Advance VPLS Technology Atachements Options

Advanced VPLS attachment technology options
Session BRKRST-2008
Patrice Bellagamba
Distinguished System Engineer
Cisco Europe
Problem statement
Problem to solve:
Ethernet service Attachment-Circuit redundancy should not imply
end to end Spanning-tree usage
STP is struggling with
- Topology diameter
- Fault isolation between sites

- Resilience to WAN type connections
Position Architecture: STP domain isolation thru dual homing

- Carrier-Ethernet
- Data-Center Interconnect
- Ethernet Services
Presentation_ID
2011 Cisco and/or its affiliates. All rights reserved.
Cisco Public
Access Multi-homing Solutions Summary

Highlights
Multi-chassis LAG
Simple solution for spoke-and-hub topology, works for both bridging and nonbridging access device
Standard based solution by using 802.3ad
Sub-second convergence
Phase 1 implement is active/standby mode.
Phase 2 is per VLAN load balancing
Ring topology support is under investigation
REP /REP access

gateway
Sub 200msec convergence

Good access ring isolation
Now standard based G.8032 (XR4.1 release)
Spoke-and-hub and ring topology, not works well for mesh network
MST/PVST access
gateway
Standard based solution as long as access network support MST/PVST

Works for any access network topology
Good access domain isolation
Work with 802.1ah PBB
Convergence time depends on access network STP
Node clustering
VSS (Catalyst / Cisco 7600-Sup2T)

Nv cluster (ASR9k)
One control-plane for two chassis
Easiness, Active/Active
Presentation_ID
Cisco Public
L2 node peering solutions

Making usage of an Inter-Chassis Communication Protocol
Service Type
L2VPN
Transport
Enabler
Access
Redundancy
ICCP based
Protocol / Feature
E-LINE
VPWS
Hub and Spoke

(Active / Backup)
mLACP + 2-way PW Red.

(coupled mode)
E-LAN
VPLS
Hub and Spoke

(Active / Backup)

(decoupled mode)
E-LAN
H-VPLS
Hub and Spoke

(Active / Backup)

(decoupled mode)
DCI
VPLS
CPE service
mLACP or pseudo-mLACP
DCI
EVPN
PE service
mLACP + BGP
Dont be scared of all acronyms, they will be explained along the presentation
Presentation_ID
Cisco Public
Inter-Chassis Communication Protocol for L2VPN PE

Redundancy (ICCP)
draft-martini-pwe3-iccp-06
Inter-Chassis Communication Protocol

RG
ICCP allows two or more devices to form a

Redundancy Group
ICCP provides a control channel for
synchronizing state between devices
ICCP uses TCP/IP as the underlying
transport
ICCP over Dedicated Link
-ICCP rides on targeted LDP session, but

MPLS need not be enabled
RG
Various redundancy applications can

use ICCP:
-mLACP
-Pseudowire redundancy
Under standardization in IETF:

-draft-ietf-pwe3-iccp-06.txt
ICCP over Shared Network
Presentation_ID
Cisco Public
Inter-Chassis Communication Protocol

draft-ietf-pwe3-iccp-06.txt
ICC Protocol Transport Requirements
Reliable Message exchange
In-order Delivery
Sequence Numbers
Timeouts/Retransmissions
use widely deployed LDP protocol.
Extend LDP with a small set of new messages:

RG Connect Message
RG Disconnect Message
RG Notification Message
RG Application Data Message
Use LDP Capability to bootstrap ICCP.
Application layer specific TLVs.
7
Presentation_ID
Cisco Public
Node Failure Detection

Two mechanisms:
RG
BFD Bi-Directional Forward Detection

Detection upon loss of BFD keepalives
Requires nodes to be co-located, with a direct link connection
No split-brain protection, mandates link to be port-channel
dispatched over two different line cards
ICCP over Dedicated Link
/32 IP Route-watch
RG
Detection upon loss of IP routing adjacency

Geo localization of nodes
Split-brain tie-break via MPLS network

Depends on IGP timers
OSPF/ISIS fast convergence tuning is required
Ensure not to have less specific route for detection to work
ICCP over Shared Network
ip routing protocol purge interface

This is requiring to avoid route-watch flapping on peer link
failure
Default in IOS-XR & IOS Rls15
Presentation_ID
Cisco Public
Multi-Chassis Link Aggregation Group

(ASR 9K - MC-LAG)
&
Multi-Chassis Link Aggregation control Protocol
(Cisco 7600 - mLACP)
Background: Link Aggregation Control Protocol

