This chapter deals with a description of the MPLS traffic engineering technology behavior on two heterogeneous, but nowadays the most commonly used network vendors are Cisco and Huawei. Compatibility and functionality between network devices Huawei and Cisco were verified by testing the appropriate network topology. In this topology, we mainly focused on the useful feature of MPLS TE called Fast Reroute (FRR) protection. It provides link protection, node protection and also bandwidth protection during the failure of the primary link, especially on backbone networks. After successful validation, compatibility and functionality of the network topology between the heterogeneous routers using the Fast Reroute protection will be possible to use this MPLS TE application in the real networks.
- Fast Reroute
In the 1990s, asynchronous transfer mode (ATM) was considered an ideal solution in transmission networks to operate with different demands . In earlier times, this technology provided traffic engineering by a virtual channel as well as Frame-Relay. But subsequently IP began to replace the ATM technology, which became the most popular network protocol for transmission. On the other hand, the ATM was still widely used by telecommunication providers at that time. Since 1999, the draft of multiprotocol label switching (MPLS) has become the IETF  standard and internet service providers started to use this concept for IP/MPLS transmission over older ATM technology. In this chapter, we focus on the application of MPLS called MPLS traffic engineering. MPLS TE can be understood as “effective planning utilization” . Instead of the normal routing of IP packets, MPLS TE routes traffic according to the source IP addresses. This application can choose the most appropriate links according to the speed of individual lines, delay, delay variability and can also react automatically to the change of these parameters [3, 4]. In addition, the applications of MPLS are also used for an effective creation of separate virtual private networks among the company branches, or for addressing QoS issues in communication networks, such as satellite and mobile cellular networks. This chapter is focused on the most used function of MPLS TE called Fast Reroute. Fast Reroute can be used in the case of a link or node failure in the MPLS network. Both vendors Huawei and Cisco support MPLS TE, but each vendor can use a different function model. The main motivation of this chapter is to bring complex view on usage and cooperation between routers of two different vendors using Fast Reroute protection.
2. State of the art
Multiprotocol label switching (MPLS) is a backbone technology, which uses labels attached to the packets for their transmission. Packets are not transmitted based on the destination IP addresses but according to the MPLS labels. The protocol allows most packets to be forwarded at Layer 2 (switching) rather than at Layer 3 (routing). The term “multiprotocol” means that it can transport various protocols on Layer 3 such as IPv4, IPv6, IPX, and protocols of Layer 2, e.g., Ethernet, HDLC, Frame-Relay, or ATM .
As shown in Figure 1, source A sends a packet to the router CE1. CE1 handles the packet according to its routing table in a standard way. According to the destination IP address of each packet, the ingress router (PE1) inserts a label in front of the IP header at the edge of the backbone network. All the subsequent routers ignore the IP headers and perform the packet forwarding based on the labels in front of them. This MPLS label determines a path that is used for the routing of a particular packet. Paths through MPLS network are called LSPs [5, 7].
Each label has its local importance and every MPLS backbone router processes the packet based on the MPLS label. Finally, the egress router (PE2) removes the label and forwards the original IP packet toward its final destination.
Nowadays, practically, computer networks are not built only on a homogeneous infrastructure, but they use heterogeneous devices.
As depicted in Figure 2, the basic MPLS topology consists of two Huawei routers—the first one AR3200 and the second one AR2200 (marked in the red frame) and two Cisco 2800 series routers. The first goal was to verify MPLS functionality and interoperability among these above-mentioned routers.
Huawei routers have only two CLI modes (basic view and the system view). The basic configuration of Huawei routers is as follows:
[PE1-ospf-1-area-0.0.0.0]network 220.127.116.11 0.0.0.3
[PE1-ospf-1-area-0.0.0.0]network 10.0.0.1 0.0.0.0
[P1]mpls lsr-id 10.0.0.2
[PE1-LoopBack0]ip address 10.0.0.1 255.255.255.255
[PE1-GigabitEthernet0/0/1]ip address 18.104.22.168 255.255.255.252
[PE1-GigabitEthernet0/0/0]ip address 192.168.10.1 255.255.255.0
All routers use OSPF as a routing protocol. Unlike Cisco routers, LSR identification must be configured on every Huawei router. For identification of Huawei routers, the loopback IP addresses were applied. The command lsp-trigger all allocates label for each IP prefix in the routing table. Then the LDP protocol for exchange of MPLS labels had to be activated for each MPLS physical interface.
