes d) To keep servers running in the event of a computer power supply failure

B. UPSs can provide servers with battery backup power, but usually only for a few minutes, so that the servers can be powered down safely, without the potential for data corruption or loss. UPSs cannot keep servers running for two hours. UPSs can protect against power spikes, but that is not their primary function. A computer power supply failure will bring a server down, regardless of the presence of a UPS.

Which of the following backup job types does not reset the archive bits of the files it backs up? a) Full b) Incremental c) Differential d) Supplemental

C. Differential backups use the archive bit to determine which target files to back up. However, a differential backup does not reset the archive bit. Full backups do not pay attention to the archive bit, because they back up all of the files. A full backup, however, does clear the archive bit after the job is completed. Incremental backups also use the archive bit to determine which files have changed since the previous backup job. The primary difference between an incremental and a differential job, however, is that incremental backups clear the archive bit so that unchanged files are not backed up. There is no such thing as a supplemental backup job.

A First Hop Redundancy Protocol (FHRP) is designed to dynamically alter which one of the following Internet Protocol (IP) configuration settings on a network host? a) IP address b) Subnet mask c) Default gateway d) DNS server address

C. FHRPs provide a fault tolerant default gateway for network hosts by automatically failing over to an alternative router address in the event of a router failure. Protocols such as Virtual Router Redundancy Protocol (VRRP) and Hot Standby Router Protocol (HSRP) create a virtual router that hosts use for their default gateway. The virtual router contains the addresses of multiple physical routers, to which it sends packets transmitted by the hosts. This provides fault tolerance in the event that a router fails, and in some cases provides load balancing as well. FHRPs do not affect the hosts' IP addresses, subnet masks, or DNS server addresses.

Which of the following is an element of high availability systems that enables them to automatically detect problems and react to them? a) Backups b) Snapshots c) Failover d) Cold sites

C. Highly available systems often have redundant failover components that enable them to continue operating even after a failure of a router, switch, hard disk, server, or other component. Backups, snapshots, and cold sites can all contribute to a system's high availability, but they do not function automatically.

Which of the following is not a fault tolerance mechanism? a) Port aggregation b) Clustering c) MTBF d) UPS

C. Mean Time Between Failures (MTBF) is a hardware specification used to predict the approximate lifetime of a component. It does not refer to any type of fault tolerance mechanism. Port aggregation, clustering, and Uninterruptible Power Supplies (UPSs) are all mechanisms that provide fault tolerance in the event of network adapter, server, and power failures, respectively.

Which of the following RAID levels uses disk striping with distributed parity? a) RAID 0 b) RAID 1 c) RAID 5 d) RAID 10

C. Redundant Array of Independent Disks (RAID) is a technology for storing data on multiple hard disk drives, providing fault tolerance, increased performance, or both. The various RAID levels provide different levels of functionality and have different hardware requirements. RAID 5 combines disk striping (blocks written to each disk in turn) with distributed storage of parity information, for fault tolerance. RAID 0 provides data striping only. RAID 1 provides disk mirroring. RAID 10 creates mirrored stripe sets.

Which of the following is not a fault tolerance mechanism? a) UPS b) RAID 1 c) SNMP d) Clustering

C. Simple Network Monitoring Protocol (SNMP) is a means of tracking the performance and functionality of network components. Software or firmware components called agents are embedded in network devices and communicate with a central monitoring console. SNMP does not provide fault tolerance. An Uninterruptible Power Supply (UPS) is a battery backup device that enables a computer to continue functioning in the event of a power failure. Redundant Array of Independent Disks (RAID) level 1 is a disk mirroring mechanism that provides fault tolerance by maintaining duplicate copies of all stored data. Clustering is a mechanism by which multiple servers function as a single unit, running the same application, so that if a server should fail, the others continue to function.

Which of the following is the criterion most commonly used to filter files for backup jobs? a) Filename b) File extension c) File attributes d) File size

C. The archive bit that backup software uses to perform incremental and differential jobs is a file attribute, so this is the most commonly used filter type. It is possible to filter files based on their names, their extensions, and their size, but these are not used as often as the archive file attribute.

