Layer 2 Protocol Troubleshooting and Commands
sw4# show int fa0/21
■ Runts—Runts are frames smaller than 64 bytes.
■ CRC errors—The frame’s cyclic redundancy checksum value does not match the one
calculated by the switch or router.
■ Frames—Frame errors have a CRC error and contain a noninteger number of octets
■ Alignment—Alignment errors have a CRC error and an odd number of octets.
■ Collisions—Look for collisions on a full-duplex interface, or excessive collisions on a halfduplex
interface.
■ Late collisions on a half-duplex interface—A late collision occurs after the first 64 bytes of
a frame.
show controllers fastEthernet 0/0
##############################################
Troubleshooting EtherChannels
Remember that the following rules apply to all ports within an EtherChannel:
■ Speed and duplex must match.
■ Interface type—access, trunk, or routed—must match.
■ Trunk configuration—encapsulation, allowed VLANs, native VLAN, and DTP mode—must
match.
■ If a Layer 2 EtherChannel is not a trunk, all ports must be assigned to the same VLAN.
■ No port in the EtherChannel can be a Switched Port Analyzer (SPAN) port.
■ On a Layer 3 EtherChannel, the IP address must be on the PortChannel interface, not a
physical interface.
L3SW4# show etherchannel summary
####################################
Layer 3 Troubleshooting Process
You can expect that many difficulties that appear at Layer 3 are not really Layer 3 problems at all,
but rather are the result of troubles in other layers of the protocol stack. Here are some examples
of issues at other layers that can impact Layer 3 protocols in subtle or misleading ways:
■ An MTU mismatch on a link
■ A unidirectional link
■ A duplex mismatch
■ A link with a high error rate in one or both directions
■ Layer 2 configuration issues
■ Access list (ACL or VACL) filtering with unintended consequences (don’t forget that implicit
deny!)
■ Security policy that blocks required traffic
■ A TTL setting that’s too low for Layer 3 protocol operation
■ Two or more Layer 3 subnets configured in the same VLAN, which is especially problematic
with Layer 3 protocols that use broadcast or multicast traffic to form adjacencies
■ Mismatched subnet masks within a subnet.
■ TTL too short can cause some routing protocol adjacencies (specifically eBGP) to fail to
form, or stop IP communications from taking place across a path with multiple Layer 3 hops.
■ MTU too low on a link can cause large packets to be dropped.
■ MTU mismatch on a link can cause large packets to be dropped on the low-MTU end when
they arrive.
■ Multicast traffic is not supported, disabled, or rate-limited on one or more links.
■ An overloaded link can result in packet loss, long latency, and jitter.
■ QoS configuration can cause packet loss, especially of keepalives.
After you’ve gotten past these core IP issues, you can begin to look for more in-depth issues at
Layer 3. These are likely to be specific to routing protocol configuration or operation. However,
in keeping with the scope of this section, we won’t consider simpler, one-command issues such as
adjacencies failing to form or authentication failures. These issues are covered in the earlier
chapters of this book. Some of the common sources of problems in routing include the following:
Incorrect split-horizon configuration. This is challenging to find quickly because the result is
usually that most routes are propagated correctly, but some are not propagated.
■ Incorrect redistribution configuration, especially with multiple points of redistribution or mutual
redistribution. Incorrectly configured filtering or a lack of filtering can cause routing loops.
■ Protocols not advertising routes when they appear to be configured to do so.
■ Protocols not redistributing routes when they appear to be configured to do so.
■ Incorrect route filtering because of incorrect masks applied in an access list or prefix list.
■ EIGRP stuck-in-active (SIA) issues.
■ Incorrect summarization.
■ Administrative distance manipulation causing fundamental routing rules to be superseded.
■ Metric calculations configured differently on different routers (particularly affecting metric
calculations in OSPF or mismatched EIGRP k values).
■ Metric manipulation on a router.
■ NAT configuration with unintended consequences.
■ Policy-based routing configuration issues or unintended consequences.
■ Interface dampening activity causing intermittent or flapping operation.
■ Mismatched timer settings, which sometimes result in adjacencies flapping.
阅读(736) | 评论(0) | 转发(0) |