System attributes:
-System MAC address: MAC address that uniquely identifies the switch
-System priority: determines which switchs Port Priority values win
Aggregator (bundle) attributes:

-Aggregator key: identifies a bundle within a switch (per node significance)
-Maximum links per bundle: maximum number of forwarding links in bundle used
for Hot Standby configuration
-Minimum links per bundle: minimum number of forwarding links in bundle, when
threshold is crossed the bundle is disabled
Port attributes:
-Port key: defines which ports can be bundled together (per node significance)
-Port priority: specifies which ports have precedence to join a bundle when the
candidate ports exceed the Maximum Links per Bundle value
-Port number: uniquely identifies a port in the switch (per node significance)
Key: 10
System Priority: 3
System MAC: M2
Key: 35
Presentation_ID
Key: 10
Port #: 1, Priority: 10
Agg3
Agg1
Agg4
Agg2
Port #: 4, Priority:
6
Cisco Public
System Priority: 5
System MAC: M1
Key: 20
11
Extending LACP Across Multi-Chassis: mLACP

mLACP uses ICCP to synchronize LACP configuration & operational state
between PoAs, to provide DHD the perception of being connected to a single
switch
All PoAs use the same System MAC Address & System Priority when
communicating with DHD
-Configurable or automatically synchronized via ICCP
Every PoA in the RG is configured with a unique Node ID (value 0 to 7). Node
ID + 8 forms the most significant nibble of the Port Number
For a given bundle, all links on the same PoA must have the same Port Priority
PoA = Point of Aggregation
DHD = Dual-Homed-Device
Port #: 0x9001, Port Priority 1
PoA1
Node ID: 1
DHD
System MAC:
aaaa.bbbb.cccc
ICCP
System Priority: 1
Node ID: 2
LACP
Presentation_ID
PoA2
Port #:0xA001, Port Priority 2
Cisco Public
12
Operational Variants
PoA-Based Control
Max Links per
Bundle: L
Each PoA is configured to limit the

maximum number of links per bundle
-Limit must be set to L, where L is the minimum
number of links from DHD to any single PoA
Standby PoA
DHD max link should be set > L

-
In order to insure that it is slave of the POA
This will allow faster convergence
L=2
DHD
ICCP
Selection of active/standby links is the

responsibility of the PoAs
Advantages: Faster switchover times

compared to other variants, and Minimum
Link policy on PoA can be flexible
Disadvantage: If ICCP transport is lost,
Split Brain condition could occur
LACP
Active PoA
Max Links per

Bundle: L
This is the most used variant

Presentation_ID
Cisco Public
14
Protected Failure Points

mLACP Offers Protection Against 5 Failure Points:
A: DHD Port Failure
B: DHD Uplink Failure
C: Active PoA Port Failure
D: Active PoA Node Failure

E: Active PoA Isolation from Core Network
Standby PoA
DHD
D
E
Active PoA
Presentation_ID
Cisco Public
15
Failover Operation
Max Links per

Bundle: L (=2)
Min Links per
Bundle: M (=2)
Standby PoA
Port/Link Failures
3. LACP Exchanges
DHD
ICCP
3. Dynamic Port Priority

or Brute-force failover
2. Signal failover over

ICCP
C
Active PoA
Step 1 For port/link failures (A,B,C), active PoA evaluates number of surviving in bundle:
If >= M, then no action
If < M, then trigger failover to standby PoA
Step 2 Active PoA signals failover to standby PoA over ICCP

Step 3 Failover is triggered on DHD by one of:
Dynamic Port Priority Mechanism: real-time change of LACP Port Priority on active PoA to cause the standby
PoA links to gain precedence
Links are either Hot-Standby or Up
Brute-force Mechanism: change the state of the surviving links on active PoA to admin down
Links are either Err-disabled or Up
Step 4 Standby PoA and DHD bring up standby links per regular LACP procedures
Presentation_ID
Cisco Public
16
2. LACP Exchanges
Standby PoA
Failover Operation
Node Failure
DHD
ICCP
1B. Links Down
1A. IP RouteWatch or BFD

timeout
D
Active PoA
Step 1A Standby PoA detects failure of Active PoA via one of:
-IP Route-watch: loss of IP routing adjacency
-BFD: loss of BFD keepalives
Step 1B DHD detects failure of all its uplinks to previously active PoA
Step 2 Both Standby PoA and DHD activate their Standby links per
regular LACP procedures
Presentation_ID
Cisco Public
17
3. LACP Exchanges
Standby PoA
Failover Operation
PoA Isolation from Core
DHD
ICCP
2B. Dynamic Port
Priority or Bruteforce failover
2A. Signal failover

