IP address runs out?

Status
Not open for further replies.
I

ismu

::phenomenon::.
Registered Senior Member
Since IP adrdesses only constructed by 4x2bytes numbers, and rapid growing of internet communities, how about possibilities that IP addresses for the internet will runs out? No matter the rule of IP address classification, Class A, B, C, and branching system using WAN, DHCP, Illegal IP-address, etc...

Is IP address won't be runs out anyway? How come?
How long current addressing rule will valid (if there new rule then)?
Is there anyone here have info about this?
Is there any solution on this problem?

I think this stuff will also usefull for other readers, specially IP based programmers/sys analyst. Tnx.
 
Last edited:
Yup, there's a solution :) People have been worried about this very thing for a while. The solution is called IPv6, and it simply adds 2 additional numbers, so the theoretical limit of addresses is something like 65536 x 4.2 billion (IPv4's limit). Even if Earth's population rose to 40 billion, there'd still be 6 thousand IPs per person. Of course, we might need more than six thousand robotic servants per person to maintain our lifestyles, in which case we can just go to war to reduce the population rather than reengineer things again (because reengineering is a pain in the ass).
 
I think IPv6 has been implemented in browser and server software for the past 1.5 or 2 years or so.
 
Compatibility?

Thank's Porfiry and Adam.

How about compatibility of IPv6 and IPv4? About softwares and programmable devices, it's realtively more cheap to upgrade. But on fixed harwares such as router etc., are those gonna unusable anymore?

Anyone has good links about this IPv6? Tnx.
 
(Warning: Hamster opinions are a simplified overview of a complex and changing situation...)

IPv6 addresses are a superset of IPv4 addresses. Old IP addresses will work under the new system. New IPv6 routers are aware of IPv4 routers and may interoperate by encapsulating IPv6 headers inside IPv4 headers. IPv6 network islands may communicate across an IPv4 backbone.

IPv6 adds complexity and cost to routers so few companies want to be first to commercially deploy. (Alll the router companies have IPv6 projects.)

There is little economic pressure to deploy IPv6. Reassigning blocks of unused IP addresses and using dynamically allocated IP addresses satisfies most companies buying router equipment today.

The companies who would benefit most from the increased address space of IPv6 are makers of IP devices such as cell phones or coffee makers. That need could be satisfied by an IPv6 network operating in conjunction with the main IPv4 Internet.

The big advantage of IPv6 is advanced protocols to support quality of service for applications such as Internet telephony and video.
 
I would like to see, instead of simply releasing expanded sets which will always run out eventually, a different way of doing it. Instead of just adding more numbers I mean. Maybe make the first set of digits a class meaning: government, business, private, educational institution, et cetera. Next set of digits a geographical code maybe. Third set a state or city code, et cetera. Have the sets of numbers represent something which can not be used to represent two things. For example, there could only possibly be one address which is Educational.Australia.Victoria.Clayton.Monash. The classes as they are now are basically a mess.
 
Adam...

That would be easier to remember...
What do you think if using Postal Code as first digits numbers? So, just by take a look at the IP we know where the server was sitting. :D
 
I discovered the last 5 days that I'm no expert in IP addresses...

But what's this IPv6 thing actually?
I see different lengths and seperations so far I found things...
Can anyone display an IPv6 Address here?

But I don't believe there will litteraly be an end to these IP addresses...
If the Internet does run out of these IP addresses, why don't they just add 1 Byte to it so you have a 5 byte address...

Then there will be a limit to *255.*255.255.255.255.
Let's get the calculator :)

1.099.511.627.776 addresses instead of 4.294.967.296...
Usefull for off-planet *if ever* communication by assigning the first byte to a planet or what-ever.
*And assign each part to something in some logical way...*

Or we could make network layers.
When the Internet get's full, make gateways to another network.
A second ARPA-net or Internet.
 
Psy-learner...

Then there will be a limit to *255.*255.255.255.255.
Let's get the calculator
1.099.511.627.776 addresses instead of 4.294.967.296...

Not all of those addresses can be used, isn't it? There a rules that several addresses are reserved and/also used for special purposes.
Beside that there are peoples who 'buy' lots of legal addresses but not operate those addresses (, yet).

If the Internet does run out of these IP addresses, why don't they just add 1 Byte to it so you have a 5 byte address...

Adding 2 more bytes, will longer lasting isn't it?

Or we could make network layers.
When the Internet get's full, make gateways to another network.
A second ARPA-net or Internet.

Using too many level gateways will slows down the speed connection.

Umm... i'm not sure about my comment also. I think we need to hear more info from Imahamster2\777 here... :D
 
(Imahamster is operating from memory so be warned…)

IPv6 is meant to fix many shortcomings of IPv4. One issue is addressing. The problem is not just the number of addresses but also routing tables. How does a router know where a particular address resides?

In IPv4 the network is essentially divided into local network islands interconnected through the public Internet. Locally, if a router doesn’t know an address it passes the packet to its default gateway. Eventually that packet is passed to a router that can identify the proper IP location for the packet. The packet is then forwarded from router to router based on route tables until the packet reaches the destination network.

Routers regularly broadcast routing tables to other routers that then use those tables to update their own route tables. (This is very simplified explanation of a complex process.) As the Internet has grown the size of those routing tables has grown and the bandwidth needed for network maintenance has increased. Basically every local network is represented in those route tables and there are now very many local networks.

