Posts Tagged ‘NXOS’

Never thought I would be writing about how to utilize IPv6 in 2017 because of all the excellent material on the Internet; however, I have discovered a few things:

  1. There are still technologies which have horrible support for IPv6 (including new stuff)
  2. There are people still resistant to implementing it
  3. There is material on the Internet which shows up early in Google searches which references deprecated standards

Without any further delay, I am going to outline a few items you should keep in mind when deploying your IPv6 network:

Subnet mask size

In IPv6, barring a few exceptions like point-to-point links, you should always utilize a /64 for each deployed subnet. Why? Well, if you wanted to use DHCPv6 you’ll find Microsoft’s implementation won’t even allow you to change from a /64 and even a DHCPv6 server in Linux, while it will actually run with a mask larger than a /64, it will only hand out a /64. Also, you’ll find the use of anything larger than a /64 breaks a lot of the auto-discovery mechanisms in the switch/router, namely around EUI-64, and just doesn’t make sense.

What subnet size should I get from ISP/provider/administrator?

If you’re not going to “own” your IPv6 network, that is you’re not getting an assignment with an ASN to advertise, you’re either looking to obtain a public block of addresses for use and/or you’re internal and need your networking administrators to assign you a prefix which you can further subnet yourself. There is a standard most follow to assign prefixes to “customers”.

An ISP, for instance, may have numerous /32’s (or maybe a bit larger) assigned to them for their use to distribute to customers. Lets call them ISP and you work for “company” and you’re an internal IT organization within “company” who uses “ISP”. Your company would request from the ISP an IPv6 block assignment. From one of the ISP’s /32’s you’ll get, lets say, a /48 just for the hell of it. This is how your company can break it down internally for assignment:

  • 65,536 =  /64’s
  • 32,768 = /63’s
  • 16,384 = /62’s
  • 8192 = /61’s
  • 4096 = /60’s
  • 2048 = /59’s
  • 1024 = /58’s
  • 512 = /57’s
  • 256 = /56’s
  • 128 = /55’s
  • 64 = /54’s
  • 32 = /53’s
  • 16 = /52’s
  • 8 = /51’s
  • 4 = /50’s
  • 2 = /49’s

How your company doles these out, is up to them. However, almost no one is going to just directly carve out /64’s from the assigned /48 block, that is stupid. Generally, you’re looking to summarize and aggregate where possible throughout your network and we’ll assume you’re in location “A” at “company”.

We’ll go ahead and assume the company has decided each location is assigned a /58, which gives each location a total of 64 available /64’s to use. As you see, no different than standard IPv4 in the sense of ensuring proper aggregation, except now you’re no longer having to worry about the size of a VLAN’s subnet mask, you’ll always use /64.

What about private IPv6 address space?

If you do not want a Globally Unique IPv6 address you can indeed have what is called a “Unique Local IPv6 address = ULA”. There is a guide on how to properly generate these addresses, which includes a variable which references the time and date, along with other factors to ensure absolute uniqueness.

Why does this matter with private address space? Have you ever been involved with a merger/acquisition, or having to aggregate two offices together which use the same private IPv4 subnet range? I need not say anymore because this can be a PITA! Thus, ULA, when done right, ensures this will never happen; however, there is absolutely nothing stopping you from selecting your own, basic, prefix.

IPv6 ULA uses the FC00::/7 prefix, divided into two groups:

  1. fc00::/8 – The idea for this prefix is to be administered by some authority, but no one can agree to it, so just forget about it
  2. fd00::/8 – Is defined for the generation of /48 prefixes only, using the last 40 bits to generate a random, unique, prefix, according to the algorithm in RFC4193

You will want to use option 2 and you can use online generation tools like those from SiXXs or use a tool from another resource, either way, make sure it generates a proper /48 prefix for you and is, by some degree, RFC4193 compliant.

Finally, your company’s IT department is likely to have this /48 already and is almost very likely to have assigned you a prefix according to the same standards for which they’ll dole out their Globally Unique IPv6 addresses; thus, no additional explanation needed.

Get your DNS infrastructure setup for IPv6 AAAA and PTR-record resolution

I won’t delve into this much more other than you absolutely must make sure your DNS infrastructure is setup for IPv6 AAAA-record and IPv6 PTR-record solution or you WILL have issues!

