|
Per-Port Configuration
|
Global Configuration
|
PortFast State
|
PortFast BPDU Filtering State
|
|
Default
|
Enable
|
Enable
|
Enable
|
|
Default
|
Enable
|
Disable
|
Disable
|
|
Default
|
Disable
|
Not applicable
|
Disable
|
|
Disable
|
Not applicable
|
Not applicable
|
Disable
|
|
Enable
|
Not applicable
|
Not applicable
|
Enable
|
●
Root Guard - Prevents switches connected on ports
configured as access ports from becoming the root switch. Enforces the position
of the root bridge. When a root guard port gets a BPDU with a higher priority,
it transitions into a special STP state (root-inconsistent) which is the
effectively same state as listening and no traffic is passed through. When it
stops receiving superior BPDUs, it will actually transition to regular STP.
Enable root guard using the following interface-level command: [no] spanning-tree guard root
●
show
spanning-tree summary totals
●
show
spanning-tree summary
●
show
spanning-tree interface fa0/1 detail
●
show
spanning-tree inconsistentports
●
Black holes in a network are created when a device that
receives frames has no forwarding information for that packet and thus
essentially drops all such packets.
●
Loop Guard - Improves the stability of Layer 2 networks
by preventing bridging loops. When a blocked port stops receiving BPDUs,
instead of transitioning to a forwarding state, it’ll go into a
loop-inconsistent state. Can be configured on a per-port basis and will only
block on a per-VLAN basis. Enable Loop Guard on all non-designated ports.
Should be enabled on root and alternative ports for all possible combinations
of active topologies. Interface level command to enable: spanning-tree guard loop. You can enable it globally on all ports
using the following command: spanning-tree
loopguard default. To disable loop guard on an interface: no spanning-tree loopguard. To verify
loopguard on an interface: show
spanning-tree interface interface-id detail
●
UDLD - Detects and disables unidirectional links
(traffic is being transmitted between neighbors in one direction only - can
cause topology loops). Will shut down an interface with this issue. Useful on a
fiber port to prevent network issues resulting in miswiring at the patch panel
causing the link to be up/up status but the BPDUs are lost. Layer 2 protocol
that works with layer 1 mechanisms. With UDLD enabled, the switch periodically
sends UDLD protocol packets to it’s neighbor and expects the packets to be
echoed back before the predetermined timer expires. If the timer expires, the
switch determines the link is unidirectional and shuts it down. Default
interval is 15 seconds which is configurable.
●
UDLD uses MAC 01-00-oc-cc-cc-cc with Subnetwork Access
Protocol (SNAP) High-Level Data Link Control (HDLCC) protocol type 0x0111.
Packet contains info about sending port’s device ID and port ID and the
neighbor’s device ID and port ID.
●
In normal mode, UDLD simply changes the port to an
undetermined state if it stops receiving UDLD messages from it’s directly
connected neighbor. Aggressive mode UDLD tries to reestablish the connection
with the neighbor and after 8 failed retries, the port state changes to the
err-disabled state.
●
Aggressive mode is the preferred method of configuring
UDLD.
●
In EtherChannel, UDLD shuts down only the physical link
that failed.
●
Renabling the port is simply shutting it down and no
shutting it.
●
In order for UDLD to work, it must be enabled on both
ports on both sides
UDLD Default Config Status:
|
Feature
|
Default Status
|
|
UDLD Global enable state
|
Globally disabled
|
|
UDLD per-interface enable state
for fiber-optic media
|
Enabled on all ethernet fiber-optic
interfaces
|
|
UDLD per-interface enable state
for twisted-pair (copper) media
|
Disabled on all Ethernet 10/100
and 1000BAST-TX interfaces
|
●
To enable UDLD on an interface: udld enable [aggressive]
●
To enable UDLD globally on fiber-optic interfaces: udld {enable | aggressive}
●
To disable UDLD on individual non-fiberinterfaces: no udld enable
●
To disable individual fiber-optic interface: no udld port
●
Verify config: show
udld gigabitethernet 0/1
●
A designated port transmits BPDUs whereas a
nondesignated port receives BPDUs
Comparison between Loop Guard and Aggressive Mode UDLD
|
|
Loop Guard
|
Aggressive Mode UDLD
|
|
Configuration
|
Per port
|
Per port
|
|
Action granularity
|
Per VLAN
|
Per port
|
|
Auto-recovery
|
Yes
|
Yes, with err-disable timeout
feature
|
|
Protection against STP failures
caused by unidirectional links
|
Yes, when enabled on all root
ports and alternative ports in redundant topology
|
Yes, when enabled on all links
in redundant topology
|
|
Protection against STP failures
caused by problem in software resulting in designated switch not sending
BPDUs
|
Yes
|
No
|
|
Protection against miswiring
|
No
|
Yes
|
●
Enabling both aggressive mode UDLD and LoopGuard
provides highest level of protection against bridging loops and black holes in
multilayer switches
●
Flex Links is a layer 2 availability feature that
provides an alternative solution to STP and allows users to turn off STP and
still provide basic link redundancy. Can coexist with STP in distribution
switches. Can enable a convergence time of less than 50 milliseconds
consistently regardless of the number of VLANs or MAC addresses configured on
the uplink ports.
●
Based on
active/standby link pair on common access switch. Pair of layer 2 interfaces
(switchports or port channels) that are configured to act as a backup to
another layer 2 interface.
●
Configured on one L2 interface (active) by assigning
another L2 interface as the Flex Links or backup links. Can be on the same
switch or other switch in the stack. When one link is up and forwarding, the
other link is in standby and ready to forward if the other link shuts down. If
active goes down, standby goes up. When active comes back up, standby goes down
again. Supported only on L2 ports/port channels, not VLANs or L3 ports.
●
Guidelines on Flex Links:
○
Can configure only one Flex Links backup link for any
active link and it much be a different interface from the active interface
○
Interface can only belong to one Flex Links pair and
can be a backup for only on active link. Active link cant’ below to another
Flex Links pair.
○
Neither of the links can be a port that belongs to an
EtherChannels but you can configure two port channels (Etherchannel logical
interfaces) as Flex Links. Can configure a port channel and a physical
interface as Flex Links with either the port channel or the physical interface
as the active link.
○
Backup link does not have to be the same type of media
as the active link but you should configure both to have similar
characteristics so there aren’t loops or changes in behavior if standby begins
to forward traffic
○
STP is disabled on Flex Links ports. That port doesn’t
participate in STP even if the VLANs present on the port are configured for
STP. When STP is not enabled, be sure there are no loops in the configured
topology.
●
Flex Links are configured at the interface level with
the command: switchport backup interface
○
Example:
■
interface
fa1/0/1
■
switchport
backup interface fa1/0/2
■
end
■
show interface
switchport backup
●
To limit the limitations of not having multipathing and
networkwide failure domain, the general recommendation is to use L3
connectivity at the distribution or core layer of the network and keeping L2
for the access layer.
●
Using L3 between the distribution and the core layer
allows multipathing (up to 16 paths) using Equal-Cost Multipathing (ECMP)
without dependency of STP and preferred unless there is a need to extend L2
across a data center pod (distribution block).
●
ECMP refers to situation where router has multiple
equal-cost paths to a prefix and load-balances traffic over each path.
●
RSTP superior to 802.1D and PVST+ from a convergence
perspective. If network includes other vendor switches, isolate the different
STP domains with L3 routing to avoid STP compatibility issues
●
To configure a VLAN instance to become the root bridge,
enter the spanning-tree vlan vlan_ID root command to modify the
bridge priority from the default value (32768) to a lower value. Manually
placing the primary and secondary bridges enables you to support a
deterministic configuration where you know which ports should be forwarding and
blocking
Troubleshooting STP:
●
Duplex mismatch on point-to-point links can cause
issues in STP which could lead to a bridging loop. Collisions on one side would
prevent BPDUs from being sent and assumes the root bridge is no longer present
and then shifts another port into forwarding state. Correct this by checking
and adjusting to full duplex
●
Unidirectional Link Failure can lead to bridging loops
and routing black holes. Resolve by configuring aggressive mode UDLD to detect
incorrect cabling or unidirectional links which will put the affected port in
err-disabled state. Recommended practice is to use aggressive mode UDLD on all
point-to-point interfaces in any multilayer switched network
●
Frame corruption - high rate of physical errors may
result in lost BPDUs which leads the interface to move to forwarding state.
Very rare this happens since the port would need to miss BPDUs for over 50
seconds and successful BPDUs received breaks the loop. More common for
nondefault STP parameters and aggressive STP timer values. This is generally a
result of duplex mismatch, bad cable, or incorrect cable length
●
Resource Errors - If CPU is over-utilized, might lack
resources to send out BPDUs. Look at product documentation for recommended
number of VLANs and STP instances as not to exhaust resources
●
PortFast Configuration Error - If PortFast configured
on the wrong ports, the fast transition can lead to bridging loops.
●
Troubleshooting Methodology:
○
Develop a plan
○
Isolate the cause and correct an STP problem
○
Document findings
●
Develop a Plan - Understand the following
○
Topology of the bridged network
○
Location of the root bridge
○
Location of the blocked ports and, therefore, the
redundant links
●
Isolate the cause and correct an STP problem:
○
1 - Identify a bridging loop
■
Capture traffic on saturated link and check for
duplicated packets
■
Check port utilization on devices - Look for abnormal
values
■
Look for other protocols (like HSRP) breaking down
■
Check for lapping MAC addresses which doesn’t happen in
a stable network
■
Don’t associate a bridging loop with a packet storm
○
2 - Restore connectivity
■
Break the loop - Can involve disabling every port that
is redundant manually. Identify the part of the network that is more affected
and start there. If possible, start by disabling ports that should be in
blocking state first.
■
Enable logging and increase the logging level of STP
events on switches with failures. Use debug commands (debug spanning-tree). Use following command to capture debug
information into the logging buffer: logging
buffered
○
3 - Check port status
■
Blocked ports - check that switch reports receiving
BPDUs periodically on root and blocked ports. Issue the following to display
the information: show spanning-tree vlan
vlan-id detail
■
Duplex mismatch - Use the show interface command on specified ports
■
Port utilization - show
interface will show if interface utilization using the load of the
interface and the packet input and output rates
■
Frame corruption - look for increases in input error
fields of the show interface command
○
4 - Check for resource errors
■
Use the show
processes cpu command for this. Switches prioritize control packets such as
BPDU over any lower-priority traffic. If the CPU exceeds 70 percent, action
should be taken immediately or the network should be considered for
re-architecture
○
5 - Disable unneeded features
■
Disable as many features as possible to reduce
troubleshooting complexity. Reapply after root cause is found.
●
Document findings to improve upon future issues
No comments:
Post a Comment