over ICCP
1. Detect core
isolation
E
Active PoA
Step 1 Active PoA detects all designated core interfaces are down
interchassis group 21
backbone interface TenGigabitEthernet4/1
backbone interface TenGigabitEthernet1/4
Really useful if no direct connection between POA or using one only

module toward core
Step 2A Active PoA signals standby PoA over ICCP to trigger failover
Step 2B Active PoA uses either Dynamic Port Priority or Brute-force
Mechanism to signal DHD of failover
Step 3 Standby PoA and DHD bring up standby links per regular
LACP procedures
Presentation_ID
Cisco Public
18
E-LAN Availability Models

Active/Backup Access Node Redundancy (mLACP)
Works similarly with H-VPLS (MPLS Access)
E-LAN availability model
VPLS
Active / Backup Access Node Redundancy (mLACP)

Port / Link Failures
Events
PW - PseudoWire
VFI Virtual Forwarding Instance
VFI
A
FA
2
1A
ICCP
ICCP
LACP
4
LACP
1A
Active PoA detects

failure and signals
failover over ICCP
1B
Failover triggered on
DHD
Standby link brought

up per LACP proc.
Standby PoA flushes

MAC table and
triggers LDP MAC
add. withdrawal to
remote peers
Remote PEs flush

MAC addresses
VFI
For VPLS Decoupled Mode, VFIs PWs always

advertised in Active state, regardless of AC state
Presentation_ID
VFI
FA-C
A
FB
Initial state
VFI
A
FC
1B
Cisco Public
Forwarding EoMPLS PW
Non-Forwarding EoMPLS
20 PW
E-LAN availability model
VPLS

Port / Link Failures (cont.)
Events
VFI
A
End State
FB
Remote PEs flush

MAC addresses
VFI
A
FC
FA
ICCP
ICCP
LACP
VFI
Presentation_ID
LACP
VFI
Cisco Public
21 PW
E-LINE Availability Models

Active/Backup Access Node Redundancy (mLACP)
Pseudowire Redundancy with LDP

Background
MPLS
CE
PW 2
<IP1, VCID1>
PE1
PW 1
<IP2, VCID2>
Designate Pseudowires as either primary or backup

-Primary Pseudowire used for traffic forwarding, and backup
takes over in case of failure
Signaling Redundant Pseudowires in targeted

LDP session
-Cold Redundancy: Backup PWs were not signaled until
required to take over
-Warm Redundancy: Backup PWs were signaled up in the
control-plane but held down in the data-plane. Use AC Fault
code-point in LDP Status Message to indicate a backup PW
-Hot Redundancy: Use PW Preferential Forwarding Status Bits
Presentation_ID
Cisco Public
23
Two-Way Pseudowire Redundancy

Overview
MPLS
CE
PW 1
uPE1
PW 3
PW 2
uPE2
uPE3
PW 4
uPE4
Allows dual-homing of two local PEs to two remote PEs

Four pseudowires: 1 primary & 3 backup provide redundancy
for a dual-homed device
Presentation_ID
Cisco Public
24
Pseudowire Redundancy with LDP

PW Status Signaling
tLDP
CE
PW
PE1
PE2
-0x00000000 - Pseudowire forwarding (clear all failures)
-0x00000001 - Pseudowire Not Forwarding

-0x00000002 - Local Attachment Circuit (ingress) Receive Fault
-0x00000004 - Local Attachment Circuit (egress) Transmit Fault
RFC 4447
-0x00000008 - Local PSN-facing PW (ingress) Receive Fault

-0x00000010 - Local PSN-facing PW (egress) Transmit Fault
-0x00000020 - PW Forwarding Standby
-0x00000040 PW Request to go Active
draft-ietf-pwe3-redundancy-bit
Only this bit is required/used (with help of ICCP)

Presentation_ID
Cisco Public
25

Independent Operation Mode
Every PE decides the local status
of the PW: Active or Standby
MPLS
A PW is selected as primary for

forwarding if it is active on both
local & remote PEs
PW
PE1
A PW is considered as backup if
it is declared as Backup by either
local or remote PE
PE1
PE1
PE1
Presentation_ID
Cisco Public
Active
Active
Active
PE2
Standby
PE2
Standby
Standby
PE2
Standby
Active
PE2
26

