6PE Primer: Enabling IPv6 in Your Network
This is the last post in the series on Primers. The technology has mostly been geared towards SP networks, but as with all things networking - everything apply’s to everything. As matter of fact, I believe we can all do good to know both sides to the coin. Enjoy.
MPLS Layer 3 VPNS, a Primer
It’s time for another primer. This week we take a look at MPLS Layer 3 VPNs. Enjoy!
The Plight of An Engineer
Do you ever find yourself doing all the right things for the wrong reasons? If you’re anything like me, you do this more often than you would like to admit. It is far more common for me to get carried away with wrong thinking than it is to have pure motives. And it’s because of that, that I have some things I need to get off my chest.
There is a question that I constantly, almost daily, have to ask myself, “Why do I start with the best intentions and always come up short when it comes to motivations.”
You see, I love teaching, nurturing, and watching people around me grow in their technical ability. I get some kind of natural high when I am able to help someone out. I think it is the greatest thing that there are people on twitter who have never met me, but are willing to help me with some technical problem. It is equally as great that people spend their time writing blogs (or podcasts) that they put out for free for my consumption.
Seriously, think about that for a moment; what an amazing field we have chosen to work in. In fact, I find it so invigorating that I started this blog and try to help people as much as I can. There is something inside of me that wants to give back to the community that has helped me so much as I have grown and continue to grow as an engineer.
Here’s the problem, I find myself equally, if not more excited when I get retweeted, or one of my blog posts hits it “big”. You know, when you find yourself hitting F5 over and over on your Google analytics page to see how many people aren’t reading your blog. Or you’re having a conversation with someone; suddenly a *great* tweet pops in your head; you tweet it and wait anxiously for the retweets to begin. This becomes the validation for your awesomeoness. You know how great you are, but for some reason you need others to validate how great you are.
OK, maybe it’s just me. I am ashamed to admit it but I constantly fall into this trap. I find myself wrapped up in my own mind wanting more and more attention from my virtual peeps. Somehow I feel as though my identity, my self-worth is wrapped up in how *popular* I am online. That my networking skillz (yes, skillz with a Z) can only be validated by how much ataboys and “Wow you’re a genius” comments I get on a daily basis. If I’m not getting it in person, well then heck I have to get it online, after all there is WAY more people to behold my glory online.
Here’s the thing, I don’t think I am alone in this, while it may not be as big of a struggle for some; it is definitely an issue I have seen time and time again in the field of IT.
Where does it come from?
I believe it stems from the fact that we are paid by how much we know. So we must strive and struggle to prove our worth. It’s almost as if the further into our career we go, the more pride and egotistical our behavior must become. If we aren’t the best and brightest around then why would a company want us. We constantly feel the pressure that we must earn our kept day in and day out. And forgot it if we don’t know something, that is just death as we know it.
Is this reality?
Do we really need to think like this? No, I think engineers put this pressure on themselves – while not entirely, but for the most part it is our fault.
What can we do?
Honestly, I wish I had the answer to that. I have tried time and time again to let it go. My family does not judge me based on my OSPF skills or how well I can explain the traffic flow internally inside a Cat 6500. And at the end of the day it doesn’t really matter.
So why am I publicly admitting to my own egotistical behavior? I have no idea – really I don’t. Maybe it is just my narcissism, unabashedly releasing itself into the wild. Or maybe I am just looking for some empathy from some fellow engineers who struggle the same. Or maybe yet, it is a my way of holding me accountable to change. At any rate, I cannot stop trying to help and give back to the community, even if that means my motives aren’t always as pure as I would like them to be.
If one out of every 100 times I strike it rich and my motives are pure, it is worth the 99 times that they aren’t. If I can help in any small way, if I can give back just an ounce what has been given to me, it is worth it no matter what.
So here is my advice, just have fun. We get to work with some really cool technology so let’s just have fun and forgot about being the best and the brightest. If you have fun and just keep learning, I believe it is only natural that you will become a great engineer, one that isn’t so paranoid and annoying.
Let’s just keep putting material out there, let’s keep helping as best as we know how and maybe, just maybe our egos well take a side step and we can get on with it.
