Network Bulls
www.networkbulls.com
Best Institute for CCNA CCNP CCSP CCIP CCIE Training in India
M-44, Old Dlf, Sector-14 Gurgaon, Haryana, India
Call: +91-9654672192
Like voice traffic, video traffic is latency-sensitive. Therefore, many of the same design
and troubleshooting considerations for voice traffic (for example, QoS considerations)
also apply to video traffic.
Class Bulk:
No data found.
Class Scavenger:
No data found.
Class Management:
No data found.
Class Routing:
Recommended Minimum Bandwidth: 0 Kbps/0% (AverageRate)
Detected applications and data:
Application/ AverageRate PeakRate Total
Protocol (kbps/%) (kbps/%) (bytes)
-------- -------- -------- --------
eigrp 0/0 0/0 1110
icmp 0/0 0/0 958
Class Best Effort:
Current Bandwidth Estimation: 44 Kbps/<1% (AverageRate)
Detected applications and data:
Application/ AverageRate PeakRate Total
Protocol (kbps/%) (kbps/%) (bytes)
-------- -------- -------- --------
http 44/<1 121/1 372809
unknowns 0/0 0/0 232
Suggested AutoQoS Policy for the current uptime:
!
class-map match-any AutoQoS-Voice-Fa0/0
match protocol rtp audio
!
policy-map AutoQoS-Policy-Fa0/0
class AutoQoS-Voice-Fa0/0
priority percent 1
set dscp ef
class class-default
fair-queue
R4# conf term
R4(config)# int fa 0/0
R4(config-if)# auto qos
R4(config-if)# end
R4#
www.CareerCert.info
www.CareerCert.info
www - CareerCert - info
Chapter 11: IP Communications Troubleshooting 347
IP
V
Gateway/
Gatekeeper
Call Agent
Unified
Messaging
Server
Cisco Unified Video
Advantage
Cisco TelePresence
System
Cisco
TelePresence
System
H.323 Video
Conferencing
System
H.323 Video
Conferencing
System
Cisco Unified Video
Advantage
TelePresence
MCU
IP WAN
ISDN
ISDN
PSTN
ISDN
V
Gateway
V
Figure 11-5 IP-Based Video Network
An additional consideration for video traffic, however, is multicasting. Specifically, a video
server might send a single video stream to a multicast group. End-user devices wanting to
receive the video stream can join the multicast group. Multicast configuration, however,
needs to be added to routers and switches to support this type of transmission. With the
addition of multicast configurations comes the potential of additional troubleshooting
targets. This section describes the basic theory, configuration, and troubleshooting of a
multicast network.
This section then concludes with a listing of potential video network troubleshooting
targets.
Introduction to IP-Based Video
Figure 11-5 illustrates a sample IP-based video network.
www.CareerCert.info
www.CareerCert.info
www - CareerCert - info
348 CCNP TSHOOT 642-832 Official Certification Guide
Key
Topic
Table 11-5 Recommended QoS Metrics for Video
QoS Metric Cisco Unified Video
Advantage
Cisco TelePresence Video Surveillance
One-Way
Delay
200 ms maximum 150 ms maximum 500 ms maximum
Jitter 10 ms maximum 10 ms maximum 10 ms maximum
Packet Loss 0.05 percent
maximum
0.05 percent maximum 0.5 percent maximum
Many components in the figure are identical to components seen in a voice network.
Three types of video solutions are shown, as follows:
■ H.323 Video Conferencing System:Multiple third parties offer H.323 video conferencing
systems, which can be used to set up a video conference over an IP or ISDN
network.
■ Cisco Unified Video Advantage: The Cisco Unified Video Advantage product uses
a PC, a video camera, and a Cisco IP Phone as a video conferencing station. Specifically,
the camera connects to a USB port on the PC. Software is loaded on the PC,
and the PC is connected to the PC port on a Cisco IP Phone. Alternately, the Cisco IP
Phone could be the software-based Cisco IP Communicator running on the PC.
When a voice call is placed between two users running the Cisco Unified Video Advantage
product, a video call can automatically be started, with the video appearing
on each user’s PC.
■ Cisco TelePresence: The Cisco TelePresence solution uses CD-quality audio and
High Definition (HD) video (that is, 1080p) displayed on large monitors to create lifelike
video conferences.
Design Considerations for Video
Due to the bandwidth-intensive and latency-sensitive nature of video, consider the following
when designing or troubleshooting a video network:
■ QoS: Like voice, video packets need to be allocated an appropriate amount of bandwidth
and be treated with high priority. Table 11-5 shows the QoS metrics that Cisco
recommends for various types of video applications.
■ Availability: Also like voice, video networks should be built on an underlying data
network with reliable components and redundancy, such that the availability of the
video network can approach an uptime of 99.999 percent.