One area to ponder is the hostnames that’ll resolve when you’re in a dual-stack environment. Do you want the same hostname to return on both a A-record and AAAA-record? Well, some say no, some say yes. Me? I say you should discuss this with your vendor to ensure their solution doesn’t have a problem with this, especially in a dual-stack environment. I was told, by co-workers who know more about Vmware vCenter than I do right now, this is a problem and the returned hostnames must be different when using dual-stack based environments.

Always research and question IPv6 support on your devices

This goes for hardware and software vendors, many have made claims their stuff works with IPv6; however, what, if any, testing was done isn’t known and there are a variety of scenarios to consider. For instance:

  • Does it support native IPv6 from installation-to-operation?
  • Does it support dual-stack, from installation-to-operation?
  • How does it handle DNS requests in dual stack?
    • Does the system start with IPv6 AAAA requests and then fails over to IPv4 A-record requests?
    • If so, what is the timeout if a AAAA record is not available and it must try for an IPv4 A-record?
    • Is the order of DNS resolution preference configurable? (Can you choose to have IPv4 A-records first?)
  • What forms of address configuration are available for IPv6? (SLAAC, static, DHCPv6?)
  • What IPv6 address types are supported? (Globally Unique and/or ULA?)
  • Are there specific “sections” of configuration which cannot support IPv6?
    • For instance, in Cisco NX OS, you cannot reference an IPv6 address for use on a vPC peer keep-alive link.

More questions will come to mind, but these are from experience and I can promise you are a lot of reasons why most IPv6 implementations in the enterprise, and data center, fail. Question all vendors!

This is it for now, hope this clears up some stuff for you out there who’re thinking about their IPv6 implementation


DNSMASQ is both a DNS and DHCP server that is quick and efficient to run on Linux systems and is likely already running on your Linux box. If you’re in need of a quick DHCP server to run your environment to serve multiple DHCP scopes for different subnets in your VLAN, of which we all know the best practice is subnet == VLAN == Broadcast domain, then DNSMASQ is your go to guy and I prefer it over the ISC DHCPD server. This quick tutorial will go over the basics of how to get this setup and running and assumes you’re not going to utilize the DNS service.

Create a directory for your DHCP leases file:

sudo mkdir /opt/dnsmasq

Setup dnsmasq.conf:

#
#Disable the DNS server
#
port=0
#
#Setup the server to be your authoritative DHCP server
#
dhcp-authoritative
#
#Set the DHCP server to hand addresses sequentially
#
dhcp-sequential-ip
#
#Enable more detailed logging for DHCP
#
log-dhcp
#
#Set your DHCP leases file location
#
dhcp-leasefile=/opt/dnsmasq/dnsmasq.leases
#
#Create different dhcp scopes for each of the three simulated subnets here, using tags for ID
#Format is: dhcp-range=<your_tag_here>,<start_of_scope>,<end_of_scope>,<subnet_mask>,<lease_time>
#
dhcp-range=subnet0,10.0.0.5,10.0.0.250,255.255.255.0,8h
dhcp-range=subnet1,10.0.1.5,10.0.1.250,255.255.255.0,8h
dhcp-range=subnet2,10.0.2.5,10.0.2.250,255.255.255.0,8h
#
#Setup different options for each of the unique subnets, since default gateways will be different
#The format for this is: dhcp-options=<your_tags_here>,<option>,<option_value> - 3 is router
#
dhcp-options=subnet0,3,10.0.0.1
dhcp-options=subnet1,3,10.0.1.1
dhcp-options=subnet2,3,10.0.2.1

Once this is complete, enable your DHCP service to start automatically. You should also check your systems firewall/IPTABLES service(s) to ensure you have created rules to allow UDP traffic over port 67 and port 68, or you can just flush your IPTABLES and/or disable your firewall, your choice, this isn't a security blog so I'll leave the choice to you, the person who knows their environment better.


If you’re looking to use command line variables for scripting stuff you have some predefined variables in the NX-OS environment to use and you can also create your own. For now, I’ll just show you how to use the most common, the switches hostname. In some environments you’ll have to save the output of a show tech file and later on upload it via SCP. However, if you’re doing this to 2 or more switches, you’ll need unique file names to make your life easier. Instead of going to each one, you can just use the variable SWITCHNAME in the file. So, if you’re using a script or something like cluster-ssh, this makes your job easier.


sh tech all > bootflash:///shtech-$(SWITCHNAME)


See my video on this very specific topology, what I’ve encountered, and the solution I found to work for me: