[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vm / vmg / vr / vrpg / vst / w / wg] [i / ic] [r9k / s4s / vip / qa] [cm / hm / lgbt / y] [3 / aco / adv / an / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / pw / qst / sci / soc / sp / tg / toy / trv / tv / vp / vt / wsg / wsr / x / xs] [Settings] [Search] [Mobile] [Home]
Board
Settings Mobile Home
/g/ - Technology


Thread archived.
You cannot reply anymore.


[Advertise on 4chan]


File: IPv6-image.jpg (64 KB, 500x292)
64 KB
64 KB JPG
>2001:0db8:0000:0000:1234:0ace:6006:001e

Setting up anything IPv6-related is unbelievably clunky.

Why did they have to make it so goddamn long? 128-bit for 340 undecillion possible IP addresses? Why couldn't they just have made it 64-bit (18 quindecillion)? Or just add a few bits to IPv4? Not to mention it's fixing a "problem" that was already fixed by NAT. Fuck everyone who worked on this retarded IPv6 standard
>>
Sometimes, I wish this board displayed the poster's IQ next to the post number
>>
>>96100176
>adding a few more bits to IPv4
What the fuck do you think Ipv6 is?
You'll also then run into the issue of how to differentiate addressing between different bitnets.
And you'd run into the issue of running out of that address space later
Ipv6 is just ipv4 but longer and with Hex. You can even shorten addresses with ::
Not to mention you no longer have to deal with the bullshit that is NATting
>>
>>96100176
oh look, it's this shit thread again
>>
File: EKM9pjTVAAIudzZ.jpg (51 KB, 680x680)
51 KB
51 KB JPG
>IPv4
>IPv6
>no IPv5
>>
>>96100176
Truly IPv6 was designed by complete autists.
Somehow they went with a 128bit implementation that isnt at all backward compatible with IPv4 instead of just extending the address space of IPv4 to 64 bit.

If they were gonna go with absurd overkill, why not go for 512bits so that a source/dest pair is exactly 1kb?
Just think of all the information they could encode into the address then hahaha (:
Every grain of sand on every non-gaseous planet in the solar system could be assigned 10,000,000,000 unique IP addresses w/ ipv6.
Not enough imho.
>>
>>96100594
ipv6 is for niggers
>>
>>96100362
I can setup cron and spam it every 4 hours if it bothers you that much.
>>
64 bits is too few considering jeets and niggers are being born exponentially
>>
File: 1665958967890913.jpg (977 KB, 2600x3631)
977 KB
977 KB JPG
>>96100176
>>96100594
>waaah it's so long
>waaaah waaaaah I can't make the same joke about 192.168.0.1 and 127.0.0.1
>waaaaaah it's soooooooo long it's triggering my tiny penis complex
>waaaaaaaaaaaaah ping fe80::5 instead of ping 192.168.0.5 it's literally unusable
>wah wah wah wah
>>
>>96100802
kys tranime
>>
>>96100349
>Not to mention you no longer have to deal with the bullshit that is NATting
This doesn’t mean each of my device’s ip address is exposed instead of one router address for all my network data does it?
>>
>>96100464
>>96100176
>>96100594
We did have ipv5, retards: https://datatracker.ietf.org/doc/html/rfc1819
it never left the Proposed Standard stage (was later archived as experimental) but it was pointed out that using 64 bits would only delay the address shortage by another 10 years or so (this was in 1981, before home NAT became at all standard), which is why it was never actually adopted outside of very niche fields where it was quickly scrapped for RFC 2460 (ipv6).
>>
They did not make 64bit addresses because you are supposed to get a /64 subnet, ie. you get a 64bit address and then every device can get an allocated IP by just claiming one of the 64 bit postfixes. This basically solves the IP allocation problem in a different/better way than DHCP.
>>
>>96100176
it's not fixed by NAT
>>
>>96100989
WHO THE FUCK NEEDS A SUBNET WITH 2^64 HOSTS ON IT YOU UTTER CRETIN
>>
>>96101052
Yes it fucking is. I followed the "omg it's running out soon" doomposting of the late 90s and early 00s and it never happened. Fuck yourself.
>>
>>96100987
thank you for this informative post, anon
>>
>>96101054
I do
>>
>>96100176
Stop thinking in IPv4. Idea is that everything has a subnet.
>>
>>96101170
Shut the fuck up YOU FAGGOT.
It is POINTLESS.
You MICROPENIS.
>>
>>96101198
It's not pointless to have subnets.
>>
>>96101218
IT IS POINTLESS FOR EVERYTHING TO HAVE A SUBNET YOU GOALPOSTFUCKING FAGGOT
>>
>>96101054
because there's no broadcast in ipv6 so it doesn't matter how large your subnets are it's all unicast and multicast so there is no broadcast domain to contend, you utter cretin.

>>96100176
average filtered sysadmin brain
I bet you don't even understand ipv4 tell me how do you differentiate RA marked packets so you are not punting unnecessary packets.

>>96100594
you're a moron, ipv6 was conceived of a very long time ago and they realized a weakness of ipv4 was the namespace and that is not even the main things delivered by ipv6, flow label next header and fixed packet lengths are a godsend ipv4 comparatively looks like the swarthy ugly cousin.