■ Security: Just as an eavesdropper could capture unencrypted voice packets and interpret
the information contained in those packets, unencrypted video packets could
also be captured and interpreted. Therefore, appropriate security measures, such as
encryption and authentication, should be in place in a video network.
www.CareerCert.info
www.CareerCert.info
www - CareerCert - info
Chapter 11: IP Communications Troubleshooting 349
■ Multicasting:More common in a video environment rather than in a voice environment
is the use of multicasting technologies. Multicasting allows a multicast server to
send traffic (for example, a video stream) to a destination Class D IP address known
as a multicast group. End stations wanting to receive the traffic sent to the multicast
group can join the group, thus allowing the multicast server (for example, a video
server) to send a single stream of traffic, which is received by multiple recipients
wishing to receive the traffic.
Multicasting
QoS, availability, and security considerations were discussed in the previous section.
However, this chapter has not yet addressed multicasting in detail. Therefore, this section
introduces you to multicasting operation, configuration, and troubleshooting.
Introduction to Multicasting
Consider a video stream that needs to be sent to multiple recipients in a company. One approach
is to unicast the traffic. The source server sends a copy of every packet to every receiver.
Obviously, this approach has serious scalability limitations.
An alternate approach is to broadcast the video stream, so that the source server only has
to send each packet once. However, everyone within a broadcast domain of the network
receives the packet, in that scenario, even if they do not want it.
IP multicast technologies provide the best of both worlds. With IP multicast, the source
server only sends one copy of each packet, and packets are only sent to intended recipients.
Specifically, receivers join a multicast group, denoted by a Class D IP address (that is, in
the range 224.0.0.0 through 239.255.255.255). The source sends traffic to the Class D address,
and through switch and router protocols, packets are forwarded only to intended
stations. These multicast packets are sent via UDP (that is, best effort). When doing a multicast
design, also be aware of the potential for duplicate packets being received and the
potential for packets arriving out of order.
Figure 11-6 shows a simple multicast network, where a multicast server is sending traffic
to a destination IP address of 224.1.1.1. Notice that there are three workstations in this
network. However, only two of the three workstations joined the multicast group. Therefore,
only the two multicast group members receive traffic from the multicast server.
Internet Group Management Protocol
The protocol used between clients (for example, PCs) and routers to let routers know
which of their interfaces have multicast receivers attached is Internet Group Management
Protocol (IGMP). As an example, Figure 11-7 shows a PC sending an IGMP Join message
to a multicast-enabled router. The router receives this IGMP Join message on its Fast Ethernet
0/0 interface. Therefore, the router knows that when it receives traffic destined for a
particular multicast group (as identified in the IGMP Join message), that traffic should be
forwarded out its Fast Ethernet 0/0 interface.
www.CareerCert.info
www.CareerCert.info
www - CareerCert - info
www.CareerCert.info
www.CareerCert.info
www - CareerCert - info
Chapter 11: IP Communications Troubleshooting 351
IGMPv2 Router
X
IGMPv2 Receiver IGMPv1 Receiver
Leave Message
Figure 11-8 Mixed IGMPv1 and IGMPv2 Topology
type appears to be invalid, and it is ignored. Therefore, an IGMPv2 host must send
IGMPv1 reports to an IGMPv1 router.
In an environment with an IGMPv2 router and a mixture of IGMPv1 and IGMPv2 receivers,
the version 1 receivers respond normally to IGMPv1 or IGMPv2 queries.
However, as illustrated in Figure 11-8, a version 2 router must ignore any leave message
while IGMPv1 receivers are present because if the router processed the IGMPv2
leave message, it would send a group-specific query, which would not be correctly interpreted
by an IGMPv1 receiver.
As mentioned earlier, multicast routers can periodically send queries out of an interface to
determine if any multicast receivers still exist off of that interface. However, there might
be a situation where more than one multicast router exists on a broadcast media segment
(for example, Ethernet). Therefore, one router must be designated as the querier for that
segment. This IGMP designated querier is the router that has the lowest unicast IP address.
To determine which router on a multi-access network is the querier, you can issue
the following command:
Router# show ip igmp interface [interface-id]
The output from this command identifies the IP address of the IGMP querier. Additionally,
the following command displays the IP multicast groups of which a router is aware:
Router# show ip igmp group
When a Layer 2 switch receives a multicast frame on an interface, by default, the switch
floods the frame out all other interfaces. To prevent this behavior, the switch needs awareness
of what interfaces are connected to receivers for specific multicast groups. IGMP
snooping is a feature that can be enabled on many Cisco Catalyst switches, which allows
a switch to autonomously determine which interfaces are connected to receivers for specific
multicast groups by eavesdropping on the IGMP traffic being exchanged between
www.CareerCert.info
www.CareerCert.info
www - CareerCert - info
352 CCNP TSHOOT 642-832 Official Certification Guide
clients and routers. To globally enable IGMP snooping on a Cisco Catalyst switch, issue
the following command:
Switch(config)# ip igmp snooping
Once enabled globally, individual VLANs can be enabled or disabled for IGMP snooping
with the following command:
Switch(config)# ip igmp snooping vlan vlan_id
Multicast Addressing
In a multicast network, the multicast source sends multicast packets with a Class D destination
address. The 224.0.0.0 through 239.255.255.255 address range is the Class D address
range, because the first four bits in the first octet of a Class D address are 1110.
Some ranges of addresses in the Class D address space are dedicated for special purposes:
■ Reserved Link Local Addresses: 224.0.0.0–224.0.0.255. These addresses are used,
for example, by many network protocols. OSPF uses 224.0.0.5 and 224.0.0.6. RIPv2
uses 224.0.0.9, and EIGRP uses 224.0.0.10. Other well-known addresses in this range
include 224.0.0.1, which addresses all multicast hosts, and 224.0.0.2, which addresses
all multicast routers.
■ Globally Scoped Addresses: 224.0.1.0–238.255.255.255. These addresses are used
for general-purpose multicast applications, and they have the ability to extend beyond
the local autonomous system.
■ Source-Specific Multicast (SSM) Addresses: 232.0.0.0–232.255.255.255. These
addresses are used in conjunction with IGMPv3, to allow multicast receivers to request
not only membership in a group but also to request specific sources from
which to receive traffic. Therefore, in an SSM environment, multiple sources with different
content can all be sending to the same multicast group.
■ GLOP Addresses: 233.0.0.0–233.255.255.255. These addresses provide a globally
unique multicast address range, based on autonomous system numbers.
■ Limited Scope Addresses: 239.0.0.0–239.255.255.255. These addresses are used
for internal multicast applications (for example, traffic that doesn’t leave an autonomous
system), much like the 10.0.0.0/8 address space is a private IP address space.
In addition to Layer 3 addresses, multicast applications must also have Layer 2 addresses
(that is, MAC addresses). Fortunately, these Layer 2 addresses can be constructed directly
from the Layer 3 multicast addresses. A MAC address is a 48-bit address, and the first half
(that is, 24 bits) of a multicast MAC address (in hex) is 01-00-5e. The 25th bit is always 0.
The last 23 bits in the multicast MAC address come directly from the last 23 bits of the
multicast IP address. Consider the following example:
Given a multicast IP address of 224.1.10.10, calculate the corresponding multicast MAC
address.
Step 1. First, convert the last three octets to binary:
0000.0001.0000.1010.0000.1010
www.CareerCert.info
www.CareerCert.info
www - CareerCert - info
www.CareerCert.info
www.CareerCert.info
www - CareerCert - info
www.CareerCert.info
www.CareerCert.info
www - CareerCert - info
www.CareerCert.info
www.CareerCert.info
www - CareerCert - info
356 CCNP TSHOOT 642-832 Official Certification Guide
4. If a receiver comes up on a router that was previously pruned from the tree, that
router can rejoin the tree by sending a Graft message.
A major consideration for PIM-DM, however, is that this flood-and-prune behavior repeats
every three minutes. Therefore, PIM-DM does not scale well. A better alternative to
PIM-DM is PIM Sparse Mode (PIM-SM).
PIM-SM Mechanics
Next, consider the formation of a PIM Sparse Mode distribution tree:
1. A receiver sends an IGMP Report message to its router indicating that it wants to participate
in a particular multicast group. The receiver’s router (that is, the last-hop
router) sends a Join message to the RP, creating (*, G) state along a shared tree between
the RP and the last-hop router.
2. A source comes up and creates a source tree between its router (that is, the first-hop
router) and the RP. (S, G) state is created in routers along this path. However, before
the source tree is completely established, the source sends its multicast packets to the
RP encapsulated inside of unicast Register messages.
3. After the RP receives the first multicast packet over the source tree, it sends a
Register Stop message to the source, telling the source to stop sending the multicast
traffic inside of Register messages. Two trees now exist: (1) a source tree from the
first-hop router to the RP and (2) a shared tree from the RP to the last-hop router.
However, this might not be the optimal path.
4. The last-hop router discovers from where the multicast traffic is arriving, and the lasthop
router sends a Join message directly to the first-hop router to form an optimal
path (that is, a source path tree) between the source and the receiver.
5. Because the last-hop router no longer needs multicast traffic from the RP, as it is receiving
the multicast traffic directly from the first-hop router, it sends an (S, G) RP-bit
Prune message to the RP, requesting the RP to stop sending multicast traffic.