IPv6 supports a hierarchical addressing scheme similar to postal addresses. A router no longer needs to know about all networks. It knows about neighboring local networks and regional routers. Route tables may be smaller.

Modern high-speed routers use special hardware controlled by micro-code for IPv4 address resolution. The IPv6 addresses had to be compatible with such switching technology.

And of course the new address format must be compatible with the IPv4 format. As Ismu says, some addresses are used for special purposes such as broadcasting, tests, etc.

All these reasons and many more contributed to the IPv6 address design decisions. (Lots of politics in the process as well.)

(Addressing is only a small part of the IPv6 standard.)

Addressing is not a crisis that will stunt the growth of the Internet.

Ismu, you wanted comments? This hamster believes Psy-learner is correct that network topology may be used to alleviate the addressing problem. And IPv6 networks may co-exist with IPv4 networks through tunneling with encapsulated packets.

As for too many gateways slowing down connections…modern routers are very fast. As old routers are replaced with new, router delays are less of an issue. This hamster strongly favors direct high-speed connection topology rather than a hierarchical network topology but only partly because of router queuing delays.
 
Imahamster777...

Good explanation. Thank's for your comments, Imahamster. I'll search more info about IPv6 somewhere else. Bye. :)
 
The reason for starting out late with this thead was to talk just about various issues regarding IPv6.And not the basics as they can be explained better by others than my own detailed IPv6 Headers complte description;).

The controversies regarding IPv6 were tremendous because of the design issues,that i am sure many of you will be aware of.
lets take them one by one.
First a fight developed over the length of hop limit field.one camp felt strongly that by limiting maximum number of hops to 255(implicit in using an 8-BIT field) was a gross and a <B>BIG</B> mistake.after all,32 hops are common now and 10 yeasr from now the longer paths may become a common place.
the response to these arguments could be made to increase every field,leading to Bloated header.also function of Hop limit field is to keep packets from wandering around for a long time and 65,535 is far too long.As internet grows,more and more long distance links will be built making it possible to get from any country to another in half a dozen Hops. at most,if it takes more than 125 hops to get from the source and destination to their international gateways,something is wrong with national backbones.the 8-BITTERS won this one.

Another shit was the maximum packet size Crap.the supercomputer community wanted packets in excess of 64 KB.when a supercomputer gets started transferring,it really means business and doesnt want the god-damm interuption of any kind every 64 KB.the arguement against large packets is that if a 1 MB packet hits a 1.5 MBPS T1 line(which i have always desired and dreamt off) the packet will tie line up for over 5 secs,producing a very noticeable delay for interacive users sharing the line.A compromise was reached here:Normal packets are limited to 64 KBs,but the hop-by-hop extension header can be used to permit jumbograms.

A third hot one(in fact as hot as it gets)was the IPv4 Checksum shit.some people likened this move to remove the brakes of car,doing so makes it faster without any delay,and erm...;)but if unexpected event happens erm...like...you got a problem BOB.
the arguement against checksum was that any application that really cares about data integrety has to have a transport layer checksum anyway,so having one in IP is Overkill.(One more infact in data link layer checksum),furthermore it ws expensive.


to be continued...

;)
bye!
 
Contd.

Moblie hosts were also a point of contention.if a portable computer flies halfway around the world(like my dad's;)),can it continue operating at the destination with same IPv6 address?or does it have to use a scheme with Home agents and foreign Agents?

Mobile pcs also introduce some assymetry into routing systems.it may well be the case that a small mobile comp could hear the powerful signal put out by a large stationery router,but router cant hear the feeble signal put by a mobile host.
Consequently some people wanted to build the explicit support for mobile hosts into IPv6.the effort fell flat when no consensus could be found for any specific proposal.(Talk about people!ha!...)


to be continued...

bye!
 
Probably the biggest issue was security...
Everyone agreed that it was needed.The was about where and how exactly the implementation should be done.
Let us take Where first:
the arguement for putting it in Network layer is that it becomes a standard service that all applications can use without planning.the arguement against it is that reaally secure applications want nothing less than end to end encryption,where the source applicaion does the encryption and destination application undoes it.with anything less,the user is at mercy of potentially buggy network layer implementations which has no control of.the response to this arguement is that these applications can just refrain from using IP security features and do the job themselves.the rejoinder to do that is that the people who dont trust the network to do it right,dont want to have to pay the price of slow,bulky IP implementations that have this capability,even if it is disabled.

Finally the controversy is surrounding the implementations of which algorithms?While MD5 was thought to be kool,recent advances in "Kryptuh" "graph" may weaken the whole thing.
Tell a serious cryptographer that DES is most secure and he's gonna laugh out loud at you.the compromise was thus mandate to security in IPv6,use a state of art checksum algorithm for good authentication and a weakish shit algorithm for secrecy,at the same time gives users the option of replacing it with their own...


bye!

bye!
 
Don't forget RFC 1913

Or the fact that you can use the 10.0.0.0, 172.195.0.0 and 192.168.0.0 addresses in your private network. All that is needed is $BIGCORP use the 10 network and so can $BIGCORP[1].

I think some ISP's do this now as well.
 
Wooops

Of course, I really meant RFC1918.

For some reason I can never remember all 3000+ RFC's and their titles.:p
 
Status
Not open for further replies.
Back
Top