Once again an ipv6 thread full of people's who's only exposure to IP is through DNS to give them their Youtube anycast gateway IP to watch linus cuck tips videos. When will you dumb niggers learn that an IP packet is substantially more than just src/dst addressing? Or that there's a significant difference between JUST global addressing on the internet and addressing within the network.

ipv6 is suitable for forwarding plane operations which is why it is used in all serious networks that matter your trashy little SOHO or home network doesn't matter because it doesn't have to move data anywhere and it doesn't have to consider 60 unique paths all with different QoS and SLA's (which ipv6 can handle) IPv4 is not a proper forwarding plane.
>>
>>96100884
your question is weirdly worded

NAT lets multiple devices use one router IP address. ports on the router correspond to each device connected. external IP packets go to port x on the router, which then go to device y. theres a table in the router that keeps track of which port on the router maps to which device

t. learned off youtube cus 3rd year networking class was useless in teaching this
>>
>>96101268
Irrelevant. nobody needs it.
>>
>>96101296
Your brain would explode trying to figure out how datacenter or mobile networks don't implode in 15 seconds, iqlet
>>
>>96101296
consumers don't really need ipv6, but it's easier to just have everything use ipv6 instead of mixing ipv4 for consumers and ipv6 for businesses
>>
>>96101318
Burn them both down.
>>
>>96101332
Ok based, finally the correct answer
IP and IPv6 are intrinsically abhorrent and lead to allowing Melvin ass niggers to be terminally online
>>
>>96101268
how/where did you learn all this, anon?
do you recommend any specific books?
>>
>>96100219
It's already the case: last 2 digits of the post (there is no one above 100 IQ here).
>>
>>96101324
i think that will be the right move.
waitbros we're at it again
>>
>>96100176
>Or just add a few bits to IPv4
It means doing the same adoption process as with IPv6 (old hw can't recognize it anyway, there's no drop-in hack), which means it can be re-done from the ground up, which they did.
>>
>>96101275
>NAT lets multiple devices use one router IP address
Yeah so with nat every device’s packets use the same public ip address. So without nat do all devices have their own public ip address or do I still have one public ip address for my whole network?
>>
>>96101444
>So without nat do all devices have their own public ip address
Sure. 80s style.
>>
>>96101444
All devices have their own public ip address, yes.
And that is good.
>>
It might not be soon, it might not even be in my lifetime, but one day we're going to be happy we made IPv6 128 bit
>>
>>96101489
No. It is GAY and unusable.
>>
>>96101489
great. now i have to tunnel all my devices through my home server
>>
>>96101377
rolling
>>
>>96101370
If you just want the topics then all the big Network vendors and very large Network Operators have good documentation.
Cisco Libary is probably the best because it's not insufferable to parse, Nokias SROS help docs are also very good
https://www.cisco.com/c/en/us/tech/index.html
https://infocenter.nokia.com/public/7750SR222R1A/index.jsp
https://developers.cloudflare.com/

Also a lot of Youtube playlists
https://www.youtube.com/@aarontechtalk
https://www.youtube.com/@ping-factory-com
https://www.youtube.com/@DavidMahler
https://www.youtube.com/@NetworksLearning
https://www.youtube.com/@OrhanErgunCCDE
https://www.youtube.com/@TeamNANOG

I won't directly advocate you purchase actual books but
TCP/IP implementation vol 2


I originally started learning networking from the perspective of a developer and implementing things so that helped greatly since if you skip a lot of the flags and significant bits you miss a lot of why the protocol behaves the way it does or what the point was in the first place, and then you end up like most /g/ posters who totally misunderstand the concept.
>>
>>96101489
I can at least regularly have them repick an address though right?
>>
>>96101610
Don't BARGAIN with faggots, you pushover.
>>
>>96101602
I'm sorry.
>>
>>96101636
i am heartbroken. it is still low even with 3 digits.
>>
>>96101606
based, ty anon, I'll check these out!
I'm not a dev, so maybe that's why I struggle to understand the nitty gritty of networking protocols and such, but I'll probably end up learning some programming language along the way in the future so I'll likely revisit this
>>
>>96101602
blessed retard
>>
>>96101667
All of these are oriented towards understanding the protocols but not implementing them beyond just configurations on boxes so I could not recommend these more they are probably the highest quality networking related channels. They're not exactly beginner oriented but I am of the opinion that you osmosis most of what you need anyways and you can work backwards on some of it. If you try to study from the ground up CCNA style I think you get stuck like the vast majority of people in what is essentially ELO Hell where you will never understand or study anything beyond implementing vlans and inter-vlan routing with some basic stuff like DHCP or NTP.
>>
>>96101377
feeling lucky
>>
>>96101705
yeah that sounds exactly like what I need, my peers tend to recommend I read books like Microsoft Network Essentials by Darril Gibson and while it did help me understand some more terms relating to standards organizations, it didn't really enlighten me on the reasoning behind why things are implemented the way they are or what any of the advanced stuff even does or means.

Osmosis is definitely a valid strategy and ELO hell is a great way to describe what I felt like when going through CCNA.
>>
>>96101370

This is all easily available in any ccna course in the ipv6 unit.

All the retards in this thread claiming it's useless or overcomplicated are stupid niggers who don't understand the fundamental problems of ipv4 that it is designed to resolve.

Retards think nat fixed ipv4. Everyone else knows that it just saved ipv4 for the time being. Kinda like how recycling doesn't fix our consumptive self-destructive habits-- it just delays the inevitable.

We will all switch to ipv6 eventually but faggot tech journalists keeping shifting the goalpost as to when that will be.
>>
>>96101377
I feel like 00 should count for 100+ IQ
>>
>>96101602
have a you

>>96101377
rolling
>>
>>96101848
yeah, I don't really get why other anons don't do even the slightest bit of research in regards to the reasoning behind ipv6 outside of "heehoo bigger bit number". But then again, it's /g/.
>>
>>96101930
Agreed. Part of the problem too is that NAT and related bandaids to ipv4 work almost too fucking well, and they're so retard friendly that it essentially obfuscates the utility of ipv6.