What are the three elements in the Grandfather-Father-Son media rotation system for network backups? a) Hard disk drives, optical drives, and magnetic tape drives b) Incremental, differential, and full backup jobs c) Monthly, weekly, and daily backup jobs d) QIC, DAT, and DLT tape drives

C. The generational media rotation system uses the terms _grandfather_, _father_, and _son_ to refer to backup jobs that are run monthly, weekly, and daily. The jobs can be full, incremental, or differential, and the terms have nothing to do with whether the backup medium is a hard disk, optical, or any type of tape drive.

Which of the following datacenter disaster recovery mechanisms is the least expensive to implement? a) A cold site b) A warm site c) A hot site d) A cloud site

D. A cloud site is the least expensive to implement. Cold, warm, and hot backup sites differ in the hardware and software they have installed, but they all require the maintenance of a facility for a new datacenter. A cold site is just a space at a remote location. The hardware and software must be procured and installed before the network can be restored. It is therefore the least expensive of the cold, warm, and hot options and takes the most time. A warm site has hardware in place that must be installed and configured. A hot site has all of the necessary hardware already installed, configured, and ready to go in the event of a disaster. A warm site is more expensive than a cold site, and a hot site is the most expensive and takes the least amount of time to be made operational. A cloud site is a virtual facility maintained with a cloud service provider. The cloud site does not require physical space or physical hardware, so it is therefore the least expensive.

Which of the following can be provided by clustering servers? a) Fault tolerance b) Load balancing c) Failover d) All of the above

D. A cluster is a group of computers configured with the same application that function as a single unit. The cluster can function as a fault tolerance mechanism by failing over from one server to the next, when necessary, or provide load balancing by distributing traffic among the servers.

Which of the following best describes the difference between cold, warm, and hot backup sites? a) Whether the backup site is owned, borrowed, or rented b) The age of the most recent backup stored at the site c) The cost of the hardware used at the site d) The time needed to get the site up and running

D. Cold, warm, and hot backup sites are a disaster recovery mechanism that enables a network to be activated at a remote location when a catastrophe occurs. The temperature refers to the sites readiness to assume the role of the network. A cold site is just a space at a remote location. The hardware and software must be procured and installed before the network can be restored. A warm site has hardware in place that must be installed and configured. It takes less time to restore the network than at a cold site, but more than at a hot site. A hot site has all of the necessary hardware installed and configured. The network can go live as soon as the most recent data is restored.

Disk mirroring and disk duplexing are both fault tolerance mechanisms for hard disk data storage. Which of the following statements about disk mirroring and disk duplexing is true? a) Disk mirroring enables a server to survive the failure of a disk drive. b) Disk duplexing enables a server to survive the failure of a disk controller. c) Disk mirroring enables a server to survive a failure of a disk drive or a disk controller. d) Disk duplexing enables a server to survive a failure of a disk drive or a disk controller.

D. Disk mirroring and disk duplexing both use multiple hard disk drives to store duplicate copies of all data. However, disk duplexing calls for each disk to be connected to a separate controller so that the data remains available despite a disk failure or a disk controller failure.

hich of the following is not a First Hop Redundancy Protocol (FHRP)? a) CARP b) VRRP c) HSRP d) RARP

D. FHRPs provide a fault tolerant default gateway for network hosts by automatically failing over to an alternative router address in the event of a router failure. The Reverse Address Resolution Protocol (RARP) is a deprecated Internet Protocol (IP) address assignment protocol; it is not an FHRP. Common Address Redundancy Protocol (CARP), Virtual Router Redundancy Protocol (VRRP), and Hot Standby Router Protocol (HSRP) are all FHRPs.

Which of the following is the most commonly recommended fire suppression system for a datacenter? a) Carbon Dioxide b) Water c) Foam d) HFC-125

D. Hydrofluorocarbon (HFC) 125 is a heat-absorbing gas that is frequently used for fire suppression systems. Many large datacenters include HFC-125 total flooding fire suppression systems because the products resulting from the flame retardation process are less toxic than carbon dioxide and do not damage electric and electronic equipment as water and foam do.

Which of the following storage techniques prevents version skew from occurring during a system backup? a) Incrementals b) Differentials c) Iterations d) Snapshots

D. Version skew can occur when a data set changes while a system backup is running. A file written to a directory that has already been backed up will not appear on the backup media, even though the job might still be running. This can result in unprotected files, or worse, data corruption. A snapshot is a read-only copy of a data set taken at a specific moment in time. By creating a snapshot and then backing it up, you can be sure that no data corruption has occurred due to version skew. Incrementals and differentials are types of backup jobs, and iteration is not a specific storage technology.