Wednesday, June 30, 2010

IPv6 Tutorial


Till now I have published various IPv6 tutorials on MPLS cloud. Those IPv6 tutorials tell about how to connect IPv6 sites overs MPLS cloud. Given is the list of the IPv6 MPLS Tutorials:-
1. IPv6 From MPLS
2. IPv6 Addressing Architecture For Service Providers
3. IPv6 Internet In VRF Over MPLS Cloud
4. IPv6 Internet in VRF Over IPv4 MPLS Cloud
5. Troubleshooting Command Difference Between VPNv4 and VPNv6

But I have not posted any good material on the basics of IPv6. I am planning to post some good tutorials of IPv6 subnetting. But in this post I am adding one of the IPv6 videos which clears all about the IPv6 w.r.t design and implementation.




Click Here To Read Rest Of The Post...

Sunday, June 27, 2010

RSVP Tunnels are up but all traffic forwarding is stopped


I found one weird issue during my testing lab where in end to end RSVP-TE was configured and all the customer traffic forwarding was done on those tunnels. But suddenly I found that the traffic started drop but the tunnels were up. Then after deep analysis of FIB and LFIB, I found that the one of the node was advertising label for RSVP TE but the next node was not able to receive that label. After refreshing the labels, everything started work.
Now the question comes, if some of the node start misbehaving with label, in that case my path and resv messages should get dropped or not. I am not sure about it.
But the given post helped me lot to find the RCA.
Missing labels can forward labels somewhere else
How to troubleshoot MPLS VPN?
When and where PHP happens?

Click Here To Read Rest Of The Post...

Friday, June 25, 2010

Is LDP required for VPNv4 Labels?



In case of layer 3 VPN, two labels are normally carried by packet. But the differentiation between the labels is ipv4 and vpnv4. Ipv4 label is used for IGP and vpnv4 label is used for customer route. Normally a question comes in mind, “Is LDP responsible for both the labels”. The answer is no because LDP is only responsible for the top most label is IGP label and MP-iBGP is responsible for vpnv4 label which is present under beneath of IGP label. Even if the core network is not running LDP, but MP-iBGP is enabled from PE to PE, we can easily see the vpnv4 labels exchange. But the problem is that traffic forwarding will not happen because the core network doesn’t understand the labels.
How BGP does forward the vpnv4 label?
In the given logs, we can see that from 100.100.100.1 to 100.100.100.3 MP-iBGP tunnels are created. In the update message, under Border Gateway Protocol, we have update and path attribute messages. Path attribute is again having six different BGP attributes:-
  1. Origin
  2. AS_Path
  3. MED
  4. Local Preference
  5. Extended Communities
  6. MP Reach NLRI (Network Layer Reach ability Information)

In the depicted image, Carried Extended Communities are the route target values which we define in vrf. In this we have defined 65500:1 as route target value.



Another attribute is MP_REACH_NLRI which plays vital role in MP-iBGP. In this first value is Address Family: IPv4(1); 1 is reserved for IPv4 only. For more see RFC 1700. According to RFC 2858 Address Family Identifier is defined as 
         This field carries the identity of the Network Layer protocol
         associated with the Network Address that follows. Presently
         defined values for this field are specified in RFC 1700 (see
         the Address Family Numbers section).”


According to RFC 2858 SAFI is defined as “Subsequent Address Family Identifier:
 
         This field provides additional information about the type of
         the Network Layer Reachability Information carried in the
         attribute.”
 
In the given figure SAFI value is 128 because from 128 to 255 values are not being reserved by IANA. These are for private use. It means, if some vendor wants IOT with CISCO MP-iBGP then they might need to use the same SAFI (Not Sure About It).According to draft draft-ietf-l3vpn-rfc2547bis-03 section 4.3.4 SAFI 128 is used for labeled VPN IPv4 addresses.
 
 
Next field is next hop network address which is carrying the information of next hop address i.e. router id of advertising router and CISCO is using route distinguisher RD 0:0 for IGP routes. So it means by default global routing is also a part of a vrf which could be called global vrf(not sure).
  
Last field is Network Layer Reachability Information which is actually carrying VPNv4 label 19 also known as bottom label, route distinguisher and ipv4 prefix of vrf.   



Click Here To Read Rest Of The Post...

Thursday, June 24, 2010

How Does Label Distribution Protocol(LDP) Initializes?



Highest loopback id starts the Label Distribution Protocol initialization process by sending common session parameter TLV which includes a sub TLV of parameters containing session protocol version, session keepalive time, advertisement method, loop detection and session path vector. In the given diagram, I have started MPLS LDP firstly on 100.100.100.1 (R1) and 100.100.100.2(R2). So in this case R2 send a first initialization message to R1 by adding all the above TLV parameters.



Structure of LDP initialization message sent from R2 to R1.

Figure 1 (Click on image to see in new window)




Being R2 is having highest LDP router id, so it start sending the label information to R1. We can say R2 is working as downstream LSR to R1. R2 sending Address message and label mapping message which are sub part of label distribution protocol.

Figure 2(Click on image to see in new window)

Address message is only containing the directly connected interface ip address of R2 which are 1.1.1.2, 2.2.2.2 and 100.100.100.2.