3.1. Configuration of MPLS TE on Huawei routers
The network topology of the MPLS TE network is depicted in Figure 3.
First of all, it is necessary to configure MPLS TE technology and then turn on signalling protocol RSVP-TE. In the case of a link or node failure, we configure mpls rsvp-te hello as well. It is also necessary to enable modified SPF algorithm called CSPF which excludes. Using CSPF algorithm, the ingress MPLS router do not use these lines, which not satisfying the requirements of the data flow.
It is also necessary to explicitly turn on RSVP-TE for each MPLS physical interface. The part of the configuration is setting of the maximum bit rate of a line, which can be reserved. This bit rate cannot exceed the bit rate of a physical interface. The command mpls te bandwidth bc0 10000 defines maximum total bandwidth for class type 0. In this case, the maximum bit rate of a physical interface is used.
In order to LSR routers could exchange information about set parameters such as maximum bit rate of the line, it is necessary to configure support for a special type of message OSPF LSA 10 for the OSPF area. Then this type of message is used for CSPF algorithm. By the command opaque-capability enable, we allow propagation of LSA 10 messages. Next command enable traffic-adjustment advertise includes static LSP tunnels into SPF calculation and to the routing table.
[PE1-mpls]mpls rsvp-te hello
[PE1-mpls]mpls te cspf
[PE1-GigabitEthernet0/0/1]mpls rsvp-te hello
[PE1-GigabitEthernet0/0/1]mpls te bandwidth max-reservable bandwidth 10000
[PE1-GigabitEthernet0/0/1]mpls te bandwidth bc0 10000
[PE1 ospf-1-area-0.0.0.0]mpls-te enable
[PE1-ospf-1]enable traffic-adjustment advertise
3.2. Configuration of MPLS TE on Cisco routers
To enable the MPLS TE technology on Cisco routers, it is necessary to configure mpls traffic-eng tunnels and ip rsvp signaling hello commands. To achieve establishment of the LDP signaling protocol from the loopback interface, mpls ldp router-id Loopback0 force is configured.
A part of the next configuration is to explicitly turn on RSVP-TE for each MPLS physical interface and set the maximum bit rate of a line which can be reserved. Using mpls traffic-eng area 0 command is configured a special type of message OSPF LSA 10 for the OSPF area. Each Cisco router must be uniquely identified using OSPF router-ID. If the router did not have this identification, OSPF LSA 10 will not be transmitted.
PE2(config)#mpls ldp router-id Loopback0 force
PE2(config)#mpls traffic-eng tunnels
PE2(config)#ip rsvp signalling hello
PE2(config-if)#mpls traffic-eng tunnels
PE2(config-if)#ip rsvp bandwidth 10000
PE2(config-if)#ip rsvp signalling hello
PE2(config)#router ospf 1
PE2(config-router)#mpls traffic-eng router-id Loopback0
PE2(config-router)#mpls traffic-eng area 0
3.3. Configuration of primary explicit path on Cisco router PE1
To define an explicit path for the primary line through the MPLS network via routers PE1-P1-P3-P2-PE2, each next hop is defined by the IP address of the LSR router.
MPLS TE technology includes configuration of MPLS tunnel connections. As a tunnel source, a loopback interface is defined by IP address unnumbered interface LoopBack0 command. Last next hop IP address of an explicit path must match the destination of the tunnel. In our case, 10.0.0.4 is used. Identification of the MPLS tunnel is made by mpls te tunnel-id 1 command. Priority is set by the command mpls te priority 0, where zero indicates the highest priority. A part of the configuration must be mpls te record-route label which records the links during the initiation of the tunnel.