Determining Pseudowire State
VPWS / H-VPLS two-way

coupled:
-When AC changes state to Active1,
both PWs will advertise Active
-When AC changes state to Standby1,
both PWs will advertise Standby
Active
Active
Active
Standby
Standby
Standby
pseudowire-class <class name>

encapsulation mpls
status peer topology dual-homed
H-VPLS two-way decoupled:

-Regardless from AC state, Primary
PW and Backup PWs will advertise
Active state
For H-VPLS, all PWs in VFI (at nPE)

are Active simultaneously, for both
access & core PWs
(1) Active / Standby AC states determined for example by mLACP
Presentation_ID
Cisco Public
Active
Active
Active
Standby
Active
Active
pseudowire-class <class name>

encapsulation mpls
status peer topology dual-homed
status decoupled
27
E-LINE availability model
VPWS

Events
I
AS
FC
FB
1A
FA
LACP
ICCP
ICCP
1B
2
4
SA
LACP
Initial state
FA-C
1A
Active PoA detects

failure and signals
failover over ICCP
1B
DHD

up per LACP proc.
Active PoA
advertises Standby
state on its PWs
Standby PoA
advertises Active
state on its PWs
For VPWS Coupled Mode, attachment circuit (AC) state

(Active/Standby) drives PW state advertised to remote peers
Presentation_ID
Cisco Public
28 PW
E-LINE availability model
VPWS

Port / Link Failures (cont.)
Events
I
FC
FB
FA
LACP
ICCP
ICCP
LACP
Presentation_ID
FA-C
1A
Active PoA detects

failure and signals
failover over ICCP
1B
DHD

up per LACP proc.
Active PoA
advertises Standby
state on its PWs
Standby PoA
advertises Active
state on its PWs
End State
Local site access failure does not trigger LACP

failover at remote site (i.e. control-plane
separation between sites)
Cisco Public
Initial state
29 PW
Implementing MC-LAG
Attachment Circuit Using mLACP/MC-LAG

Multi-Chassis LACP synchronization:
LACP BPDUs (01:80:C2:00:00:00) are exchanged on each Link
System Attributes: Priority + bundle MAC Address
Port Attributes: Key + Priority + Number + State
Redundancy
Group
Active POA
DHD
ICCP
redundancy
iccp
group <ig-id>
mlacp node <node id>
mlacp system mac <system mac>
mlacp system priority <sys_prio>
member
neighbor <mpls device>
interface <bundle>
mlacp iccp-group <ig-id>
mlacp port-priority <port prio>
interface <physical interface>
bundle id <bundle id> mode active
Presentation_ID
MPLS
Standby POA
Terminology:
mLACP :
MC-LAG :
DHD :
DHN :
CiscoPOA
Public :
Multi-Chassis Link Aggregation Control Protocol

Multi-Chassis Link Aggregation Group
Dual Homed Device (Customer Edge)
Dual Homed Network (Customer Edge)
Point of Attachment (Provider Edge)
31
Active/Standby Deployment Options
MPLS Core
Active
PE
VFI
VFI
ICCP
MPLS Core
Active
PE
Standby
PE
ASR9Ks as
POAs
VFI
VFI
ICCP
Standby
PE
ASR9Ks as
POAs
Cluster node
as a DHD
Cluster node
as a DHD
LACP Hot-Standby
Traffic may take sub-optimal path

via the DHD inter chassis link
Presentation_ID
Optimal path by bundling 4 links in vPC

Direct path exist between each N7K
and active PE
Cisco Public
32
MC-LAG to VPLS Testing

http://www.cisco.com/en/US/docs/solutions/Enterprise/Data_Center/DCI/vpls/vpls_asr9k.html
2
1
4
5
8
Si
MPLS
core
Si
Only error 2/3/4 are leading to ICCP convergence

Rem: 2 & 4 are dual errors
500 VLAN Unicast: Link error sub-1s & Node error sub-2s
1200 VLAN unicast: Link error sub-2s & Node error sub-4s
Presentation_ID
Cisco Public
33
MC-LAG Not-Supported Service Example

Local Connect
CE1
CE1
CE2
Active POA
ICCP
CE3
Local Connect
Standby POA
Active POA
CE2
ICCP
CE3
Local Bridging
Standby POA
Local connect service between bundle port and bundle port, or between bundle port and regular port are
configurable. And it works in the normal condition
However, if bundle failover happen, then it wont work. For example, for CE2, if it failover to the bottom
POA, then there is no local connection to CE 1 or CE3
In summary, although you can configure the local connect service for the MC-LAG bundle port, but it
could cause problem. So its not supported
Local
bridging is supported
there
L2 link between
Presentation_ID
Cisco Public two POA
2011 Ciscoassume
and/or its affiliates.
All rightsis
reserved.
34
Inter-AS Option A Resiliency Solutions

