IPv6 Address Assignment Example

Rene,
Greetings from Washington, DC. I have question regarding the IPV6 subnets and hosts. Once I determine my subnets as you describe in the lesson do you simply start counting up to find the valid hosts for that subnet? For example, for the network: 2001:41f0:4060:0001::/64, for the first host on this subnet can I simply put in 2001:41f0:4060:0001::1/64 for the first host on this network?

Thanks.

Hello Willie.

You got it, thatā€™s exactly right!

Laz

Hello Willie.

You got it, thatā€™s exactly right!

Laz

19 posts were merged into an existing topic: IPv6 Address Assignment Example

Hi I had a question. we are always told no more than 500 IP addresses in a subnet yet here we have 18,446,744,073,709,551,616.00 IP addresses. whats the idea on how this should be handled?

great IPv6 lesson Iā€™m actually enjoying IPv6 for first time in my studies. in past years I actually cringed learning it now im having fun!

2 Likes

Hello Brian

Yes, I understand your concern. After spending years (and some of us decades) learning and understanding IPv4 with both its strengths and its limitations, it is very often hard to avoid viewing IPv6 in a similar manner.

Now if you have a prefix length of 48, 64 or even 96 bits which are all very common in IPv6, then of course you will have a subnet capable of supporting an ridiculously enormous number of hosts. Although IPv6 can actually handle a greater number of hosts per subnet than IPv4 (because there are no broadcasts and because it handles addressing differently) it would not be wise to actually use all those addresses.

With IPv4, in order to conserve addresses, we subnetted our scopes to sizes comparable to what we need. With IPv6 you donā€™t need to do that because there are just so many addresses available there is no concern for address exhaustion. So, in order to simplify addressing, just make all your subnets with a /64 prefix and be done with it. Just donā€™t actually physically place more than 500 devices within a subnet.

I hope this has been helpful!

Laz

2 Likes

Hello,
Is any reason that we donā€™t use subnet 0 in IPv6? I never seen any reference using subnet 0x0000. For example: 2001:41f0:4060:0000::1/64. And I couldnā€™t find any reason behind this. Is this reserved by any RFC?

Thank you.

Hello Network E

Subnet zero is a valid subnet to use. There is no reason not to use it. It could be that some ISPs prefer not to offer it as it is the first subnet in the series and can be used to refer to blocks of subnets, but that is strictly supposition. There is no technical reason not to use it.

I hope this has been helpful!

Laz

Thank you Lagapides,

I think the explanation in CCNA book is confusing:
Therefore no one answered https://learningnetwork.cisco.com/thread/101127

1 Like

Hello Laz,
Just doing some light reading on a Sunday and i came across this. Since there are no broadcast with ipv6 and it handles addressing differently, can a small company say with anywhere between 500 to 1500 nodes just use one ipv6 subnet for their complete network without any performance issues? Just curious.
Thanks,
Cecil

Hello Cecil

This is an excellent question. Thereā€™s no clear cut answer :stuck_out_tongue:. It all depends on your network architecture and design. Like you say there are no broadcasts used for IPv6 so in a pure IPv6 environment, you will only have multicast packets being sent. However, keep in mind that any L2 switches in your infrastructure will still ā€œbroadcastā€ those multicast packets out all of their interfaces unless those switches are configured with a feature called Multicast Listener Discovery (MLD) for IPv6. As Cisco documentation describes it, ā€œMLD snooping constrains IPv6 multicast traffic at Layer 2 by configuring Layer 2 LAN ports dynamically to forward IPv6 multicast traffic only to those ports that want to receive it.ā€ More on this feature can be found here.

So, if you employ pure IPv6 in your infrastructure and you enable the MLD feature on all L2 devices, then you can have a more efficient network segment/VLAN. Does that mean that you can employ a subnet of 1500 nodes? The answer is an emphatic ā€œprobablyā€. It also depends on the applications youā€™re running and the nature of the traffic and the devices on the subnet.

Having said all of this, in general, having smaller subnets is not only beneficial for network efficiency, but also for network organization and security. So even if you can create a subnet with 1500 users without any performance degradation, you would rarely do so, not because of ā€œbroadcast problemsā€ but because of security and management issues.

I hope this has been helpful!

Laz

1 Like

Hello Laz,
Thanks for the reply.

