Content Overview Troubleshooting Layer
2 problems can be a challenging process. Although data link
protocols are usually quite simple, the configuration and
operation of these protocols is critical to creating a
functional, well-tuned network. This module explores the
following Layer 2 protocols and technologies and outlines
troubleshooting strategies and common problems for each of
them: - Switches and switched technologies
- PPP
- ISDN
- Frame Relay
The
interaction of Layer 3 protocols with various Layer 2 protocols
is often the source of network problems. Specific examples and
trouble shooting techniques for routed and routing protocols in
various Layer 2 environments are also examined.
Content
4.1 Characteristics of Data Link Layer Problems
4.1.1 Data link layer problems overview Data
link layer problems have common symptoms that assist in
identifying Layer 2 issues. Recognizing these symptoms will
help the troubleshooter narrow down the number of possible
causes of a problem. Before examining the individual protocols
and their unique characteristics, it is helpful to know the
symptoms that are associated with data link layer problems in
general. Lab Activity Lab Exercise: Isolating Physical
and Data Link Layer Problems After completing this lab, the
student will be able to analyze user feedback and end-system
data to decide at which OSI layer to begin isolating problems.
Content 4.1 Characteristics of Data Link
Layer Problems 4.1.2 No component above the
data link layer is functional Because each layer of the OSI
model encapsulates the layer above it, any failure of a Layer 2
protocol on a link will prevent the valid exchange of Layer 3
through Layer 7 information. Some Layer 2 problems can stop the
exchange of frames across a link, while others only cause
network performance to degrade. For example, consider a simple
Layer 2 problem. A leased line WAN link is misconfigured such
that the router at one end of the link is configured using
Frame-Relay as the encapsulation type and the router at the
other end is using HDLC as the encapsulation type. Obviously,
HDLC frames received by the Frame-Relay interface will be
meaningless to the router. Therefore, no information can be
exchanged at Layer 2. In a Cisco environment this could be
confirmed using the show cdp neighbor command. If no
information is received about a neighbor when CDP is enabled,
that is a very good indication that Layer 2 frames are not
being received. Since Layer 2 is not functioning, a Layer 3
test such as ping would also fail. Layer 4 TCP services would
also fail, which might be indicated by the failure of a BGP
router to peer with neighboring routers. All applications which
require the services of lower layers of the OSI model will be
affected. Although this is a simple concept, it underpins one
of the fastest ways to narrow down a problem. Network faults
are often first indicated by the failure of Layer 7
applications which require network connectivity. The first
instinct of the troubleshooter would be to test connectivity at
lower layers to attempt to isolate the failure at a specific
OSI model layer or layers. The network layer ping
command is the most convenient method for testing connectivity.
If a ping is successful, the lower layers can be
eliminated as possible sources of the problem. However, if the
pings fail, even intermittently, the next step is to find the
boundary of the Layer 3 problem. In other words, how far does
the ping work? After examining the Layer 3 configuration of the
router, the Layer 2 operation at the edge of this boundary can
be confirmed. In a Cisco environment this is likely to involve
the show cdp neighbor command. The troubleshooter could
also examine the output of the show interfaces command
to determine the status of the line protocol. In most instances
Layer 2 problems are not difficult to troubleshoot because the
problems are confined to communications between directly
connected devices, and Layer 2 protocols tend to be less
complex than Layer 3 protocols. The troubleshooting process is
enhanced by the ability to drill down through the OSI model to
determine if Layer 2 issues exist. Lab Activity Lab
Exercise: Correcting Problems at the Physical and Data Link
Layers In this lab, the student will use various Cisco commands
to correct network problems.
Content 4.1
Characteristics of Data Link Layer Problems
4.1.3 The network is functional, but operating at less
than the baseline level While complete data-link layer
failures are relatively easy to identify and rectify, it is
much more difficult to isolate problems caused by suboptimal
Layer 2 operations. The fact that some, or all, frames do
actually reach their destination can make it difficult to
determine the boundary of the problem and focus on the devices
or link causing the problem. There are two distinct types of
suboptimal Layer 2 operation that can occur in a network:
- Frames take an illogical path to their destination, but do
arrive.
- Some frames are dropped.
Although
these are very different problems that can be caused by Layer 2
or Layer 1 faults or design issues, they share some similar
symptom: - Network delays.
- Lack of throughput.
- Poor network application performance.
It
is important to remember that the loss of some frames can be
tolerated by many applications, since TCP will retransmit the
lost segment. For delay sensitive traffic such as voice, video,
or SNA of IBM, suboptimal levels of performance can actually
bring the application to a halt. An example of a problem which
would cause frames to take an illogical path through the
network would be a poorly designed Layer 2 spanning-tree
topology. In this case, the network might experience high
bandwidth utilization on links that should not have that level
of traffic. Problems that cause frames to actually be dropped
can be identified through error counter statistics and console
error messages that appear on the switch or router. In an
Ethernet environment, an extended or continuous ping will also
reveal if any frames are being dropped. There are Layer 2
protocols, such as X.25, that have their own retransmission
mechanisms which will ensure that dropped frames are
retransmitted. In this case, it can be difficult to determine
if frames are being dropped. Variations in round trip times for
ping replies can indicate a problem, but to be certain
that frames are being lost or corrupted, an examination of the
error counters on the interface would be required.
Content 4.1 Characteristics of Data Link Layer
Problems 4.1.4 Framing errors The data-link
layer uses standardized frame formats and physical addressing
to establish communications within a broadcast domain. If
devices disagree on the frame format communications will fail.
Problems with the frame format usually translate to
encapsulation incompatibilities between nodes. Actual framing
errors occur when a frame does not end on an 8-bit byte
boundary for one of the following reasons: - Noisy
serial line.
- Improperly designed cable. This could be
caused by serial cable that is too long, or the cable from the
CSU or DSU to the router is not shielded.
- The CSU
line clock is incorrectly configured. For example, the clock on
one of the CSUs may be configured for local clocking.
- Ones density problem on T1 link. This is caused by
incorrect framing or a coding specification.
All of
these problems result in a receiver having difficulty
establishing where one frame ends and another frame starts.
When the interface is capable of recognizing this condition,
the show interfaces command will reveal an incrementing
frame error count. These problems are critical as they will
cause the communications protocol itself to fail or frames will
be incorrectly addressed and fail to reach their destination.