Using MC-LAG
MC-LAG on
both pair of
ASBRs
ASBR3
ASBR1
VFI
VFI
VFI
AS2
AS1
VFI
VFI
VFI
VPLS or H-VPLS
ASBR2
Link bundle
ASBR4
VPLS or H-VPLS
From each pair of ASBR point of view, the other side is just a single virtual LACP device
Pros:
o Simple configuration which provide both link and node redundancy
o Advantage of option A: Flexible per interface/sub-interface features: QoS, MAC limit, ACL, etc
o Convergence time is fast: sub-second for both link and node failures
Cons:
o Bundle is in active/standby mode, 50% usage of the inter-AS link bandwidth
o Need manual configuration to inter-connect different VPLS VFIs
o limited to be 4K L2 sub-interfaces per bundle port
o ASBR link has to be Ethernet
Presentation_ID
Cisco Public
35
MC-LAG & L3 Service

IRB/BVI, decoupled mode
BD
BVI
Active POA
LACP
ICCP
MPLS/IP
Recommanded Option:
DHD can run IGP with both POA. Routing
session will be up with both POAs
Bundle/POA failover wont cause
the L3 topology change
BD
Standby POA
BVI
Configure L2 sub-interface on the bundle, and then configure L2 PW between two POA. Both L2 subinterface and L2 PW are in the same bridge-domain. Configure IRB/BVI for the bridge-domain for the L3
service
L3 features like HSRP, VRRP, routing, etc could be configured under BVI interface
BVI interfaces are up on both POA regardless of the bundle status
Bundle failover only impact the bundle itself. BVI and related L3 topology is not aware fast L3
convergence
Presentation_ID
Cisco Public
36
mLACP Platform Support
Feature
Cisco
Catalyst
3750-ME
One-Way PW Redundancy
Cisco
7600
Cisco
ASR 9000
Two-Way PW Redundancy
LDP MAC Address

Withdrawal
mLACP
ICCP
LACP
Cisco ME
Cisco ASR
3600X / 3800X 901/903
Presentation_ID
Cisco Public
37
Pseudo mLACP
Pseudo MLACP (P-mLACP) concept
P-mLACP provides an Active / Active dual homing redundancy

mechanism with per-VLAN load balancing
- higher bandwidth utilization
Backup time does not require link state changes

- Improve scale and convergence time
Supports today Dual Homed Device (DHD)

- Will support Dual Homed Network (DHN) in future
New extensions to Inter-Chassis Communication Protocol (ICCP) used

for communication to control the failover process
Available in Cisco 7600 with 15.1(3)S release

- ES / ES+ linecards
Presentation_ID
Cisco Public
39
Pseudo MLACP (P-mLACP) concept (cont.)

P-mLACP provides VLAN based redundancy by allowing one to configure one
primary and one secondary interface pair for each member VLAN
Dual Homed Device (DHD) configured with two separate port-channels
aggregating to one LAG on Point of Attachment (PoAs)
-
Pseudo-LACP application on PoAs ensure that two port-channels on DHD will get bundled to one
logical port-channel on PoA
Allow: VLAN 1-10

Block: VLAN 11-20
DHD
POA1
Allow: VLAN 1-20

Pseudo
mLACP
ICCP
Aggregation / Core
Allow: VLAN 1-20

Block: VLAN 1-10
Allow: VLAN 11-20
Presentation_ID
Cisco Public
POA2
40
Attachment Circuit options

DHD
Mono-Device Dual Homed Device(DHD)

ICCP
DHD
ICCP
DHN
Presentation_ID
Multi-Devices Dual Homed Device (DHD)

The one analyzed in this presentation
Dual Homed Network (DHN)

7600 do support MVRP
But not officially supported as it would
require DHN to support MVRP to flush
local MAC
4K VLAN Scale (no multi-chassis
etherchannel)
Cisco Public
41
Traffic path
DHD
Per VLAN group balancing over Etherchannel

Presentation_ID
Cisco Public
42
Traffic path
DHD
With VPLS, on the ES+ facing edge card

EVC load balancing over lacp bundle is only on per EFP
With one EFP per VLAN, repartition is OK
If one EFP for all VLAN, then no balancing
Presentation_ID
Cisco Public
43
EFP static load repartition over bundle
interface TenGigabitEthernet2/3
channel-group 2 mode active link 1
Optional, but useful to control

which link is primary
interface Port-channel2
port-channel load-balance link 1
service-instance 4021
!
service instance 4021 ethernet
description ### VDC-2 VLANs ###
EFP
encapsulation dot1q 400,1000-1999

bridge-domain 4021
Presentation_ID
Cisco Public
44
Active path failures
DHD
N
C
A
B
ICCP
A = Nexus port failure (shut)

B = Attachment-Circuit failure
C = Dual AC failure
Presentation_ID
D = 7600 crash
E = MPLS link failure
Cisco Public
N= DHD member crash
45
Failure condition and MAC flush consideration

MAC learning on nominal mode
Mr
Ml
Mr
Ml
Mr
Ml
ICCP
Mr
Mr
LAG
PW
Forwarding
Blocking
Presentation_ID
Cisco Public
Assuming no VLAN repartition

46

MAC learning on PE failure
Mr
Ml
Mr
Mr
Ml
Ml
Mr
ICCP
Mr
Ml
Mr
Mr
Ml
LAG
Mr
PW
Forwarding
Blocking
Presentation_ID
Cisco Public
Assuming no VLAN repartition

47

MAC learning on PE up time
Due to lack of per VLAN MVRP, second path VLAN may be affected during recovery
Manual trigger is prefered
Mr
Ml
Mr
Mr
Ml
Ml
Mr
ICCP
Mr
Ml
Mr
Mr
Ml
LAG
PW
Forwarding
No MVRP
Requires to flap port-channel,
Blocking
Presentation_ID
Mr
Cisco Public
Reversion
trigger
Manual
procedure
48
Reversion procedure
scripted for easiness
event manager applet Revert_po2
event none
action 1.0 cli command "en"
action 1.1 cli command "conf t"

action 2.0 cli command "int po2"
action 2.1 cli command "shut"
action 2.2 cli command "no shut"
action 2.3 cli command "end"
action 3.0 syslog msg "Preemption requested for port-channel 2
alias exec Revert_po2 event manager run Revert_po2
Presentation_ID
Cisco Public
49
Active path failures measurement

Medium scale = 1000 VLAN per vPC *2
Failure type
Traffic
Down
Up
Traffic
Down
Up
DHD link
shut
0.5s-1s
1.5s
<0.1s
1.8s
7600 link
shut
0.8s
1.8s
<0.1s
1.8s
7600
crash
0.8s
1.7s
0.8s
1.7s
MPLS link
down
0.1s
<0.1s
0.1s
<0.1s
DHD crash
0.3s
<0.1s
0.3s
<0.1s
Comment
HW and SW considerations
Supported today by Cisco 7600 with following HW:
- Access linecard: ES+ and ES-20 (Core linecard: Any)
- SUP-720 and RSP-720
Supported on port-channel with service instances (evc-based port-channel)

- EFP with bridge-domain and SVI for VPLS uplinks
- EFP with xconnect (PoA with EoMPLS uplinks VPWS and H-VPLS spoke)
- EFP with C-MAC bridge-domain (PoA with native 802.1ah Ethernet uplinks)
PBB configurations support only one (1) member link per PoA in mLACP/P-mLACP port-channels
- EFP with C-MAC bridge-domain with SVI for VPLS uplinks
Supported by Cisco 7600 starting with IOS 15.1(3)S
Planned support with ASR9K

Presentation_ID
Cisco Public
51
P-mLACP Configuration Guidelines / Restrictions

Only one Port-channel of type mLACP or pmLACP is supported on a single
RG. There can be one mLACP port-channel and one more pmLACP portchannel on a single RG, but not two Port-channel of same type.
Bridge-domain associated to P-mLACP port-channel should not be part of any
other non P-mLACP interfaces
Fo a given Bridge-domain (SVI)
ICCP
ICCP
Working in normal mode
Presentation_ID
Blocking in failure mode !

Not supported
Cisco Public
52
P-mLACP Configuration Guidelines / Restrictions

DHCP / IGMP snooping state synchronization over ICCP is supported from 15.2S
Routed PW will not be supported in Core
Connectivity Fault Management (CFM) maintenance points (MEP/MIP) not
supported on P-mLACP port-channel
- CFM Port MEP (untagged CFM) can be configured under port-channel member
Link OAM can be configured under port-channel members

Recommended configuration sequence:
- Configure interchassis group
- Configure port-channel interface
- Configure port-channel memberlinks
Any Out-Of-Sequence Configuration may need flapping of the Port-Channel to ensure

proper pmLACP functionality
Presentation_ID
Cisco Public
53
SSO considerations
Pseudo-mLACP is supporting SSO

- With LDP graceful
Software crash protected gracefully
But manual command redundancy force-switchover

- Is subject to timing constraints and could lead to
medium term blackout
- Under investigation
- Recommendation is to preventively swap traffic to
alternate node
Presentation_ID
Cisco Public
54
Sample Configuration
PE1 Configuration
PE2 Configuration
redundancy
redundancy
monitor peer bfd
monitor peer bfd
member ip 2.2.2.2
member ip 1.1.1.1
backbone interface GigabitEthernet2/21
backbone interface GigabitEthernet1/12
mlacp node-id 1
mlacp node-id 2
Prim: 4041
Sec: 4042
PE-1 (POA1)
1.1.1.1
Gig2/20
Gig2/21
mLACP-AA
ICCP
Redundancy
Group 100
Gig1/12
DHD
Gig1/11
Prim: 4042
Sec: 4041
Presentation_ID
PE-2 (POA2)
2.2.2.2
Cisco Public
55
Sample Configuration (cont.)

PE1 Configuration
PE2 Configuration
no ip address
no ip address
mlacp interchassis group 100
mlacp interchassis group 100
mlacp mode active-active
mlacp mode active-active
mlacp load-balance primary vlan 4041
mlacp load-balance primary vlan 4042
mlacp load-balance secondary vlan 4042
mlacp load-balance secondary vlan 4041
ethernet vlan color-block all*
ethernet vlan color-block all*
description vpls_coupled
description vpls_coupled
encapsulation dot1q 4041
rewrite ingress tag pop 1 symmetric
bridge-domain 4041
bridge-domain 4041
description vpls_decoupled
description vpls_decoupled
bridge-domain 4042
bridge-domain 4042
(*) Configuration
required if EVC cross
connect or PBB is used on the P-mLACP Cisco
port-channel
Presentation_ID
Public
56
Debugging/Troubleshooting
Show commands
show lacp multi-chassis group

show lacp multi-chassis port-channel
show lacp multi-chassis load-balance group
show lacp multi-chassis load-balance port-channel <x>

show lacp internal
show lacp neighbor
show etherchannel [id] [detail | port | port-channel | summary]

show redundancy interchassis
show mpls ldp iccp
Presentation_ID
Cisco Public
57
Show Commands
Checking ICCP connectivity
PE1#show redundancy interchassis
Redundancy Group 100 (0x64)
Applications connected: mLACP, Pseudo-mLACP, DHCP Snooping
Monitor mode: BFD
member ip: 2.2.2.2 "PE2", CONNECTED
BFD neighbor: GigabitEthernet2/9, next hop 192.168.41.2, UP
mLACP state: CONNECTED
Pseudo-mLACP state: CONNECTED
ICRM fast-failure detection neighbor table

IP Address
Status Type Next-hop IP
Interface
==========
====== ==== ===========
=========
2.2.2.2
UP
GigabitEthernet2/9
Presentation_ID
BFD
192.168.41.2
Cisco Public
58
Show Commands
Checking P-mLACP state
PE1#show lacp multi-chassis load-balance port-channel 34
Interface Port-Channel 34
Local Configuration:
P-mLACP Enabled:
Yes
Redundancy Group:
100
Revertive Mode:
Revertive
Primary VLANs:
4041
Secondary VLANs:
4042
Local Interface State:
Interface ID: 34
Port State:
Up
Primary VLAN State:
Active
Secondary VLAN State: Standby
Peer Interface State:
Interface ID: 34
Primary VLAN State:
Active
Secondary VLAN State: Standby
Presentation_ID
Fail flags (bitfield):

MLACP_FAIL_PEER
= 0x1,
MLACP_FAIL_MIN_LINK = 0x2,
Cisco
Public
MLACP_FAIL_CORE
= 0x4,
/* Node failure */
/* A/B/C failure */
/* Core failure */ 59
Debugging/Troubleshooting
SP Debugging commands
-
debug lacp load-balance

debug lacp load-balance vlan
debug lacp load-balance database
debug lacp load-balance redundancy-group
debug lacp load-balance all
debug redundancy interchassis [all | application | error | event | monitor]
debug ethernet service instance error
debug ethernet service instance xdr
debug ethernet service instance event
debug mvrp all
debug mvrp error
debug mvrp packet
debug mvrp verbose
RP Debugging commands
-
debug lacp load-balance vlan

debug lacp load-balance database
debug redundancy interchassis [all | application | error | event | monitor]
debug mpls ldp iccp
debug ethernet service error
debug ethernet service api
Presentation_ID
Cisco Public
60
Service Multi-Homing
Building blocks
IP/Service Edge
IP/MPLS
L2 Ethernet Ring
(MST/REP-AG,
G.8032)
Ethernet spokeand-hub
(MC-LAG)
L3 Router dualhoming (L3 ECMP)
Presentation_ID
Point to point L2 Service PW redundancy

