MPLS cisco QOS VPN full 03 mpls te toi

-Information Distribution -Path Calculation -Path Setup -Forwarding Traffic Down A Tunnel... Path Calculation• Constrained SPF – find shortest path to a specific node • Consider more t

Trang 1

MPLS TE TOI

eosborne@cisco.com

Trang 2

• How MPLS TE works

• What Code Is MPLS TE In?

• Platform Issues in Implementation

• Lab Demo - config

Trang 3

How MPLS TE Works

• Prerequisites

• How MPLS-TE Works

• Basic Configuration

• Knobs! Knobs! Knobs!

• Deploying and Designing

Trang 4

You should already understand…

push/pop/swap, EXP, and LFIB

works

Trang 5

• Deploying and Desiginig

Trang 6

How MPLS-TE Works

-What good is MPLS-TE?

-Information Distribution -Path Calculation

-Path Setup -Forwarding Traffic Down A Tunnel

Trang 7

What Good Is MPLS-TE?

• There are two kinds of networks

1 Those that have plenty of bandwidth

Trang 8

What Good Is MPLS-TE?

• MPLS-TE introduces a 3 rd kind:

1 Those that have plenty of bandwidth everywhere

2 Those with congestion in some places, but not in others

3 Those that use all of their bandwidth to its

maximum efficiency, regardless of path routing!

Trang 9

shortest-What Good Is MPLS-TE?

Substitute which is Totally

Effortless

This stuff takes work, but it’s worth it!!!

What is MPLS-TE? What is it not?

Trang 10

Information Distribution

• You need a link-state protocol as your IGP

IS-IS or OSPF

• Link-state requirement is only for MPLS-TE!

Not a requirement for VPNs, etc!

Trang 11

Need for a Link-State Protocol

• Why do I need a link-state protocol?

1 To make sure info gets flooded

2 To build a picture of the entire network

Trang 12

Need for a Link-State Protocol

Consider the following network:

- All links have a cost of 10

- RtrA’s path to RtrE is A->B->E, cost 20

- All traffic from A to {E,F,G} goes A->B->E

Trang 13

What a DV Protocol Sees

Node Next-Hop Cost

Trang 14

What a LS Protocol Sees

• RtrA sees all links

• RtrA only computes

the shortest path

• Routing table doesn’t change

Trang 15

The Problem With Shortest-Path

Trang 16

What MPLS-TE Addrs

• RtrA sees all links

• RtrA computes paths on

properties other than just shortest cost

• No congestion!

Trang 17

-Information Distribution

-Path Calculation -Path Setup

-Forwarding Traffic Down A Tunnel

Trang 18

• OSPF

-Uses Type 10 (Opaque Area-Local) LSAs -See draft-katz-yeung-ospf-traffic

Trang 19

• IS-IS

-Uses Type 22 TLVs -See draft-ietf-isis-traffic

Trang 20

• IS-IS and OSPF propagate the same information!

-Link identification -TE Metric

-Bandwidth info (max physical, max reservable,

available per-class)

-Attribute flags

Trang 21

• TE flooding is local to a single {area|level}

• Inter-{area|level} TE harder, but possible (think PNNI)

Trang 22

-Information Distribution

-Path Calculation

-Path Setup -Forwarding Traffic Down A Tunnel

Trang 24

Path Calculation

• Normal SPF – find shortest path

across all links

Trang 25

Trang 26

Trang 27

Trang 28

Trang 29

Trang 30

Trang 31

Trang 32

Trang 33

• Constrained SPF –

find shortest path

to a specific node

• Consider more than

just link cost!

Trang 34

just link cost!

Trang 35

just link cost!

Trang 36

just link cost!

Trang 37

just link cost!

Trang 38

just link cost!

Trang 39

just link cost!

Trang 40

just link cost!

Trang 41

just link cost!

Trang 42

just link cost!

Trang 43

just link cost!

Trang 44

• “But Wait! There’s nothing different between the two SPF results!”

• ….but….

Trang 45

• What about the 2nd

path?

• Available bandwidth has changed!

Trang 46

path?

Trang 47

path?

Trang 48

path?

Trang 49

path?

Trang 50

path?

Trang 51

path?

Trang 52

path?

Trang 53

path?

Trang 54

path?

Trang 55

path?

Trang 57

• Happy! Happy!

• Joy! Joy!

Trang 58

• What if there’s more than one path that

meets the minimum requirements (BW, etc)?

• PCALC algorithm:

1 find all paths with the lowest IGP cost

2 then pick the path with the highest minimum bandwidth

along the path

3 then pick the path with the lowest hop count (not IGP

cost, just hop count)

4 then just pick one path at random

Trang 61

4!