Figure 3(Click on image to see in new window)





Along with this R2 is also sending label mapping message which is containing the information about route, label and address family. Address family means whether it is ipv4 route or vpnv4 route. R2 is sending a information about 1.1.1.0 prefix by including label 3 which is used for PHP (Penultimate Hop Popping).

Figure 4(Click on image to see in new window)

The same way R2 is sending a label 16 to R1 for 100.100.100.1 which is loopback address of R1 itself and label 16 becomes as local label in R2 forwarding table.

Figure 5(Click on image to see in new window)

Figure 6(Click on image to see in new window)

Once this is done then R1 initiates a label mapping process by sending the label and FEC information to R2 ( R2 is upstream LSR to R1).

After this, R3 is started which is having loopback address of 100.100.100.3. In this case R3 will start the initialization process of LDP being R3 is having highest LDP router id.

Figure 7(Click on image to see in new window)

Now R3 becomes downstream LSR to R2 and R2 becomes upstream LSR to R3. Once R3 will exchange all the label mappings then R2 will initiate the process and send the information to R3.

Now how the updates of new route will flood. To test this, a new route 100.100.100.100 is installed in R3. After this we saw a label mapping message was sent from 100.100.100.2 to 100.100.100.1 and vice versa but the communication was being started by 100.100.100.2

Figure 8(Click on image to see in new window)

How does LDP sends new updates to its peer?
LDP uses label mapping message to send the partial label updates to its peer.
Label mapping message is containing the information of new route 100.100.100.100 with label 18.
Figure 9(Click on image to see in new window)



Click Here To Read Rest Of The Post...

Technorati Claim


HMJ9589TJHS5

Click Here To Read Rest Of The Post...

Wednesday, June 23, 2010

What is downstream and upstream router in MPLS?



Downstream router is the one which is responsible to advertise the label first to upstream router in case of downstream on demand method is selected.
Upstream router is the one which advertise the labels to its downstream router after receiving label bindings from it.

Click Here To Read Rest Of The Post...

Tuesday, June 22, 2010

Requirement of Synchronous Ethernet aka syncE


From the last two decades Ethernet is becoming the dominant technology in service provider as well as in Telecom domain. Being it’s asynchronous behavior, providers faced lot of issues to integrate it with the existing TDM technologies. TDM mainly works on clock which means synchronization is required among all the nodes.

Now service providers are moving towards the Next Generation Network aka NGN or Packet Converge Network aka PCN. The challenge which they are facing is the clock synchronization between the TDM networks over Ethernet network. By nature Ethernet doesn’t provide any clock mechanism but thanks to synchronous Ethernet aka syncE which helps to make communication possible between TDM to TDM cloud over packet core network.
In synchronous Ethernet frequency is transferred over physical level than using any upper layers. Every syncE network equipment contains an internal clock called Ethernet Equipment Clock aka EEC. By doing this master slave interconnected tree is built and clock is passed from one node to another.  One primary reference clock (PRC) is used which normally serves the clock to the equipments.

One more category of clock is used which is also known as Synchronous Supply Unit (SSU). SSU are mainly required for given three reasons:-
  1. Reduce the jitter in clock.
  2. Act as a node clock.
  3. Serve the clock in case of failure of PRC.

According to ITU-T Rec. 803 standard, maximum depth of EEC could go up to 60 EEC nodes. The hierarchy is shown in depicted picture:-




Click Here To Read Rest Of The Post...

Friday, June 11, 2010

BWA Spectrum Auction


The highest bid of Rs 2,293 was achieved by Mumbai and the second highest of Rs 2,241 crore, by Delhi. Qualcom and Infotel won the licence for both Mumbai and Delhi. Bharti and Qualcom won the bids for four broadband circles.


BWA auction had two slots of 20 MHz of pan-India spectrum. BSNL and MTNL will have to pay the equivalent of the winning bid in each service area. BWA spectrum provides the high-speed internet access as well as Internet telephony and TV services. It also enables voice and high-speed data services.

Around 11 companies including Bharti Airtel, Reliance, Idea Cellular, Aircel, Vodafone and Tata Communications Internet Services, participated in the auction for Broadband Wireless Access spectrum.

Here is the beginning of my post.
Source:-One India



Click Here To Read Rest Of The Post...

Sunday, June 6, 2010

Types of PseudoWire


Pseudowire emulation aka PWE3 that emulates the attributes of service over packet switched network (PSN). Pseudo means no physical existence only virtual. By using pseudowire, service provider can emulate any circuit end to end. E.g. if customer is looking for TDM bandwidth end to end, but SP is having a packet core network but no TDM backhaul, in that case pseudowire help SP to deliver end to end circuit which uses packet core network and provide TDM drop to customers. This is the case where in both termination points are having same output but in case of different output like one side Ethernet and another side frame-relay or atm, the best is to provision inter network circuit.
Types of Pseudowire
1. CESoPSN:- Circuit Emulation over Packer Switched Network supports framed and channelized TDM services over packet switched network.
2. SAToP:- Structure Agnostic TDM over Packet, is a TDM Pseudowire technology which treats the TDM traffic as data traffic and ignore the framing bits. It supports unframed TDM services.

Advantages of SAToP:-
1. Flexible packet size.
2. Lowest end to end delay.
3. Low overhead.

Advantages of CESoPSN:-
1. Lower packetization delay.


Click Here To Read Rest Of The Post...