6. With the shared tree to the last-hop router pruned, the RP no longer needs to receive
multicast traffic from the first-hop router. So, the RP sends an (S, G) Prune message
to the first-hop router. At this point, traffic flows in an optimal path from the firsthop
router to the last-hop router. The process of cutting over from the path via the
RP to the direct path is called Shortest-Path Tree (SPT) Switchover.
Comparing PIM-DM versus PIM-SM suggests that PIM-SM offers the benefits of
PIM-DM (that is, optimal pathing) without PIM-DM’s flood-and-prune behavior.
You can determine a distribution tree’s topology by examining the multicast routing table
of multicast routers in the topology. The show ip mroute command displays a router’s
multicast routing table, as demonstrated in Example 11-8.
www.CareerCert.info
www.CareerCert.info
www - CareerCert - info
Chapter 11: IP Communications Troubleshooting 357
Example 11-8 show ip mroute Command Output
Notice the highlighted (*, G) and (S, G) entries. Other valuable information contained in
the mroute table includes the incoming interface (IIF), which shows on which interface
traffic is entering the router, and the outgoing interface list (OIL), which shows the router
interfaces over which the multicast traffic is being forwarded.
Rendezvous Points
In a PIM-SM network, one or more routers need to be designated as a rendezvous point
(RP). Non-RPs can be configured to point to a statically defined RP with the global configuration
mode command ip pim rp-address ip-address. However, in larger topologies,
Cisco recommends that RPs be automatically discovered. Cisco routers support two
methods for automatically discovering an RP: Auto-RP and Bootstrap Router (BSR). Although
Auto-RP is a Cisco approach, BSR is a standards-based approach to make the location
of RPs known throughout a multicast network.
Common Video Troubleshooting Issues
As with voice networks, data networking troubleshooting targets should be considered
when troubleshooting a reported video issue. Additionally, as a reference, Table 11-6 offers
a collection of common video troubleshooting targets and recommended solutions.
Router# show ip mroute
IP Multicast Routing Table
Flags: D - Dense, S - Sparse, B - Bidir Group, s - SSM Group, C - Connected,
L - Local, P - Pruned, R - RP-bit set, F - Register flag,
T - SPT-bit set, J - Join SPT, M - MSDP created entry,
X - Proxy Join Timer Running, A - Candidate for MSDP Advertisement,
U - URD, I - Received Source Specific Host Report, Z - Multicast Tunnel,
Y - Joined MDT-data group, y - Sending to MDT-data group
Timers: Uptime/Expires
Interface state: Interface, Next-Hop or VCD, State/Mode
(*, 224.0.100.4), 02:37:12, RP is 192.168.47.14, flags: S
Incoming interface: Serial0, RPF neighbor 10.4.53.4
Outgoing interface list:
Ethernet1, Forward/Sparse, 02:37:12/0:03:42
Ethernet2, Forward/Sparse, 02:52:12/0:01:23
(192.168.46.0/24, 224.0.100.4), 02:37:12, flags: RT
Incoming interface: Ethernet1, RPF neighbor 10.4.53.4
Outgoing interface list:
Ethernet2, Forward/Sparse, 02:44:21/0:01:47
www.CareerCert.info
www.CareerCert.info
www - CareerCert - info
358 CCNP TSHOOT 642-832 Official Certification Guide
Table 11-6 Common Video Troubleshooting Targets
Video Troubleshooting Target Recommended Solutions
Bandwidth Video streams can be bursty in nature and consume
large quantities of bandwidth. Therefore, although sufficient
bandwidth should be allocated for supported
video applications, you should confirm that the video
traffic is not consuming too much bandwidth (that is,
an amount of bandwidth that would negatively impact
other important traffic).
Pervasiveness of Video
Applications
The volume of video traffic on a network might be
somewhat unpredictable, because users might introduce
their own video traffic on a network without the
knowledge of network administrators. Therefore, your
policy for network use should address the types of
traffic a user is allowed to send and receive. Also, you
might want to block video from portions of your network.
Security In addition to protecting the content of your network’s
video streams, realize that security measures you have
in place might be conflicting with your video applications.
For example, if a video stream cannot be established,
you might check your firewalls and router ACLs
to confirm they are not blocking video media (that is,
RTP) packets, video maintenance (for example, RTCP)
packets, or video-signaling packets (for example,
H.323).
QoS Because video traffic is latency-sensitive, QoS mechanisms
should be in place to ensure video packets are
sent with priority treatment, and sufficient bandwidth
should be allocated for your supported video applications.
Multicast Because many video applications rely on multicast
technologies to transmit a video stream to a multicast
group, much of your video troubleshooting might be
focused on multicast troubleshooting. For example,
confirm that both routers and switches are properly
configured with multicast protocols (for example,
PIM-SM on a router and IGMP Snooping on a switch).
No comments:
Post a Comment