Официален блог на WebEKM EKM очаквайте сайта онлайн скоро.

Download Free Templates http://bigtheme.net/ free full Wordpress, Joomla, Mgento - premium themes.

MPLS Traffic Engineering fundamentals with example of use

September 1, 2018adminIT LABS, Networking & Security

MPLS works based on underlying IGP protocol. It may be OSPF or IS-IS, but whatever we choose, we are still dependant to IGP path calculation. Of course we may manipulate the cost of path and this way provide traffic enginering, but MPLS has its own mechanism.

Traffic engineering means traffic shaping and policing. In case of MPLS we may choose the exact path, we may also allocate required bandwith to the appropriate flow and set up priority. For allocation of the bandwith responsible is RSVP-TE protocol. This flow may be any critical service like VoIP that is highly important for us.

MPLS traffic engineering employs “constraint-based routing,” in which the path for a traffic flow is the shortest path that meets the resource requirements (constraints) of the traffic flow. In MPLS traffic engineering, the flow has bandwidth requirements, media requirements, a priority versus other flows. Protocol responsible for path calculation is called CSPF (Constrained-Based Shortest Path). We may also use explicitly configured path in 2 modes Strict (the flow has to pass through ALL routers) and Loose (the flow has to pass through the pointed routers and the others).

MPLS traffic engineering uses well known logical tunnels, the concept also used in GRE. Tunnels are uni-directional, what means the return traffic may goes back through different tunnel or even not through any tunnel, but just MPLS.

Let’s take the other lab from this topic http://itbundle.net/archives/4496, modify its slightly and based on its, deploy traffic engineering.

Here is the “empty” lab ready to implementation of MPLS-TE with MEGA decryption key:

https://mega.nz/#!u6hBQBIY

!Ltb0R_jH7ReLitECeuAQRXCWOIfr6WMxP8AhgKFeZQw

and rar file decryption key : www.itbundle.net

As you see I added another router R1 in MPLS core. From now on the traffic from Customer-A1 to Customer-A2 will be going through R1 becasue OSPF has calculated properly that this is the low cost path.

Let’s check if the flow is really going through the R1 router. I used traceroute between Customer-A routers. I pinged 6.6.6.6 with source 1.1.1.1 and conversly

MPLS Traffic Engineering

Now, thanks to traffic engineering we will enforce the traffic goes through R3 and R4

We will implement MPLS-TE in 3 steps.
1. Enabling MPLS-TE within OSPF
2. Configuration MPLS-TE on interfaces
3. Configuration of the tunnel

1. Enable OSPF (Area0) to support MPLS-TE on EACH router in MPLS core. Router ID has to be loopback of given router.

router ospf 1
mpls traffic-eng router-id lo0
mpls traffic-eng area 0

2. Configuration of MPLS-TE on EACH router and the reservable bandwith if required on the routers interfaces

mpls traffic-eng tunnels

(example on R2 router, do the same on the other routers respectively)
interface fa0/0
mpls traffic-eng tunnels
ip rsvp bandwidth 50000

interface fa1/0
mpls traffic-eng tunnels
ip rsvp bandwidth 50000

3. Configuration of the “tunnel interface” on both PE1 and PE2 (from Headend R2 to Tailend R5)

Firstly we enforce the traffic to go through the desirable path R2-R3-R4-R5 from Headend-R2 and respectively R2-R3-R4-R5 from Tailend-R5. To that end we have to create the list with the next hops.We will use this “list” later under tunnel configuration.

R2
ip explicit-path name R2-R3-R4-R5
next-address 3.3.3.3
next-address 4.4.4.4
next-address 5.5.5.5

R5
ip explicit-path name R5-R4-R3-R2
next-address 4.4.4.4
next-address 3.3.3.3
next-address 2.2.2.2

Now we create the tunnel. We may “borrow” the ip address from loopback interface. “Tunnel destination” is always loopback address of the second site. Next we run traffic engineering on the “interface tunnel” and allocate bandwith. The last 2 lines are the most interesting.”tunnel mpls traffic-eng autoroute announce” states that we want our tunnel takes part in IGP OSPF Shortest Path First calculation process, because logical tunnel is considered as peer to peer link (despite leads through 2 routers), will win over other paths that are going through the at least one router (R1). In the last line we set up “path options”. We have only the one option, but we could have more, then we add another line “tunnel mpls traffic-eng path-option 2” and so on. In our case we use “explicit” path calculation (the other option is “dynamic” when we lat MPLS-TE calculation of the best path) and we point out the list with next hop addresses.

interface tunnel 25
ip unnumbered lo0
tunnel destination 5.5.5.5
tunnel mode mpls traffic-eng
tunnel mpls traffic-eng bandwidth 3000
tunnel mpls traffic-eng autoroute announce
tunnel mpls traffic-eng path-option 1 explicit name R2-R3-R4-R5

interface tunnel 52
ip unnumbered lo0
tunnel destination 2.2.2.2
tunnel mode mpls traffic-eng
tunnel mpls traffic-eng bandwidth 3000
tunnel mpls traffic-eng autoroute announce
tunnel mpls traffic-eng path-option 1 explicit name R5-R4-R3-R2

How important is command “autoroute announce” we may check if we use its and when remove from configuration. We see that without this command the next hop for route 5.5.5.5 is 10.0.21.1, with we have p2p tunnel.