I mean, I get it-- from a consumer perspective ipv6 just seems exclusively clunky-- and for the most part right now it is for the.
>>
>>96100782
jeets pop growth is already slowing down
i suspect africa will slow down as well when they industrialize more. and if they dont industrialize more they wont need ip addresses anyways
>>
>>96101989
Lmfao
>>
>>96102000
checked
>>
>>96100176
how do you go to port 8080 on an ipv6 address?
>>
32-bit network addresses ought to be enough for everybody
>>
>>96101981
I think the main reason for that is also just a general problem we're facing now: technology being dumbed down and hidden behind so many layers that it's harder and harder to learn about it enough to understand these kinds of concepts. Even worse is when people don't care to learn these things because "it just werks for them so why change it?"
>>
>>96102050
[a:b:c:d]:8080
>>
>>96100176
why couldnt they just use variable length int for ip address (similar to utf8)? why didnt they find a way to make it backward compatible? a bunch of senile idiots
>>
>>96102082
ipv6 addresses can be very short, shorter than ipv4 ones
>>
>>96102062
There are already almost twice as many people as the available ip space of a 32-bit range.

While I don't disagree with you in theory, in practice we either need more address space or we need to fire up the gas chambers again.
>>
how does ipv6 solve the exhaustion problem if they will just sell /8 to a company
>>
>>96102082
Yeah. It's all boomer tech. Of course it's retarded.

Find out a way to make it backwards compatible? Nigga do you know what you're asking? Computers know what ipv4 is because it is an established standard. Yes, the people who designed it were retarded when it came to future-proofing it-- that is the point. Ipv4 is FUNDAMENTALLY BROKEN and it is fixed by establishing a new standard - - ipv6.

Ipv6 has such an unfathomably massive range because it was designed to be future proof. They learned from the mistakes of ipv4, but you can't just change ipv4. It's baked into the cake! So they made ipv6 (which you also can't change) and that's pre-installed on every computer now too.
>>
>>96102187
Stop thinking in ipv4.
In ipv6 every device gets a unique address. There are 2^128 possible addresses. That's how they solve the ip exhaustion problem.
>>
>>96101981
NAT/PAT as a security implement is actually a really good idea, it's not a good idea for every device to be internet facing. That being said ipv6 has NAT implementations so nothing changes.

>from a consumer perspective
talking to tech illiterate family and friends that have both ipv4 and ipv4/6 they literally notice 0 difference and in the latter case the routers from their ISP almost always have some "prefer ipv6" setting. It is 100% functionally the same from a Luddite perspective it only changes when you start pretending you know anything.

>>96102187
they don't check IANA right now for the block allocations
the vast vast vast majority of allocations are /64 global routing prefixes
>>
>>96102243
i get it but we still need subnets for routing. what is stopping them from selling 1/65536 of the address space to apple?
>>
>>96102243
Why are you selling an antifeature as a positive thing? Having a uniquely identifiable address for every single device is one of the reasons ipv6 is a bad idea, if you are one of those freaks that cares about outdated concepts like anonymity, privacy, and liberty.
>>
>>96102368
ipv6 doesn't predicate that every device is globally addressable, NAT exists for ipv6.
>>
>>96102368
what is your point? there is no difference between ipv4 and ipv6 anonymity-wise. you being in a cgnat doesn't mean your traffic isn't logged separately.
>>
>>96102453
>>96102409
i can reset my router and have a new IP address. jannies would be able to enforce permanent bans with ipv6
>>
>>96100176
>Or just add a few bits to IPv4?
this would have been ideal
the unwieldy addresses are the tip of the iceberg
ipv6 is a failyre top to bottom, utterly idiotic design decisions made

>>96100594
you cant trust computer people with namespace sizes, look at:
millenium bug
guid needs mac + date = 10 bytes actual size 32 bytes
usb vid and pid both being 2 bytes
dont even look at java
>>
>>96102492
you can reset your router and get a new ipv6 address
>>
>>96102562
which is a unique address
stop selling antifeatures as beneficial you fucking kike. why do you hate anonymity so fucking much?
>>
>>96102000
Please post chart of india's exponential pop growth
>>
>>96102253
Why are you niggers this stupid? Security through obscurity is retarded and your IPv6 isn't "internet facing", it's behind a firewall with an implicit deny all inbound rule.
>>
>>96102639
All cryptography is security through obscurity. You're just depending on your special number being so obscure that computer hardware won't be able to calculate it efficiently.
>>
>>96100884
If configured that way yes, however your typical consumer routers and ISPs will still use NAT with IPv6.
>>
>>96101084
>it never happened.
yes, because IPv6 was introduced
>>
>>96101054
Me, because I'm going to find a way to assign an IPv6 address to every single sperm in my body and then cum on you.
>>
>>96101377
Rawling for 99
>>
>>96102592
i legit try to understand your concern. the new address will be a unique ipv6 among millions of addresses that your isp can assign to your city. with ipv4 it is probably just thousands.
>>
>>96102315
Nothing is stopping them, but the current policy is to assign a /12 to each region. So RIPE has a /12 for the entirety of Europe and middle east and they still have ~80% of the block unallocated, iirc.
A private company like Apple can't justify the need for a /16 to any RIR so they won't get one.
They can get a /32 like any other AS and only get a second /32 if they use up the first, which they won't.
>>
>>96102592
What the fuck are you talking about? If you are using IPV6 NAT and you lease a new address from your ISP's DHCP server this is exactly the same how your current IPV4 setup works now except with far more possible addresses to have on your gateway, in a way this offers MORE obfuscation not less. The only time privacy is an issue is when end nodes like your PC or phone have a publically routeable ipv6 address

