Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!


When will IPv4 completely fade out in your opinion? - Page 3
New on LowEndTalk? Please Register and read our Community Rules.

All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.

When will IPv4 completely fade out in your opinion?

13»

Comments

  • FranciscoFrancisco Top Host, Host Rep, Veteran

    @Maounique said: So why not do it now, the more IPv4 we will have to migrate, the harder will be.

    Tons of them are on old cisco gear that doesn't support it alas :(

    Cisco is quite a dick about the whole supporting/not supporting IPV6 and literally releases a separate model to mark if it is or isn't supported. There's no reason that ipv6 couldn't be supported on those models.

    Sure they could upgrade to a new model but why if they have IANA blocks sitting around?

    Francisco

  • rm_rm_ IPv6 Advocate, Veteran

    @Aldryic said: Having an entire /64 on OpenVZ is just ridiculous; the container would take hours, if not days, to boot.

    And you STILL aren't even close to realizing how many IPs are in a /64. It would take about 2.7 TB of RAM to even store all the individual IPs (w/o doing anything with them).
    Hint: no one assigns all of them individually, you just assign one and route the whole subnet via it, so that the client can add and use any IP out of the /64 at will.

  • FranciscoFrancisco Top Host, Host Rep, Veteran

    @rm_ said: no one assigns all of them individually,

    Someone would just to be a dick.

    When we originally opened we had a spammer that we were leading on since we were bored and we told him how ipv6 gave so many ip's for free. So he kept bugging us to give him 1000's of addresses bound for 'reasons he didnt want to dicuss'

    We kept leading him on, knowing full well he was a spammer and in the end just told him to take a hike.

    You could give each client a subnet and then have a hard upper cap per VM, I guess that'd work and make things reasonable?

    Francisco

  • MaouniqueMaounique Host Rep, Veteran

    @Francisco said: Cisco is quite a dick about the whole supporting/not supporting IPV6

    So, maybe time to move ? This will happen all the time some vendor tries to push it like "my way or the highway", the more and more ppl will choose the second option.
    New players will strive to support anything they can think of, token ring and IPX if it comes to that, some will get it right and we will have a more diverse market, the netadmin's nightmare :P
    I can buy at the corner shop used cisco gear that looks almost new. It is a bad omen in my view.
    M

  • FranciscoFrancisco Top Host, Host Rep, Veteran

    @Maounique said: So, maybe time to move ? This will happen all the time some vendor tries to push it like "my way or the highway", the more and more ppl will choose the second option.

    New players will strive to support anything they can think of, token ring and IPX if it comes to that, some will get it right and we will have a more diverse market, the netadmin's nightmare :P
    I can buy at the corner shop used cisco gear that looks almost new. It is a bad omen in my view.
    M

    OK but 'why' ? If you're some old corp from the pre-bubble that has a /16 or something (remember, way back when they gave anyone a subnet without even caring), then what is the rush? Tons of those subnets are barely touched and that's why you'll sometimes see subnets randomly get announced elsewhere by spammers because they jack an email or things like that.

    Think for a second what your big selling point to these businesses would be?

    • They don't need more IP space, hell they barely sip what they have
    • Their network gear is old but they really don't need anything bigger

    I stand by the claim of 10 years honestly. Hell, didn't one of the biggest UK ISP's say 'no ipv6 till 2020'? I know shaw, canada's biggest ISP, isn't even considering IPV6 until they're 'close to out' of ip space.

    :(

    Francisco

  • FranciscoFrancisco Top Host, Host Rep, Veteran

    Also, good sir, remember this point...

    Old blocks now have a very high value on them. $16/IP is the 'gold standard' set by Microsoft for IP's.

    Francisco

    Thanked by 1klikli
  • rm_rm_ IPv6 Advocate, Veteran
    edited June 2012

    @Francisco said: When we originally opened we had a spammer that we were leading on since we were bored and we told him how ipv6 gave so many ip's for free. So he kept bugging us to give him 1000's of addresses bound for 'reasons he didnt want to dicuss'

    v6 RBLs (both as in 'online services' and in-house stuff large mail hosts have) are going to ban at /64 granularity, so it doesn't matter if he spammed from 1 or from 1000 IPv6s (but spammers commonly are dumb, not surprising he didn't know this). The whole /64 which they belong to, would be considered contaminated (by the whole world :). Oh you have other clients within the same /64? Whoops, sucks to be them.

  • MaouniqueMaounique Host Rep, Veteran

    @Francisco I never said 10 years is not realistic, in fact I agree that in about 10 years IPv6 will be so widely adopted that will make the value of old IPv4 blocks irelevant. We can tunnel them atm and will probably be even more integrated in the future, they will coexist for much more than 10 years, but the standard will be IPv6 in 10 years and even sooner. IPv4 will work tho.
    M

  • FranciscoFrancisco Top Host, Host Rep, Veteran

    @Maounique said: they will coexist for much more than 10 years, but the standard will be IPv6 in 10 years and even sooner

    We can only hope =\

  • IPv6! No more NAT!

  • nabonabo Member

    @LivingSouL said: IPv6! No more NAT!

    NAT is fine for private networks.

  • @rm_ said: Oh you have other clients within the same /64? Whoops, sucks to be them.

    That's my job to handle. We'll occasionally have a VPS get compromised and go on a spamming run, and the next morning I'll have an email from Spamhaus or someone else stating that they're listed a /32 or a /24. I investigate, resolve the situation, email them back, and the blacklist is removed shortly after (longest I've had to wait was 6 hours). We're not about to start assigning a v4/24 to each client just because an entire /24 can get listed :P

    Thanked by 1klikli
  • rm_rm_ IPv6 Advocate, Veteran
    edited June 2012

    @Aldryic said: We're not about to start assigning a v4/24 to each client just because an entire /24 can get listed :P

    In the v4 world it is very common that 1 user = 1 IP. So everyone blocks single IPs and that's it. In the v6 world however, 1 user will most often have a /64, because that's what (or more) every residential ISP gives out to a subscriber. Blocking on a finer granularity than that, will be seen as a waste of effort. You might get a special treatment from some RBLs (contacting them, explaining that "hey, we're a virtual server provider and we give out this and that per user", having them add some sort of exception), but I wouldn't count on that.

  • @rm_ said: In the v4 world it is very common that 1 user = 1 IP. So everyone blocks single IPs and that's it.

    Not quite. Please see the following:

    @Aldryic said: We'll occasionally have a VPS get compromised and go on a spamming run, and the next morning I'll have an email from Spamhaus or someone else stating that they're listed a /32 or a /24.


    @rm_ said: Blocking on a finer granularity than that, will be seen as a waste of effort. You might get a special treatment from some RBLs ("hey, we're a virtual server provider and we give out this and that per user"), but I wouldn't count on that.

    Agreed. But until the time comes when we decide to assign a /64 to each client, I will simply continue to have the listings removed as I do now.

    It's also worth mentioning that we've been offering ipv6 for quite some time now, and I've yet to receive the first abuse report regarding a v6 address. It's fair to say that providers can pretty much assign as they wish for the time being until v6 becomes commonplace enough to require stricter adherence to RFCs.

Sign In or Register to comment.