1 Like

In a customers network who has been assigned a /48 what type of prefixes are you most likely to see in the RIB? Just /64 and /127 for point to point?

Would you use a /64 to advertise a loopback address in ipv6? Or try to be equivalent to an ipv4 /32 loopback advertisement and use a /128?

I know this is a general question - just trying to get familiar with best practice for ipv6 prefix size!

Hello Gareth

If I was given a /48 IPv6 address space, I would keep it as simple as possible. Iā€™d separate my address space into equal sizes and not bother with various prefix lengths. There are just so many IPv6 addresses that creating /127 point to point subnets is just unnecessary.

Letā€™s say I was given 2001:ABCD::/48. I would take this and separate it into subnets like so:

2001:ABCD:0:1::/64
2001:ABCD:0:2::/64
2001:ABCD:0:3::/64
2001:ABCD:0:4::/64
2001:ABCD:0:5::/64
2001:ABCD:0:6::/64
ā€¦
ā€¦
2001:ABCD:0:FFFF::/64

That would give me about 65500 subnets with 2^64 hosts each. I would use the same prefix size for all of my subnets, whether they are end user subnets, point to point links, or management subnets. For most organizations this is more than enough address space to be able to do this.

Sometimes itā€™s difficult to disengage from the subnetting mentality of IPv4 whose goal was to conserve addresses. Here we donā€™t need to do that, and this is actually one of the things that simplifies IPv6 compared to IPv4, even though it may not seem simpler at first glance.

Now if 65500 subnets is not enough, and for some ISPs this may actually be the case, then you can always further subnet these subnets by using a larger prefix. For example: 2001:ABCD:0:1::/64 can be further subnettet using a /80 prefix like so:

2001:ABCD:0:1:1::/80
2001:ABCD:0:1:2::/80
2001:ABCD:0:1:3::/80
2001:ABCD:0:1:4::/80
2001:ABCD:0:1:5::/80
ā€¦
ā€¦
2001:ABCD:0:1:FFFF::/80

This will give you about 65500 networks per /64 network (a total of over 4.2 billion) with 2^48 hosts per network. (BTW, thatā€™s as many networks as there are addresses in the whole of the IPv4 address space!! It boggles the mind :stuck_out_tongue: ).

So you see there is currently no need to employ an address conservation mentality, and you can keep things simple by maintaining a constant prefix across all your networks.

I hope this has been helpful!

Laz

This is a great piece of information Laz. Thanks for sharing. I donā€™t think Iā€™d ever have considered that any aspect of ipv6 could be simple in comparison to ipv4!

The numbers involved when allocating ipv6 subnets are astounding and my ipv4 mindset can not help but panic when we talk about 2^64 host addresses on a point to point link!!!

One question on the /80ā€¦ if we implemented this how would an interface assign itā€™s 48 host bits in the case of a SLAAC assigned Global Unicast address?

Hello Gareth

I think thatā€™s the sentiment that almost everyone has when shifting from IPv4 to IPv6. However, as you become more familiar, you can see how much more elegant IPv6 is. Iā€™d prefer to create a subnetting plan for an IPv6 network over an IPv4 network any day!

In order to do this you would have to use a DHCPv6 server as SLAAC functions only with a /64 prefix.

I hope this has been helpful!

Laz

1 Like

Hi Rene/Laz
Do you happen to know whether there is any likelihood of an ipv6 lab on the CCNA exam or will ipv6 be limited to theory questions?

Regards,
Phil.

This is on the 200-125 exam incidentally.

Hello Philip

IPv6 is included in both the outgoing CCNA course (200-125) as well as the new CCNA curriculum (200-301) which will be available from February 24th 2020. For the expected content in this new exam, you can take a look at the course outline for the related lessons here:

Under the Unit 4: IP Connectivity section, there is a subsection containing most of the IPv6 topics. You will also find some IPv6 topics under 4.4 Routing as well.

You can also take a look at the official Cisco blueprint which contains all of the expected topics.

Now when you use the term ā€œIPv6 labsā€, in general, CCNA level exams donā€™t involve the creation of the examination of a particular lab topology. The topics covered are more ā€œknowledge basedā€ using multiple choice questions, but still require a good underlying understanding of the concepts involved.

I hope this has been helpful!

Laz

How would we go about assigning a point to point subnet /127 in IPv6?