>>96102639
if your pc has a public ipv6 address it's not behind a firewall by any definition and it's a PUBLIC address. If you firewall it off that's on your design there is no private ipv6 addresses by default there used to be site private addresses but they were deprecated, normally a firewall in the way that you think works at layer 3 is NAT not anything stateful or deep packet inspection
>>
File: download (3).png (5 KB, 294x172)
5 KB
5 KB PNG
Let me guess, you "need" more.
>>
>>96102639
>Security through obscurity is retarded
this is the most abused and tortured concept in security. You shouldn't rely on obscurity IN PLACE of proper security, but there's nothing wrong with having it.
ie (excluding home networks) you shouldn't rely on NAT as a REPLACEMENT for a firewall, but it's still slightly more security to have NAT AND a firewall.
It's the same thing with SSH on non-standard ports. I've heard retards say you shouldn't do that because that's
>security through obscurity!!!
but if you're using SSH keys or an acceptable password, that would be acceptable to use on SSH on the standard port, it's not any less secure to have it on a non-standard port and avoid thousands of bots a day trying to login to SSH.
>>
You niggas do know that NAT is just one of the functions of a firewall, right? It's a subset. You can't have nat without firewall, but you can have firewall without nat.
>>
>>96102903
>(excluding home networks)
I guess that exclusion isn't even necessary since home routers have a drop all incoming rule by default, so they do have a firewall as well.
>>
>>96102810
>if your pc has a public ipv6 address it's not behind a firewall by any definition and it's a PUBLIC address.

Are you fucking retarded? Do you know what a firewall even is?
>>
>>96102945
that's what I just said
also your home router isn't an actual firewall lmao your winblows PC has more security to it

None of you are running a stateful firewall
>>
>>96102963
You didn't read the rest of the post I'm going to disregard what you have to say where I explicitly talk about your insinuation, moron
>>
>>96102866
>172
kek v. good
>>
>>96102762
>they can own 1/4.3 billion of the pool and 2/4.3 billion if they use up the first
kek ip hoarders on suicide watch
>>
>>96100176
>>Why did they have to make it so goddamn long? 128-bit for 340 undecillion possible IP addresses? Why couldn't they just have made it 64-bit (18 quindecillion)?
luckily, thanks to SLAAC, it effectively is 64-bit!
>>
>>96102982
>normally a firewall in the way that you think works at layer 3 is NAT not anything stateful or deep packet inspection

I did and it just contains more retarded statements.
>>
>>96103532
waste of time, moron
>>
>>96100176
uhhh i think you mean 2001:db8::1234:ace:6006:1e sweaty :)
>>
>>96101489
I genuinely can't believe some retards actually like NAT.
>>
File: 1682456428868015.jpg (251 KB, 1080x914)
251 KB
251 KB JPG
>>96101268
>doesnt even know what forwarding plane means

get out you non network engineer nigger
>>
>>96100176
IPv6 is easy. The networking is a huge mess and needs to be simplified.
>>
>>96100219
Not OP but I agree with OP and my IQ is 142.
>>
You don't need ipv6 at home. A typical default IP range 192.168.0.1/254 can handle over 250 devices. Even if they shoved an IP in every fucking device you use at home (Toaster,Microwave,Fridge,etc) plus you rolled out 20 IP cams and 10 servers you'd still have plenty of unused IP left over. Bluntly; why fuck with something if you don't need to? It works, leave it alone.
>>
>>96100176
they could have just added area codes to IPv4 amiright? worked for phone numbers
>>
>>96104765
ipv4 cannot do TE it needs an encapsulated forwarding plane for any and all forwarding plane operations when you build your network so either an MPLS or ipv6 forwarding plane. IP in the data plane is only ever destination based routing. I bet you don't even know what an LSP is. Your 3 rack enterprise network NOC is not "network engineering" let alone "engineering"
>>
File: 1690261807737.png (82 KB, 350x263)
82 KB
82 KB PNG
>>96100884
>Not to mention you no longer have to deal with the bullshit that is NATting
>This doesn’t mean each of my device’s ip address is exposed instead of one router address for all my network data does it?
Yes but you'd use a firewall on your gateway instead which will prevent exposure anyway.
>>
>>96100176
>Why did they have to make it so goddamn long?
It's amazing retards still don't understand what DNS is in 2023.
>>
>>96104765
lol retard detected
>>
>>96106198
> A typical default IP range 192.168.0.1/254
The confidence in which people who don’t know shit talk on this site. Not only is that not how you specify an ip range you get a lot more than 256. For starters you get the range 192.168.0.0/16 (the 16 means 16 bit prefix i.e the range 192.168.0.0–192.168.255.255) you also get all of 10.0.0.0/8 and a bunch starting with 172 and I think others as well. Which is needed because companies easily need more than 255 ip address on the same network
>>
>>96100989
>/64
lolno, you shouldn't need more than a /120. Even /120 is too much, /124 is better.
>>
>>96102592
No. It's not a requirement to be unique. ISPs rotate your /48 or /64 for two things
>avoiding you to run servers
>avoid excessive fingerprinting and bans
>>
File: 1694482929356911.gif (3.47 MB, 464x524)
3.47 MB
3.47 MB GIF
>>96106983
>ipv4 cannot do TE it needs an encapsulated forwarding plane

you still aren't using the term forwarding plane correctly. yeah ipv6 has extensible headers in the frame that can be used for "traffic engineering" in a simplistic sense, but ipv4 already has mechanisms for that shit as is. are you dumber than fuck? MPLS is still used in conjunction with ipv6 networks. source based routing is a SECURITY CONCERN because of spoofing attacks, no one currently makes use of flow label fields how they were intended even to this day