Multi-point L2 Service MC-LAG
Cisco Public
61
Changing the Redundancy Paradigm

Virtual Cluster and Satellite for BNG
ASR 9000
Virtual
Cluster
Home
Cluster
ASR 9000
Distribution
Geo Redundant Dual Homing
High Availability
Single-Chassis-like look&feel and

Management of cluster Members
Stateful Failover between chassis
Active/active LAG user facing
BNG
Aggregation
(MPLS)
Core
BNG
ASR 9000
Virtual Cluster
Cluster & Satellite

ASR 9000
Distribution
Satellite
Home
Huge 1GE Fan-out
Satellites appear like ASR 9000

Linecards
Simplified topology, No Spanning

tree or other L2 redundancy
protocols needed
BNG
Aggregation
(MPLS)
Core
BNG
Presentation_ID
Cisco Public
62
Complete Your Online Session Evaluation
Give us your feedback and

you could win fabulous prizes. Winners
announced daily
Receive 20 Cisco Preferred Access points
for each session evaluation you complete
Complete your session evaluation online
now (open a browser through our wireless
network to access our portal) or visit one
of the Internet stations throughout the
Convention Center
Dont forget to activate your Cisco Live
and Networkers Virtual account for access
to all session materials, communities, and
on-demand and live activities throughout
the year. Activate your account at any internet station
or visit www.ciscolivevirtual.com.
Presentation_ID
Cisco Public
63
Thank you.
Presentation_ID
Cisco Public
64

Advance VPLS Technology Atachements Options

Hochgeladen von

Dokumentinformationen

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Advance VPLS Technology Atachements Options

Hochgeladen von

Copyright:

Verfügbare Formate

Advanced VPLS attachment technology options

- Fault isolation between sites

Position Architecture: STP domain isolation thru dual homing

2011 Cisco and/or its affiliates. All rights reserved.

Access Multi-homing Solutions Summary

REP /REP access

Sub 200msec convergence

Standard based solution as long as access network support MST/PVST

VSS (Catalyst / Cisco 7600-Sup2T)

2011 Cisco and/or its affiliates. All rights reserved.

L2 node peering solutions

Hub and Spoke

mLACP + 2-way PW Red.

Hub and Spoke

mLACP + 2-way PW Red.

Hub and Spoke

mLACP + 2-way PW Red.

2011 Cisco and/or its affiliates. All rights reserved.

Inter-Chassis Communication Protocol for L2VPN PE

Inter-Chassis Communication Protocol

ICCP allows two or more devices to form a

ICCP over Dedicated Link

-ICCP rides on targeted LDP session, but

Various redundancy applications can

Under standardization in IETF:

2011 Cisco and/or its affiliates. All rights reserved.

Inter-Chassis Communication Protocol

Extend LDP with a small set of new messages:

2011 Cisco and/or its affiliates. All rights reserved.

Node Failure Detection

BFD Bi-Directional Forward Detection

ICCP over Dedicated Link

Detection upon loss of IP routing adjacency

Split-brain tie-break via MPLS network

ip routing protocol purge interface

2011 Cisco and/or its affiliates. All rights reserved.

Multi-Chassis Link Aggregation Group

Background: Link Aggregation Control Protocol

Aggregator (bundle) attributes:

Extending LACP Across Multi-Chassis: mLACP

2011 Cisco and/or its affiliates. All rights reserved.

Each PoA is configured to limit the

DHD max link should be set > L

In order to insure that it is slave of the POA

This will allow faster convergence

Selection of active/standby links is the

Advantages: Faster switchover times

Max Links per

This is the most used variant

2011 Cisco and/or its affiliates. All rights reserved.

Protected Failure Points

D: Active PoA Node Failure

2011 Cisco and/or its affiliates. All rights reserved.

Max Links per

3. Dynamic Port Priority

2. Signal failover over

Step 2 Active PoA signals failover to standby PoA over ICCP

2011 Cisco and/or its affiliates. All rights reserved.

1B. Links Down

1A. IP RouteWatch or BFD

2011 Cisco and/or its affiliates. All rights reserved.

2A. Signal failover

Really useful if no direct connection between POA or using one only

2011 Cisco and/or its affiliates. All rights reserved.