Well, I feel better. Thanks for listening.
CJ
MPLS Primer for the interested
Here is a fast facts MPLS primer. MPLS is the most Zen protocol; it just doesn’t care what it carries in the payload. It has this unquie ability that allows MPLS to be used for a variety of things. Lets dig in and take a look.
Multiprotocol Label Switching (MPLS)
MPLS is an open standard (RFC 3031) that was designed to use labels to make forwarding decisions. One of the great features of MPLS is that it can transport many different types of payloads, e.g., Ethernet, ATM, PPP, IPv4, IPv6. It is an efficient protocol because the routers only need to do one IP lookup (when the packet first enters the MPLS domain) until it reaches the last hop router before the packet is forwarded to the host.
Use Cases:
This list is not exhaustive but includes the major use cases for MPLS.
1. Transparent tunneling over Service Provider Networks which allows for:
a. A BGP free core
b. Layer 2/Layer 3 VPN service to Customers
2. Traffic Engineering
a. Distribute load over underutilized links
b. Give bandwidth guarantees
c. Detect and repair failures quickly
While MPLS is/was primarily used in Service Provider environments, you can find use cases for the Enterprise. It is worth looking into.
MPLS Devices
MPLS device terminology is pretty straight forward. There are three primary devices that exist within an MPLS network.
· Customer Edge (CE) Device – the last hop device in the Customer’s network. This device is usually not MPLS aware, i.e., it runs only IP forwarding and not MPLS forwarding.
· Provider Edge (PE) Device – the last hop device in the Provider’s network, connects to the P devices and is sometimes referred to as the Label Edge Router (LER). This device will run IP and MPLS.
· Provider (P) Device – the core device at the center of the Providers network. It is sometimes referred to as a Label Switch Router (LSR). This device only runs MPLS, i.e., all switching occurs based on MPLS labels.
Forwarding Equivalency Class (FEC)
A FEC is defined as a group of IP packets which are forwarded over the same path (to the same next-hop), and with the same treatment.
MPLS Label (Header)
The MPLS header is 4-bytes in total comprising of:
· 20 bit label = locally significant to router
· 3 bit EXP = Class of Service
· S bit = defines last label in the label stack
· 8 bit TTL = Time to Live
The label is used to identify a FEC. The label is imposed (inserted) between the Layer 2 header and the Layer 3 header. The top label appears first in the packet where the bottom label is last.
Label Operation
Routers will perform various label operations depending on if the packet is incoming or outgoing.
· Label Push – adds a label to incoming packet; label imposition
· Label Swap – replaces the label on an incoming packet
· Label Pop – removes the label on an outgoing packet; label disposition
There is another technique referred to as Penultimate Hop Popping (PHP). Without PHP, the egress PE router must perform two lookups. First it has lookup the destination prefix associated with the label, and then once it has the prefix; the router must look up the next-hop for the destination prefix. With PHP, the next-to-last hop router (the penultimate router) will perform the label pop operation. This means the egress PE router will not have to do the label lookup because it is receiving an IP packet (the penultimate router has popped the label off before forwarding it to the egress router).
MPLS uses the implicit-null label to inform the router to perform PHP. The egress router will advertise the implicit-null label for all directly connected prefixes.
Label Distribution
There are two protocols used for distributing labels in the network: Tagged Distribution Protocol (TDP) and Label Distribution Protocol (LDP).
TDP
· Cisco proprietary protocol that use TCP port 711 to advertise labels.
LDP
· Standard protocol
· Uses UDP multicast 224.0.0.2:646 to discover neighbors
· Uses TCP port 646 once neighbors are discovered
· Advertises labels for IGP learned routes
Each router running MPLS will distribute the labels they are using for each prefix to its neighbors; think EIGRP not OSPF. Each router only knows about its neighbor’s labels and does not have a full view of the network.
MPLS switching logic is very straight forward; if traffic comes in interface-1 with label X, then send it out interface-2 with label Y. LDP is preferred and more widely used, so we will focus our efforts on LDP.
Configuration:
Router(config)#ip cef
Router(config)#mpls label protocol ldp
MPLS Data Structures
Like all protocols, MPLS has a few data structures to store its information. Before we get into the data structures there is one important fact about MPLS, you need to have CEF running. Without CEF enabled, the router will not enable MPLS forwarding. The MPLS adjacency will still come up, but you will not have MPLS forwarding functioning.
The first data structure is the Label Information Base (LIB). The LIB is a database used by MPLS to store labels learned from other LSRs and also labels assigned by the local router.
The second data structure is the Label Forwarding Information Base (LFIB). The LFIB is used by MPLS for switching labeled packets – think about it like CEFs FIB but for MPLS.
MPLS LIB:
Router#sho mpls ldp bindings
tib entry: 10.0.0.0/30, rev 14
local binding: tag: 19
remote binding: tsr: 172.0.0.4:0, tag: 20
remote binding: tsr: 172.0.0.2:0, tag: imp-null
tib entry: 10.0.0.4/30, rev 4
local binding: tag: imp-null
remote binding: tsr: 172.0.0.4:0, tag: 19
remote binding: tsr: 172.0.0.2:0, tag: imp-null
MPLS LFIB:
Router#sho mpls forwarding-table
Local Outgoing Prefix Bytes tag Outgoing Next Hop
tag tag or VC or Tunnel Id switched interface
16 Pop tag 172.0.0.2/32 0 Fa0/1 10.0.0.5
17 Pop tag 172.0.0.4/32 0 Fa0/0 10.0.0.10
18 16 172.0.0.1/32 0 Fa0/1 10.0.0.5
19 Pop tag 10.0.0.0/30 0 Fa0/1 10.0.0.5
“Special” Labels
There are a few labels that have special meanings, one of which is the implicit-null label that we discussed earlier. There is another label – rather a lack of label – that you might see in the LFIB. In the LFIB, you may see no label listed under outgoing label. This indicates that the next hop router and the local router have not formed a MPLS adjacency. So the local router will send pure IP packets to the next hop router.
So where pop label tells the router to pop the top label off, no label instructs the router to remove all labels and send a pure IP packet to the next-hop. Be aware of no label, if you are expecting MPLS to be running between the neighboring routers and you see no label, you have a problem.
Full Example Configuration
The basic configuration for MPLS is very simple. First you have to enable CEF and LDP/TDP. Next, you enable the interfaces that you want to run MPLS on; this would be the interfaces connecting all your routers together. One thing to keep in mind, you don’t need to enable MPLS on interfaces like you would OSPF or EIGRP, i.e. to enable those networks to be distributed within the protocol. You are only enabling MPLS on interfaces to enable hellos to go out and form adjacencies. MPLS will by default, take all known prefixes learned from the configured IGP and append labels to them.
Configuration Template
Router(config)#ip cef
Router(config)#mpls label protocol ldp
Router(config)#int f0/0
Router(config-if)#mpls ip
Here are a few useful commands that I have listed below.
· show mpls ldp bindings – displays LIB
· show mpls forwarding-table – displays LFIB
· show mpls ldp neighbors – displays LDP adjacencies
· show mpls interfaces – displays all interfaces that MPLS is enabled on
Well that’s what we have, I hope you were able to pay attention long enough; it will be well worth it. Stay tuned for MPLS VPNs and 6PE - that’s when the magic happens. Now go on, have some fun, get your MPLS on!
Is there anything I missed or needs to be corrected? Let me know.
More Information
http://www.cisco.com/en/US/docs/ios/12_0st/12_0st21/feature/guide/fs_rtr.html#wp1067202
http://www.cisco.com/en/US/tech/tk436/tk428/technologies_q_and_a_item09186a00800949e5.shtml
http://www.amazon.com/MPLS-VPN-Architectures-Ivan-Pepelnjak/dp/1587050021/ref=sr_1_4?ie=UTF8&qid=1328021275&sr=8-4
http://www.amazon.com/MPLS-Fundamentals-Luc-Ghein/dp/1587051974/ref=sr_1_1?ie=UTF8&qid=1328021275&sr=8-1
ACL: 17 Steps to Correcting a Misconfiguration
Read with caution. The procedure below is highly tuned and known to cause huge pay raises and new big important titles - use with care.