i work in a high performance computing environment with multiple supercomputing clusters in a multi billion dollar environment doing all sorts of shit but mostly big boy nexus. have you ever touched a nexus 7718 or asr 9k? i bet you buy shit grey market and off-brand at your wowee 14 rack datacenter very impressive
>>
>>96105092
>easy
Good luck managing a multihomed network where inside users get 3 or 4 different IPv6 addresses and OSes take retarded routing choices such as routing local traffic through the Internet. Or domain controllers and print servers having multiple, changing addresses wreck internal services. Fortunately NPTv6 and ULAs were invented.
>>
>>96107403
nexus switches are small game, until you hop on any iosxr platform I do not care
>ipv6 has it in a simplistic sense
you know nothing of segment routing or any of the extension headers
>ipv4 has mechanisms for TE
no it doesn't holy fuck you're dumb. Adjusting unicast routing metrics is not TE actual caveman mode.
The point that you missed because you only play with babby mode boxes on iosxe is that IP requires an encapsulation whereas ipv6 in and of itself does not. If you have never looked at a segment id table I do not want to hear it.
All routing across large networks is source based since lsp state information is only held at the ilsr and elsr moron. Core nodes do not contain state information that's the entire point
I also do not have patience to handhold someone through a compute to completion asic model today
>>
>>96107403
>i work in a high performance computing environment with multiple supercomputing clusters in a multi billion dollar environment doing all sorts of shit
I refuse to believe someone smart enough to land a job like this posts in this god forsaken place
>>
>>96107545
nexus is NX-OS or ACI and asr9k are IOS-XR you are proving my point that you are a retard
>>
>>96107573
I am well aware all my asr9k's and 8000's are ios xr I couldn't care less about nexus enterprise babby boxes honestly nxos, wait until you find out what the san os and nxos were originally developed in parralel with. I'm going to frame on my wall that IPv4 natively supports TE, wise words from a dc monkey. Continue thinking source based routing isn't used I can't wait until you find out how mobile network slicing works.
>>
>>96100176
ipv6 unto the future. This will be the universal schemal when the human cancer has colonized Andromeda.
>>
File: 1577169971315.png (123 KB, 686x550)
123 KB
123 KB PNG
>tards being filtered by having to deal with Hex instead of decimal.
Should've switched sooner
>>
File: FvEEp2uakAA_Mt2.jpg (105 KB, 1284x1259)
105 KB
105 KB JPG
>>96106983
>>96107626
>you have a puny datacenter
>oh your datacenter is big you are a monkey in it
>IPv4 HAS NO TRAFFIC ENGINEERING ABILITY BECAUSE YOU HAVE TO USE OTHER TECHNOLOGIES MEANT TO BE USED WITH IT, WITH IT
>uhhhhh i know asr9k devices run IOS-XR i was just pretending to be retarded i have many of these devices myself hehe

if your argument is that type4 SRH and SRv6 provide NEEDED functionality i can point you to MPLS-TE and SR-MPLS
>>
ipv6 is shit for homelab since you need to go out and rent a vps that handles both ipv6 and ipv4 to act as a gateway. If verizon ever forced me to switch to ipv6 I'd just drop them.
>>
>>96107626
>jannies arguing about which size corridors and buffing machines they've used
>>
>>96103183
It doesn't make the IP any less annoying to deal with.
>>
>>96107302
The big question isn't OP, it's why home routers default the LAN to 192.168.0.0/24 and not 10.0.0.0/8
>>
File: 1613754394860.jpg (33 KB, 720x588)
33 KB
33 KB JPG
>>96108474
What?
Just use the link local space
>>
>>96108804
Example:
You host a minecraft server on your IPv6 home ISP. Your friend who uses IPv4 from their ISP can't connect to your minecraft server.

You can get a VPS that has both an IPv6 and IPv4 IP, and just proxy port 25565 to your VPS. Set A and AAAA records for myminecraftserver.com to your VPS. I use wireguard and TCP-redir (for purposes) in a similar setup.
>>
>>96108892
Your friend should just get an Ipv6 address or you NAT it through your ipv4
>>
>>96108909
Your home ISP won't give you an IPv4 address, only IPv6. Your friend isn't going to get IPv6.
>>
>>96108936
Maybe if you have a shit isp.
My isp gives me a dualstack
>>
>>96108973
If you have dual stack (and thus availability to ipv4), why use ipv6? Are you hosting multiple services on the same port? If not, you should just use IPv4.
>>
File: 919597444640.jpg (8 KB, 255x204)
8 KB
8 KB JPG
>>96109006
Because I don't have to deal with NAT bullshit
It's easier to address. Subnet. And firewall
And I'm educated on how Ipv6 works as a result and am prepared for it's use
>>
>>96101377
mommy says im special, ROLL
>>
>>96106804
Phone numbers are constantly being recycled and it's becoming a huge headache
>>
File: lazy-anime-girl.png (95 KB, 256x256)
95 KB
95 KB PNG
>>96100176
can't we just abandon IPV6 and just add a few extra 000's to IPV4?
>>
>>96100176
>thinks NAT is a good thing
If your IQ is as stunted as the length of an IPv4 address that would entirely explain the contents of your post
>>
>>96109006
>>96109414
Both of you forget that the main reason IPv4 sucks is because it makes the routing problem harder.
Yes, IPv6 doesn't need a NAT, and yes it's easy to address, but did you consider the economic implications of not requiring a third party centralized service like Root DNS and domain registrars to figure out how to find someone on the Internet?

IPv6 is the vehicle in which truly decentralized applications are born.
>>
File: soy.png (193 KB, 348x560)
193 KB
193 KB PNG
>NAT is... LE BAD!
>Security through obscurity! Haha gottem!
>NAT is UGH! So difficult to deal with *rolls eyes*
>You don't need NAT! I wrote custom firewall rules for my smart toilet to send telemetry home safely
>>
IPv6 was engineered perfectly, an address scheme that wouldn’t ever run out. People who say overengineered don’t get how much of a band-aid NAT was on a gaping tranny wound. Personally can’t wait to see the mass layoffs of people who refused to change or learn it.
>>
>iiieeee my datacenter so biggu!!!
>satureto arru my ipv4 addressesu
this thread kek
>>
>>96110204
>Personally can’t wait to see the mass layoffs of people who refused to change or learn it.
Two more weeks!
>>
IPv6 is good but my only issue with it is it isn’t easy to hand jam into shit or remember. I know you can reduce trailing zeros and remove segments of multiple zeros but even then the address is still pretty large. Gonna be a pain in the ass to keep track of devices when the host portion of your address is literally more than a trillion possible hits.
>>
>>96110280
No one remembers public internet IP address anyways unless it's shit like 1.1.1.1. Private network address aren't any harder.
>>
Why not just add a few more bits to the old IPv4? Then you could just have shyt like

192.168.1.100.123.123
or even double them
192.168.1.100.123.123.23.2

Then it would still be readable and not that 234fa76b7f6e7765a67 crap
>>
>>96110329
There is literally zero reason for ip address to be human readable. It's meant for computers to find each others. If you want a human readable address, there is dns.
>>
>>96110329
>Why not just add a few more bits to the old IPv4?
That's what iPv6 is.
>>
IPv6 confirmed to be designed by a committee consisting of apes, monkeys, chimpanzees and gorillas
>>
>>96110446
>IPv6 confirmed to be designed by a committee consisting of niggers
ftfy
>>
so do i enable ipv6 or not
>>
>>96110557
yes
bask in the knowledge that you are not complicit in corporations taking over the internet, and NAT piercing is not necessary to connect to you in a P2P application.
>>
>>96110329
>filtered by base16
topkek ngmi
>>
the trick seems to buy some ipv6 that have 0s in the start / middle / top so you can easily say
>ffff::::::fffff
>:::::::fffff
or
>ffff:::::::
but Im not even sure where to buy ipv6 addresses.
>>
>>96111109
All globally routable addresses are from 2000::/3.
https://www.iana.org/assignments/ipv6-unicast-address-assignments/ipv6-unicast-address-assignments.xhtml
>>
>>96110143
So why is NAT good?
What makes NAT CHAD?
>>
>>96111165
Im too tired and retarded to realize what that cidr range is in ipv6, will read up later
>>
There's nothing wrong with ipv6, i just cbf switching as a home user.
>>
>>96111165
>tfw no BEEF:BEEF:BEEF:BEEF:BEEF:BEEF:BEEF:BEEF ipv6 address that serves pictures of steak via http
why even live
>>
>>96100176
>Why did they have to make it so goddamn long?
::1 is too long for you? That's objectively shorter than
127.0.0.1
>>
>>96102690
>your typical consumer routers and ISPs will still use NAT with IPv6.
They will not. NAT6 exists but you're a fucking moron if you use it.
>>96100884
Use a firewall.
>>
>>96111232
Everything that's not a ULA.
>>
So do they have a plan in place to better manage ipv6 allocation, or are we gonna be fucked again when they hand out /8s to Google and pals etc? I know it's an unfathomable number of addresses, but the promise is that we can make use of them for useful things.
>>
>>96111509
They're simply not going to do that this time.
>>
>>96111509
Each continent can request a /12 when they can show that they've used up their current assignment.
There are currently 7 /12s assigned out of the 512 available within 2000::/3.
Each /12 contains a million /32s.
To be assigned more than a /32 as an ISP to need to show your local RIR that you're using all your current address space.
>>
>>96107417
That's why I wrote that the networking is a huge mess. You have to keep reading the rest of the post, anon.
>>
>>96111330
Ladies and Gentleman, this is an American.
Not thinking about
>B00B:B00B:B00B:B00B:B00B:B00B:B00B:B00B
>>
File: Clipboard01.png (29 KB, 926x939)
29 KB
29 KB PNG
Here's a map of the current IPv6 address assignments.
https://blocks.roadtolarissa.com/sdstrowes/raw/f419083a42fe6fbf33ef4e59128ff2b0/index.html
>>
>>96101377
checkem
>>
>>96101377
check
>>
>>96101377
(c)hoping
>>
>>96108295
> i can point you to MPLS-TE and SR-MPLS
MPLS isn't IP retard, the entire point was that IP(v4) intrinsically lacks all of these things needed in very large networks or for virtualization. Retarded nigger. You probably don't how either work anyways.
>>
>>96100176
>already fixed by NAT
You sound Jewish
>>
File: 1694639429623287.jpg (148 KB, 939x498)
148 KB
148 KB JPG
>>96113100
keep moving those goalposts little isp buttbaby they probably dont let you touch the equipment and make you use scripts to do anything
>>
it is fucking retarded because even when there are not shortage of ipv6 addresses you will still have a new random one every time you reset the modem... in our third world country.

what the fucking point then.
>>
>>96113160
the goalposts haven't moved whatsoever in the very first post you contended I said the same thing. It's understandable in your """"high performance compute"""" (400 transit nodes lmao)(normal mundane datacenter with nexus and catalust enterprise shit) and 100G links, lol, lmao, and the most simple case of up down routing you would not understand the ins or outs of what I said and bothered arguing anyways.
>>
File: 3573.jpg (77 KB, 647x767)
77 KB
77 KB JPG
>>96113272
>400 "transit nodes" in terms of a spine/leaf setup is small
>nexus is mundane it is outside of my realm of comfort
>has absolutely no contact with technology besides routing, doesn't even know what a gtm or ltm is or how to do simple layer 2 stuff
>doesnt refute he's not allowed to touch the equipment

gay jewish
>>
>>96100349
>Ipv6 is just ipv4 but longer and with Hex
No it's fucking not.
- attempted to shorten routing tables, failed, actually increased their size, thanks
- tried to sell unique WAN IP as a useful feature - maybe useful for a technocratic surveillance state
- fucking SLAAC

All IPv6 had to do was lengthen the address space. That's fucking it. But they couldn't help but inject 30 other flavors of autism into their project and now we have this abomination that is IPv6. They tried to change too much.

