BackboneFast allows ports to transition immediately when an indirect (non-local) failure occurs.

UplinkFast, not BackboneFast, allows ports to transition immediately when a direct (local) failure occurs.

BackboneFast is used in environments with a network diameter greater than 3. BackboneFast must be installed on each switch in the network.

2. Answer D is correct. A hub basically connects hosts together at a central point, introducing a hub merely makes that number of hosts on that same segment or collision domain larger.

Remember, a collision domain is defined as a group of hosts that share the same physical copper wiring. If you have 10 hosts plugged into a hub, they are all sharing the same physical segment.

If that hub is plugged into another hub using the Uplink port, then all the hosts plugged into both hubs are sharing the same physical wire. Each host now has to contend (using Carrier Sense Multiple Access/Collision Detection) for transmission time on the segment, thus making collisions far more likely to occur.

A hub does nothing except repeat everything it hears in any port out all other ports.

Routers create separate broadcast and collision domains.

Switches create separate collision domains.

3. Answer D is correct. Configuring full-duplex mode on an interface requires the "duplex full" command.

The command line is shown here:

Switch(enable) set port duplex module/port {full | half}

If the interface was shut down, the word "shut" would follow the "duplex full" entry. Therefore, the commands indicate that the interface is using full duplex mode.

The 172.16.*.* network is a private IP address network.

Since the last configuration line states "duplex full", the option that the router's fa0 interface is operating in half-duplex mode is incorrect.

4. Answer A is correct. The port status LEDs are all green on the switch when the connection is ready. They illuminate when there is link and stay on.

The LED associated with each port is an indicator of connectivity (continuity) between the port and the host or workstation connected to that port. If the LED on the port is not green and illuminated, that is generally an indicator of a problem.

The system status LED is always green when there is power to the device and it is functioning properly.

A basic rule of thumb is that green is "good" and red is "bad". A light is amber when it is initializing.

5. Answer A is correct. If both sides are set to nonegotiate the link will trunk.

If the nonegotiate mode is set on either side of the link, the encapsulation method (ISL, dot1Q, dot10, etc.) will not be negotiated to agreement. Despite this fact, the ports will trunk. The other side of the link must have trunking set to either on or nonegotiate to ensure successful trunking. It is also important to remember that the encapsulation method is the same on both sides.

A port trunk can be in one of five different trunk negotiation modes:

• on - A port with its trunking method set to on causes the ports to permanently trunk (until you manually change the mode). There is no negotiation of any kind. It is important to set the type of trunking when using this method because the encapsulation method (ISL, dot1Q, dot10, etc.) is not negotiated.
• off - The port will not trunk and will actually negotiate with the other side to be sure it does not trunk. However, if the other side is set to on, there are connectivity issues.
• desirable - The port will negotiate with the other side of the link to configure the link as a trunk. This setting makes the port trunk. However, if the other side is unable to trunk, the port does not trunk.
• auto - The port is neutral in this mode. If the other side of the link attempts to trunk, they trunk. If the other side does not attempt to trunk, they do not.
• nonegotiate - This mode causes the port to trunk, and it will not negotiate in any way with the other side of the link. The other side of the link must have trunking set to either on or nonegotiate. If the other side of the link is set to negotiate or auto, no trunking will occur, and the link does not work.

It should also be noted that in all modes but nonegotiate, the switches must be in the same management domain.

6. Answers D and E are correct. CST is the use of a single version of the Spanning Tree Protocol (STP) by IEEE's 802.1q trunking method. It reduces processing power and bandwidth.

With only a single version of the spanning tree running, there will be far fewer bridge protocol data units (BPDUs) than if you use Cisco Per-VLAN Spanning Tree (PVST). This conserves bandwidth and utilization.

CST also has some disadvantages, most notably the fact that it can only maintain a single route bridge. This can cause sub-optimal pathing decisions in some cases, as well as longer convergence times because all switch ports are included in the STP calculation run.

7. Answer A is correct. The best way to keep the concept of an RSM from getting too confusing is to think of it as being a separate piece of equipment. Treat inter-VLAN routing configuration tasks no differently than you would traditional campus network routing.