[PE1-explicit-path-PE1-P1-P3-P2-PE2]next hop 22.214.171.124
[PE1-explicit-path-PE1-P1-P3-P2-PE2]next hop 126.96.36.199
[PE1-explicit-path-PE1-P1-P3-P2-PE2]next hop 188.8.131.52
[PE1-explicit-path-PE1-P1-P3-P2-PE2]next hop 184.108.40.206
[PE1-explicit-path-PE1-P1-P3-P2-PE2]next hop 10.0.0.4
[PE1-Tunnel0/0/0]tunnel-protocol mpls te
[PE1-Tunnel0/0/0]ip address unnumbered interface LoopBack0
[PE1-Tunnel0/0/0]mpls te tunnel-id 1
[PE1-Tunnel0/0/0]mpls te record-route label
[PE1-Tunnel0/0/0]mpls te priority 0
Bit rate 128 kbit/s is assigned to the MPLS tunnel. It is necessary to define an explicit path. During the failover of the primary line, command mpls te fast-reroute bandwidth guarantees switching to the backup line with a keeping bit rate of the primary line. By mplsteigpshortcut command, tunnel becomes a virtual tunnel line, which will be inserted into the IP routing table. To ensure the tunnel connection precedence over the traditional calculation by OSPF routing protocol, we define an absolute metric for this tunnel using mpls te igp metric absolute 1command.
[PE1-Tunnel0/0/0]mpls te bandwidth ct0 128
[PE1-Tunnel0/0/0]mpls te path explicit-path PE1-P1-P3-P2-PE2
[PE1-Tunnel0/0/0]mpls te fast-reroute bandwidth
[PE1-Tunnel0/0/0]mpls te igp shortcut
[PE1-Tunnel0/0/0]mpls te igp metric absolute 1
[PE1-Tunnel0/0/0]mpls te commit
3.4. Configuration of primary explicit path on Huawei router PE2
Because every explicit path is unidirectional, we need to configure MPLS tunnel in the opposite direction via routers PE2-P2-P3-P1-PE1.
Similarly, primary MPLS tunnel is configured on the Cisco router. As a tunnel source, a loopback interface is used. Because the last next hop of the explicit path is IP address 10.0.0.1, this address is defined as a destination address. By tunnel mpls traffic-eng autoroute announce command, Cisco router announces the presence of the MPLS tunnel to the IP routing table. The highest priority is set by mpls te priority 0 command.
PE2(config)#ip explicit-path name PE2-P2-P3-P1-PE1
PE2(config-if)#ip unnumbered Loopback0
PE2(config-if)#tunnel destination 10.0.0.1
PE2(config-if)#tunnel mode mpls traffic-eng
PE2(config-if)#tunnel mpls traffic-eng autoroute announce
PE2(config-if)#tunnel mpls traffic-eng priority 0 0
Same bit rate (128 kbit/s) is assigned to MPLS tunnel. The explicit path is then included in the MPLS tunnel interface. During failover of the primary line on the Cisco site, command tunnel mpls traffic-eng fast-reroute bw-protect guarantees switching to the backup line with a keeping bit rate of the primary line.
PE2(config-if)#tunnel mpls traffic-eng bandwidth 128
PE2(config-if)#tunnel mpls traffic-eng path-option 1 explicit PE2-P2-P3-P1-PE1
PE2(config-if)#tunnel mpls traffic-eng record-route
PE2(config-if)#tunnel mpls traffic-eng fast-reroute bw-protect
3.5. Configuration of backup path and Fast Reroute on Huawei router P1
In next step, the router P1 is configured. A backup explicit path is defined between routers P1 and P2. This backup tunnel will be used when the primary path PE1-P1-P3-P2-PE2 fails. The function of the backup exit node has a destination router P2 which uses 10.0.0.3 address. This tunnel interface becomes a backup link using mpls te bypass-tunnel command. Last command protects the interface GigabitEthernet0/0/2, in the case of failure of the router P3 or link between P1 and P3.
[P1-explicit-path-P1-P2]next hop 220.127.116.11
[P1-explicit-path-P1-P2]next hop 10.0.0.3
[P1-Tunnel0/0/0]ip address unnumbered interface LoopBack0
[P1-Tunnel0/0/0]tunnel-protocol mpls te
[P1-Tunnel0/0/0]mpls te tunnel-id 1
[P1-Tunnel0/0/0]mpls te record-route
[P1-Tunnel0/0/0]mpls te priority 0
[P1-Tunnel0/0/0]mpls te path explicit-path P1-P2
[P1-Tunnel0/0/0]mpls te bypass-tunnel
[P1-Tunnel0/0/0]mpls te igp shortcut
[P1-Tunnel0/0/0]mpls te igp metric absolute 1
[P1-Tunnel0/0/0]mpls te protected-interface GigabitEthernet0/0/2
[P1-Tunnel0/0/0]mpls te commit
4.1. Verification of MPLS TE technology
After the configuration, it is time to verify the correct functionality of the MPLS TE technology. Figure 4 shows the LFIB table with MPLS labels and also a created primary MPLS TE tunnel. The entry point of the tunnel is the PE1 router with IP address 10.0.0.1, which corresponds to the configured IP address on the loopback interface. The exit point is therouterPE2, which is identified by IPaddress10.0.0.4. Likewise, we can see establishment of the primarytunnelPE2_t0 to theIPaddress10.0.0.4. Each one-way tunnel route has its own identification (LSPID) and assigned MPLS label.
As it can be seen in Figure 5, the transmission rate of 128kbit/s is reserved throughout the LSP routersPE1-P1-P3-P2-PE2. The same transmission rate is reserved for the tunnellinePE2-P2-P3-P1-PE1 as well.
The records marked “T” in the LFIB table of the routerPE2indicated that packets are sent through MPLSTE tunnel. As we can see in Figure 6, Huawei router remembers only the IP address of the end of the MPLS tunnel. However, a Cisco router in the LFIB table also records the subnet 192.168.10.0/24.
Figure 7 shows the established primary tunnels, which pass through the router P1, but also established backup tunnels. The entry point of the backup tunnel is the IP address10.0.0.2andthe exit point is the router P2, which is identified by the IPaddress10.0.0.3. Likewise, we see the backup tunnel(P2_t0), which was defined on the router P2.
4.1.1. Verification of MPLS TE Fast Reroute
Explicitly configured path through MPLS tunnel was verified using trace route command fromPC1toPC2 via PE1-P1-P3-P2-PE2 routers, as depicted in Figure 8.
An Ethernet link between routers P1 and P3 was disconnected. Every single second was sent an ICMP message from PC1 to PC2. Because 5ICMP messages were lost, the reconvergence time of Fast Reroute was just 5seconds, which can be seen in Figure 9.
As depicted in Figure 10, the primary tunnel line used inner MPLS label 28, there is still maintained as the inner label. Value "zero" is used as the outer MPLS label. This explicit NULL label signals to the receiving router P2 to remove the outer MPLS label.
Fast Reroute was also tested on the Cisco site which was subsequently disconnected by means of a serial link between routers P2 and P3. Every single second was sent an ICMP message from PC2 to PC1. Because only two ICMP messages were lost, the convergence time of Fast Reroute was just 2 seconds, which can be seen in Figure 11.
As depicted in Figure 12, the reconvergence time of the OSPF protocol was also measured without the function of Fast Reroute. The measured time was 15 seconds.
The goal of this chapter was to test a network scenario of interoperability between different vendor’s network devices for MPLS TE technology using the Fast Reroute function. Our goal was to verify the compatibility and functionality between the Cisco and Huawei devices. Although MPLS technology is standardized by RFC, some of our practical experience showed us problems in interoperability between different vendors within various RFC standardized technologies. The basic MPLS configuration was without any problems. The appropriate IP prefixes were successfully exchanged. LIB and LFIB tables were filled up. The major disadvantage of Huawei routers during the MPLS TE configuration is necessity to have the appropriate license. After the license activation, the MPLS TE technology worked properly and the primary and backup MPLS tunnels were established. Without using the technology MPLS TE, the OSPF reconvergence lasted about 15 seconds, after disconnecting Ethernet cable. Due to function Fast Reroute of MPLS TE, the reconvergence lasted only 5 seconds between routers P1 and P3, which is 1/3of convergence time within the OSPF protocol. When using Fast Reroute, the convergence lasted only 2 seconds after disconnecting serial link between routers P2 and P3. It is 1/8 of convergence time within the OSPF protocol. If more routers were added to the network topology, it would lead to a longer convergence time of OSPF but the reconvergence time within Fast Reroute would remain unchanged.
Because nowadays the fast convergence is very critical, this chapter showed that the ISPs can use these heterogeneous network routers together with Fast Reroute technology, which can greatly reduce the convergence time.
This publication was created within the project support of VŠB-TUO activities with China with financial support from the Moravian-Silesian Region and partially was supported by the grant SGS reg. no. SP2016/170 conducted at VSB-Technical University of Ostrava, Czech Republic.