This is what happens when fragile academics apply theory to a problem. They think they know better. IPv4 had a bunch of antifragile features that some autistic commies thought they could just throw out and replace with their own perfect little world. They're non-practitioners with no skin in the game. Bunch of fucking pseuds. IPv6 is a complete and utter failure.
>>
>>96113631
>>96113631
>spine-leaf
ok so not high performance compute enjoy your buffer dependencies and literally no outscaling, hpc anywhere that matters hasn't been that for a decade you work in a trivial dc with equipment that only the untouchable caste cares about. Let me know when you make it to folded clos so I can make fun of you for being behind yet again.
>hurr you don't touch equipment
lol console jockey I'm who they call in when people like you break something yesterday was fixing some vBBU pools. I turn up equipment from companies you don't even know exist pretty neat actually.
>layer 2 vpc adjacent load balancing cope question
pathetic, I'm sure bridge domains confuse you ahh man why isn't this vlan tag being carried so hard! So much engineering!
>>
>>96113631
at least you admit through proxy you don't actually know how routing works, this is the first step to move beyond sticking everything in the link state database. Proud of you dc anon
>>
>>96110537
B-but isn’t racism not allowed outside of /b/?
>>
Force world governments and intelligence agencies in to ipv6 relinquish their ipv4 addresses they've reserved and you've saved yourself one big headache.
>>
>>96102093
Technically no, while you can compress an IPv6 address, it will always be 128-bit in length regardless.
FE80::1 is always equal to FE80:0000:0000:0000:0000:0000:0000:0001.
>>
It should be illegal for orgs like ARIN to issue new IPv4 addresses. From now on if anyone wants an address, they can only ever get a IPv6 address. If your hardware can't handle IPv6 then tough shit, stop being a nigger.
>>
>>96100594
>Truly IPv6 was designed by complete autists.
You're just making me like ipv6 even more now. Ipv6 is based and comfy and I'm sick of pretending it's not
>>
>>96101247
Literally every medium sized institution needs this
>>
File: 1646014377133.gif (460 KB, 500x500)
460 KB
460 KB GIF
>>96113878
>only contact with layer 2 is BDIs
classic

clos is spine/leaf and so is fat tree which is what i guess you're implying from your outscaling criticism why dont you go look up cloud scale ASICs before you start saying dumb shit about nexus to me yet again

just because your senior keeps complaining about having to buy cgnat equipment doesn't mean you need to make ipv6 your entire personality
>>
If my ISP properly implements dual stack IPv4/IPv6 is there any reason NOT to use IPv6?
>>
>>96114517
lol I do layer 2 all the time your question is just infantile.
You either don't even use RDMA or don't know what it is equally pathetic.
>cloud scale ASICs
this is your brain on enterprise babby boxes once again I do not have the patience to handhold you through a compute to completion or the various pipeline models keep throwing your sales brochure buzzwords at me like I don't already work with P4 I'm sure your cope-through <1tb capacity chips will save your low performance designs from a decade ago surely it will not contend cyclic buffer dependencies. Maybe you will get a trident 3 box in the spine surely this will let you cope and put the """high performance"""" in your """hpc""" datacenter. Hardware has always solved terrible designs that's why every network that actually moves data is still legacy designs right?

Talking to you has been a waste of time other than showing the next generation of people who want to get into networking that the bar really is low.

>CGNAT equipment
lol, lmao wait until you learn how the asic does nat
ip is not a proper forwarding plane, it has literally 0 functionality and must have all kinds of cope encaps specifically designed to do anything else ipv6 does not have this issue and natively supports all mpls features and many control plane features there's a reason why it is so widley deployed in networks that matter. Sorry you were filtered by IP and still struggle at layer 2 despite how easy it is I'm sure it took everyone else months to understand lisp just like you.
>>
>>96114517
folded clos is also not spine/leaf, you don't know what you're talking about. I/O stages are not leafs and Interm stages are not spines. They do not behave the same at all. You should be forced to read just one graph theory for cs students book. They need to start requiring degrees for this shit again else you end up like this.
>>
>>96115049
>requiring degrees
the schools will just lie and hand out the degrees anyway
>>
>>96115067
filter on skintone then, it's well known jeet and thirdie degrees are fake anyways.
>>
File: Fze4BA7WYAAg1AQ.jpg (77 KB, 968x903)
77 KB
77 KB JPG
>>96114991
again, misuse of the term forwarding plane
IP operations take place WITHIN the forwarding plane

calm down
>>
>>96115180
Literal semantics from a brainlet go, ask any actual network engineer what an mpls forwarding plane is and they're not going to delve into faggot sophism about "erm akshually the data plane is dictated by the forwarding operations for this protocol". Go read textbooks they use the exact same terminology go email cisco and tell them they're wrong. Again I have literally written code that tells the chip what to do.
>>
>>96100176
>Why couldn't they just have made it 64-bit (18 quindecillion)?
Futureproofing :)
>>
>>96100176
shut up incel.
>>
ipv5
>>
>>96115357
>shut up incel.
Back to /pol/ jannie.
This is our website: not yours.
Keep pushing your luck until we show you.
>>
>>96114991
>ipv6 does not have this issue and natively supports all mpls features and many control plane features there's a reason why it is so widley deployed in networks that matter.
other than the last part, which I doubt just because muh legacy, this is 100% correct. IPv4 was always shit and that's why v6 was such a massive shift. also network engineers really are mostly incompetent shitters, but this goes with almost all computer fields it seems. I don't fucking understand why there is such massive competency issues.
>>
>>96115067
>the schools will just lie and hand out the degrees anyway
can confirm. I have a degree and had to mostly "self-teach" myself virtually 80% of my knowledge. even core math classes usually gave you ways to squeeze through with no understanding and a C at my school.
>>
File: 1690295287675634.jpg (17 KB, 264x183)
17 KB
17 KB JPG
>his ISP doesn't have dualstack support