To route traffic in a campus network, you need:

• Unique Layer 3 addresses for each router interface.
• Routing protocols or static route so routers can learn where destination networks are.

The same applies when you set up inter-VLAN routing with an RSM.

VLANs are separate logical segments, despite the fact that they exist within the switch fabric. As mentioned above, the two steps needed to properly configure inter-VLAN routing are:

• Assign an IP address to each logical segment that you want to route for.
• Implement static routes and /or dynamic routing protocols.

Issuing the "ip default-gateway" command configures a router to send all traffic with an unknown destination to this address; you can also configure a routing protocol like any other Cisco router. Of course, all local VLAN subnets are known as they are defined on the VLAN interfaces.

An "external routing protocol" gives the RSM a view of routes outside of its Autonomous System (AS).

This information is learned in the same manner as any other router inside the AS. Therefore, no special consideration is needed in this respect.

8. Answer A is correct. Unicasts and multicasts both deliver packets to multiple recipients, but they do so in different ways. Unicasts sends traffic to only a single receiver at a time whereas multicasts send traffic to a number of receivers at the same time.

Think of a unicast as like having a room full of people listening to a lecture. The person giving the lecture needs to whisper his entire lecture into the ear of each person in the room. He may need to give his speech dozens or even hundreds of times to make sure that each person heard the lecture. In a unicast, the same traffic in its entirety is transmitted to each node on the network. This can consume available bandwidth very quickly and is extremely inefficient.

Remember the equation for sending streaming video (1.5 Mbps) over a network:

1.5 x N (N equals the number of clients viewing the stream) = Mbps of required bandwidth to handle the viewing requests.

In a normal lecture, one in which the speaker does whisper his lecture individually to each attendee, the speaker performs the equivalent of a multicast or broadcast of the message one time to all the users.

9. Answer A is correct. When translating IP multicast addresses to Layer 2 MAC addresses, the last 23 bits of the IP are duplicated at the last 23 bits of the MAC address. The first 25 bits are preset at 01-00-5e, with a 0 in the 25th bit position.

Using this question yields:

227 16 28 1
227.16.28.1= 11-10-00-11.00-01-00-00.00-01-11-00.00-00-00-01
Layer 2 MC =00-00-00-01-00-00-00-00-01-01-11-10-00-01-00-00 00-01-11-00 00-00-00-01
In Hex 0 1 - 0 0 - 5 e - 1 0 - 1 c - 0 1

10. Answers A, B and F are correct. By running a different instance of the spanning tree protocol per VLAN, the size of the spanning tree is reduced, the processing time is reduced, and the time required by the switch to calculate or recalculate the spanning tree is reduced.

All of these benefits of reduced processing time decrease the time to recover from a change and allow for more efficient paths through the bridged network.

You have to remember that a bridged network has a single path through it, which may not be the most efficient approach, especially as the spanning tree gets larger.

11. Answer A is correct. NBAR allows UDP and TCP to be inspected beyond the simple port numbers so it can be arranged by host name and various other characteristics.

WRED queues traffic by keeping a minimum and maximum threshold for each queue. When the queue is between this minimum and maximum, packets are randomly dropped. When the queues are above the maximum threshold, packets are dropped as they are received. When the queue is below the minimum, packets cannot be dropped.

Tail-Drop queuing drops newly queued packets when the egress queue is full. Tail-Drop can cause serious performance issues and should not be implemented unless there is a specific reason.

FIFO queuing allows traffic to traverse the interface in the order it is received so that the first packet in, is the first packet out.

Questions and answers provided by MeasureUp. To order the full version of this exam simulation, click here.

More Pop Quiz:

• Cisco Exam #642-871: ARCH Exam (MeasureUp, set 2)
• Cisco Exam #642-811: BCMSN Exam (Whizlabs, set 2)
• Cisco Exam #642-811: BCMSN Exam (Whizlabs, set 1)
• Cisco Exam #642-871: ARCH Exam (MeasureUp, set 1)