Trang 62

{8,90M}

Trang 64

-Path Setup

Trang 65

Path Setup

• Cisco MPLS-TE uses RSVP

• RFC2205, plus draft-ietf-mpls-rsvp-lsp-tunnel

Trang 66

Path Setup

• Once the path is calculated, it is handed to RSVP

• RSVP uses PATH and RESV messages to request an LSP along the calculated path

Trang 67

Path Setup

• PATH message: “Can I have 40Mb along this path?”

• RESV message: “Yes, and here’s the label to use.”

• LFIB is set up along each hop

= RESV messages

Trang 68

Path Setup

• Errors along the way will trigger RSVP errors

• May also trigger re-flooding of TE info if appropriate

Trang 69

-Path Setup

Trang 70

Forwarding Traffic Down a Tunnel

• There are three ways traffic can

be forwarded down a TE tunnel

-Autoroute -Static routes -Policy routing

• For the first two, MPLS-TE gets

you unequal-cost load-balancing.

Trang 71

• Autoroute = “use the tunnel as a directly connected link for SPF purposes”

• This is not the CSPF (for path determination), but the

regular IGP SPF (route determination)

• Behavior is intuitive, operation can be confusing

Trang 73

This is RtrA’s logical topology

Other routers don’t see the tunnel!

RtrI

Trang 74

RtrI

Trang 76

RtrG is not routed to over

the tunnel, even though it’s the tunnel tail!

Tunnel1

Trang 77

Unequal-Cost Load Balancing

• IP routing has equal-cost load-balancing, but not

unequal-cost*

• Unequal-cost load balancing difficult to do while

guaranteeing a loop-free topology

*EIGRP has ‘variance’, but that’s not as flexible, and besides,

MPLS-TE and EIGRP are two different things

Trang 78

Unequal-Cost Load Balancing

• Since MPLS doesn’t forward based on IP header, permanent routing loops don’t happen.

• 16 hash buckets for next-hop, shared in rough proportion

to tunnel BW

Trang 79

gsr1#show ip route 192.168.1.8

Routing entry for 192.168.1.8/32

Known via "isis", distance 115, metric 83, type level-2

Redistributing via isis

Last update from 192.168.1.8 on Tunnel0, 00:00:21 ago

Routing Descriptor Blocks:

* 192.168.1.8, from 192.168.1.8, via Tunnel0

Route metric is 83, traffic share count is 2

192.168.1.8, from 192.168.1.8, via Tunnel1

Trang 80

Note that the load distribution is 11:5 – very close to

2:1, but not quite!

gsr1#sh ip cef 192.168.1.8 int

………

Load distribution: 0 1 0 1 0 1 0 1 0 1 0 0 0 0 0 0 (refcount 1)

Hash OK Interface Address Packets Tags imposed

1 Y Tunnel0 point2point 0 {23}

………

Trang 81

Q:How does 100:10:1 fit into a 16-deep bucket?

1MB

gsr1#sh ip rou 192.168.1.8

Routing entry for 192.168.1.8/32

Known via "isis", distance 115, metric 83, type level-2

Redistributing via isis

Last update from 192.168.1.8 on Tunnel2, 00:00:08 ago

Routing Descriptor Blocks:

* 192.168.1.8, from 192.168.1.8, via Tunnel0

192.168.1.8, from 192.168.1.8, via Tunnel1

192.168.1.8, from 192.168.1.8, via Tunnel2

Trang 82

A:Any way it wants to! 15:1, 14:2, 13:2:1, it depends on the order the tunnels come up.

Deployment guideline: don’t use tunnel metrics that don’t reduce to

16 buckets!

1MB

gsr1#sh ip cef 192.168.1.8 internal

………

Load distribution: 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 (refcount 1)

Hash OK Interface Address Packets Tags imposed

………

Trang 83

RtrA(config-if)#ip policy route-map set-tunnel

RtrA(config)#route-map set-tunnel

RtrA(config-route-map)#match ip address 101

RtrA(config-route-map)#set interface Tunnel1

Trang 85

Forwarding Traffic Down a Tunnel

• You can use any combination of

autoroute, static routes, or PBR.

• …but simple is better unless you have a good reason.

• Recommendation: either autoroute

or statics to BGP next-hops,

depending on your needs.

Trang 86

Trang 87

Basic Configuration

-Basic Midpoint/Tail Config

-Basic Headend Config -Path-option

-Bandwidth

Trang 88

Basic Midpoint/Tail Config

(globally)

ip cef {distributed}

mpls traffic-eng tunnels

Trang 89

(per interface) mpls traffic-eng tunnels

Trang 90

(if IGP == OSPF) router ospf <x>

mpls traffic-eng router-id Loopback0 mpls traffic-eng area <y>

Trang 91

(if IGP == OSPF)

• MPLS TE is a single area only (usually area 0)

• RID must be set (unlike OSPF RID)

It’s a Very Very Good idea to make it a /32

loopback.