lol
lmao even
>>
>>96115084
wrong. whites are just as incompetent. this is beyond race.
>>
>>96115457
ya, mine doesn't. I wish I could sue these niggers. I use my existing Tinc net to forward v6 only traffic through my VPS though. v6 tunnels like the HE ones don't work on my ISP because they filter the traffic, either at the ONT or office layer.
>>
>>96115402
>which I doubt just because muh legacy
fair point but luckily the ietf thought of this and ipv6 easily supports ipv4 tunneling and mapping where needed so there really is no excuse
>so much incompetency
because it's an ultra critical field that always needs people the issue is it needs people who understand a lot and there are very few of those people so the roles are filled with people who were filtered by IP and don't even really understand it let alone things like label switching (an ipv6 sid is essentially a label).
>>96115460
Good morning sir
>>
>>96115218
you are conflating your "ip is a forwarding plane" with mpls which is not ip? not sure what you're up to with that

surprise surprise you're completely forgetting about frame switching layer 2 operations that also happen in the forwarding/data plane because you dont know layer 2

also theres qos, segmentation/reassembly, SPAN, error checking - you dont have breadth of knowledge because you've been only studying layer 3

its very apparent why ISPs want to adopt ipv6 there's a motivation for you all to do so to save a ton of money so you'll have to forgive me while i discount your bad faith arguments to justify it
>>
>>96115480
sucks to suck i guess
>>
>>96115569
Why are you strawmanning everything I said and think I am precluding those? I didn't "forget" frame switching you are now just bringing that up as if I was supposed to define some exhaustive spec for what a forwarding plane is that you can then retort "erm akshually".
If I say IP forwarding plane to an actual network engineer he knows exactly what I am talking about and isn't going to delve into faggot semantics. You are correct those things are very important especially in large networks which is why I know quite a bit about these things both across the network and in system as well. You have no idea how important L2 and L1 is in these kinds of networks. Nowhere have I made a bad faith argument I have laid out a few very good reasons why ipv6 is deployed widely. IP does not do what anyone wants and is a cope ipv6 does. This isn't bad faith moron, you argue like a circumstantial ledditor. Everything else I have said stands though I would like to see yet another tangent strawman.
>>
>>96111357
It would actually take me longer to write ::1 than to just tap the ip into my keypad.
>>
>>96115910
That's sad you should improve your typing
>>
File: 1692149301987424.jpg (206 KB, 1266x688)
206 KB
206 KB JPG
>>96115049
Folded Clos is two tier spine/leaf, retard.
>>
>>96113696
>for a technocratic surveillance state
It's called VPNs and P2P applications you spastic boomer
You're a tool of the corporations and you don't even know it
>>
>>96100176
My ISP doesn't support it.
>>
>>96100782
doesn't matter, ip v6 is meant to be interstellar-multi-civilizations proof
>>
>>96102082
it's slower to parse variable-length headers zoomer
>>
>>96114498
does your pc need a subnet?
you missed the point by a mile
>>
>The first version of this memo, describing a possible Internet Version 7 protocol was written by the present author in the summer and fall of 1989
https://datatracker.ietf.org/doc/html/draft-ullmann-ipv7-00
>>
>>96100884
That's exactly what it means. The geniuses who designed IPv6 decided it was more (((secure))) to require per-device security than to lock up a whole household of devices behind the same IP with the same firewall. This is totally not because internet standards are developed and implemented by ad companies and state surveillance agencies btw. Not at all...
>>
>>96104652
I genuinely can't believe some retards actually like making every single device with an IP address accessible from the public internet without any form of off-device security.
>just use a firewall hurrr durrrrrrrr
Yeah, because your smart thermostat, smart refrigerator, smart microwave, smart TV, smart watch, smart phone, smart toaster, smart water heater, smart air conditioner, smart light bulbs, smart baby monitor, smart doorbell, smart stereo, smart speakers, smart gym equipment, etc will all ship with a user-serviceable firewall, right retard?
>B-b-but just don't buy a smart device then!
Sure, I'll just go build my own fucking LCD panel factory and durable goods concern, since you literally can't buy a TV or appliance without that shit built into it anymore.
>>
>>96120526
Do you not know how a firewall works? You don't need per-device security if your firewall is blocking inbound connections by default (and it should be).
>>
>>96120622
Your router ships with a firewall. How do you think NAT works via IPv4 (hint: multiple firewall chains are involved)
>>
>>96120633
Correct. But people are scared of firewalls and do not know how to set it up. NAT just works, always.
>>
>>96120669
You don't set it up. You leave it in its default state of blocking inbound connections and turn off crap like UPNP that can potentially punch holes through it.
>>
>>96120633
>>96120661
With IPv6 each device needs its own firewall rules since each device has a uniquely-addressable IP. That's what makes it a retarded pain in the ass. It's custom-designed to allow exploitation of shitty IoT devices that are broadcasting publicly at all times and require incoming traffic in order to "just werk" with their shitty "smart" apps from any device. Consumer level shit from routers down to the devices themselves is going to come preconfigured shot through with security holes and no way to even address them since those configuration options won't be exposed to the end-user for le security purposes. You retards are literally agitating for a panopticon in order to make network configuration marginally simpler for your gaymen servers.
>>
>>96120828
>With IPv6 each device needs its own firewall
False.
All traffic passes through the router, the router is the ultimate arbiter of what gets through. It can and should block incoming requests which it knows about since it is tracking all connections.
>>
>>96120622
You're retarded. You have a router for a reason.
>>
>>96120828
No, you're just a fearmongering retard that doesn't understand what he's talking about



[Advertise on 4chan]

Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.