Trang 92

(if IGP == IS-IS)

router isis <x>

mpls traffic-eng router-id

Loopback0 mpls traffic-eng level-{1,2}

metric-style wide

Trang 93

(if IGP == IS-IS)

• MPLS TE is a single level only

• RID must be set (unlike OSPF RID)

It’s a Very Very Good idea to make it a /32

loopback.

Trang 94

‘metric-style wide’ - ???

• IS-IS must have wide metrics enabled

• This is discussed in more detail later in this presentation; see also www.cisco.com.

Trang 95

• Total config tally so far:

1 line globally

1 line per interface

2 lines if OSPF

3 lines if IS-IS

Trang 96

Basic Headend Config

• Headend needs the 4-5 ‘mid/tail’ lines

• But wait – there’s more!

Trang 97

• Create the tunnel interface

interface Tunnel0

ip unnumbered Loopback0 tunnel mode mpls traffic-eng tunnel source Loopback0

tunnel destination <tunnel endpoint>

tunnel mpls traffic-eng autoroute tunnel mpls traffic-eng path-option 10 dynamic

unnumbered to Loop0

path-option tells the tunnel how to get to tail

’10’ is the priority of the path-option there are other options besides dynamic autoroute is not strictly necessary, but is useful

Trang 98

• Total config tally:

Trang 99

Trang 100

Knobs! Knobs! Knobs!

• Influencing the Path Selection

• Auto-Bandwidth

• Fast Reroute

• DiffServ-Aware Traffic Engineering

Trang 101

Bandwidth Priority

Administrative Weight Attributes & Affinity

Trang 102

ip rsvp bandwidth <x> <y>

• Per-physical-interface command

• X = amount of reservable BW, in K

• Y = not used by MPLS-TE

• default: X==Y==75% of link

bandwidth

Trang 104

• New tunnel with better setup priority will force teardown of already-established tunnel with worse holding priority

• Configuring S<H is illegal

• Default is S=7,H=7

Trang 105

= 40MB tunnel with S=7, H=7

= 40MB tunnel with S=6, H=6

Trang 107

“Should I ever set S != H?”

No Not unless you know you have a good reason to.

Trang 108

• this can be used as a per-tunnel

delay-sensitive metric for doing VoIP TE

Trang 109

Administrative Weight

tunnel mpls traffic-eng

path-selection metric {te|igp}

administrative-weight to determine shortest cost

Trang 110

Delay-Sensitve Metric with Adminastrative Weight

tunnel mpls traffic-eng path-selection

metric {te|igp}

mpls traffic-eng administrative-weight <x>

• configure admin weight == interface

delay

• configure VoIP tunnels to use TE

metric to calculate the path

• delay-sensitive metric!

Trang 111

Attributes & Affinity

• Link attribute – 32 separate link properties

• Tunnel affinity – desire for links to have certain properties set

• Invent your own property meanings

Trang 112

mpls traffic-eng

attribute-flags <0x0-0xFFFFFFFF>

• Per-physical-interface command

Trang 113

care about bits 2 and 8; bit 2 must

be set, bit 8 must be 0’

Trang 114

• Q: How should I use admin-weight?

• A: To exclude some links from consideration by some

tunnels

• …so give a satellite link an attribute of 0x2, and any VoIP tunnels can be configured with ‘affinity 0x0 mask 0x2’

Trang 115

• Auto-Bandwidth

• Fast Reroute

• DiffServ-Aware Traffic Engineering

Trang 116

tunnel mpls traffic-eng auto-bw ?

collect-bw Just collect Bandwidth info on this tunnel

frequency Frequency to change tunnel BW

max-bw Set the Maximum Bandwidth for auto-bw on this tunnel

min-bw Set the Minimum Bandwidth for auto-bw on this tunnel

<cr>

reservation based on traffic out

tunnel

more or less sensitive

Trang 117

tunnel mpls traffic-eng auto-bw ?

collect-bw Just collect Bandwidth info on this tunnel

frequency Frequency to change tunnel BW

max-bw Set the Maximum Bandwidth for auto-bw on this tunnel

min-bw Set the Minimum Bandwidth for auto-bw on this tunnel

<cr>

reservation based on traffic out tunnel

or less sensitive

tradeoff: quicker reaction vs more churn

Trang 118

auto-bw: (86400/86259) 0 Bandwidth Requested: 100

• 86400 = reoptimization time (default 24h)

tunnel mpls traffic-eng auto-bw frequency <x>

• 86259 = time left to reoptimization

• 0 = BW measured at end of last reopt

interval

• bw requested = signalled tunnel BW

tunnel mpls traffic-eng {max-bw|min-bw} <bw>

Định dạng
Số trang	184
Dung lượng	1,35 MB