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.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Comments
well, OVH provides /128 so is Kimsufi. Not sure why they are doing this.
There are posts back to 2014 telling you that the entire /64 is yours here on LET, but it isn't necessarily true. I've seen IPs directly above/below on a /128 used in KS. That said, if you just set yourself as a /56 rather than a static route to the gateway interface on your physical node outside of your network, shit just works.
That has nothing to do with it and you know it. I meant when the AS is blacklisted, because a provider which turns a blind eye to spam will eventually be blacklisted or, even more, if the whole /32 is blacklisted (ipv6), did you get your /64 removed?
Everyone had to deal once in a while with blacklisted IPs, it is too much of a hassle to open a ticket so we may have removed it ourselves, however, when the whole provider is blacklisted, the choice is to move if you need to do anything more than p2p or some storage/personal processing/experiments/development there.
To give more detail: almost all subnets used by Arubacloud are widely blacklisted at the /24 level. I went ahead and got my IPs removed at the /32 level.
Good for you, I don't need my IPs to send mail from there outside my own servers, but, in the past, I wasnt able to remove any IP if the provider was blacklisted, got the nudge to move or no reply at all.
The right thing is to give people access to IPs from unique (to them) /64s. It is obvious you cannot really route a /64 to OVZ so RFCs are not really applying, therefore, the purpose being served, with 1-4 IPv6 each from a different /64 not shared with anyone should do more than nicely. People can't choose cafe's and dead beefs? Too bad, it is irrelevant here.
Our Xen/KVMs have even more /64s if they have more IPv4s: http://www.prometeus.net/returnipv6.php ON DEMAND, so that means people know what they are doing, and what IPv6 is.
Providers are always looking for reasons to upsell so why not give people the choice. Single IPv6 or charge for /64. Although I don't think anyone is trying to charge extra for IPv6. So maybe you can't get away with that yet.
I really hope you're joking.
"Want us to follow industry standards? Go pay an extra fee."
Seriously, WTF.
Imagine your car dealer selling headlights as an optional add-on: "You don't really need those anyway, just make sure to drive by daylight or on streets that are well lit."
Skimmed through the post and just heard REEEEEE screeching
Blah ipv6 blah not a /64 blah
You obviously don't know much about the hosting business.
I'm a big fan of one or two of the anecdotes on this page...
"2^52 (four quadrillion) addresses for every observable star in the known universe."
"So we could assign an IPV6 address to EVERY ATOM ON THE SURFACE OF THE EARTH, and still have enough addresses left to do another 100+ earths. It isn’t remotely likely that we’ll run out of IPV6 addresses at any time in the future."
2^128 is mindnumbingly big
As long as the big ISPs don't care about IPv6 it doesn't matter what you assign, because the routes suck anyways.
Routes suck ? Can you elaborate on this please ?
I don't think assigning /128 is a big problem as long as I don't use it as part of my business. I'd be pretty happy to set up, for example, a private VPN on a /128 for personal use. Yes it's like a NAT port in IPv4. But look, we are still selling and buying NAT port IPv4 VPSs as LEBs. Why are providers allowed to provide 10 NAT ports but should be blamed for providing 10 IPv6 addresses?
As long as the provider gives a full /64 upon a ticket, as a customer I'm happy.
Because many of them act like those 10 IPv6 addresses are equivalent (or even superior) to a dedicated IPv4. That's a very clear sign that they either don't know or don't care to do things "the right way"™.
I'm kinda scratching my head over that comment as well. My IPv6 ping times and raw speed is generally about the same as IPv4. So not only are the routes just as good, the whole "IPv6 header is larger therefore it's slower" argument is a myth as well.
This sounds like something you got in 2008 and forgot to re-check since then. Nowadays I very often get better routes and lower ping on IPv6 than IPv4.
Same here. That's usually because some backbone routers do not support IPv6, carriers have to upgrade them, so v6 runs on the upgraded part of the core, while v4 is likely not migrated there yet.
Other common example includes different routing policy or QoS/CoS, or lack thereof, for IPv6 in some networks.
It depends, some people still have to tunnel, but, yeah, it is long past 2008. On the other hand, some people do not implement traffic shaping over ipv6 since very few people use it so, if you need a burst, you can really burst it.
I'd like to point out that there is no technical reason a provider can't provide each client a /64. HE's Tunnelbroker offers a /48 for free and can be routed over a single IPv4 address without datacenter involvement or special hardware/software. While tunneled IPs aren't ideal, it's better than nothing and gets the job done where this RFC is concerned (this is how we do it for our one datacenter that doesn't offer IPv6 connectivity).
We did it too that way in Dallas but with own IPs.
The only problem is with OVZ which has some limitations. There is no RFC to apply in that case, so everyone does as sees fit.
We limit our OVZ VPSs to 128 IPs by default although we have had clients with up to 512 IPv6 addresses assigned, it didn't impact the VPS at all but it did break the vzlist commands (it would error out before it finished listing all of the VPSs).
What i meant is that routing provisions in those RFCs cannot apply to OVZ, so there is no real standard applying there.
Yes, you can, with
veth.How I wish it would be used more, I personally used it in all my proxmox installs, great thing, however, I havent met a provider offering it yet, prometeus included.
How so? There is still the relationship of a provider and a customer, and IP space on the public Internet getting assigned. The RFC fully applies.
Did you read https://clientarea.ramnode.com/knowledgebase.php?action=displayarticle&id=103 ? They do this without veth.
Indeed, however, the routing does not require a /64 in this case as a minimum, it is needed in autonfiguration situations, for example, but in venet situations, not.
So the customer can select the IP he likes more instead of being assigned at random. I think random is best, if people cant take a deal because they like dead beef cafe more, their business, once we had a customer say he wants the IPv4 to end in 88 because will bring him good luck. This is childish, if you really want it that bad, you get a KVM or Xen and have plenty of dead beef cafes to choose from.
Asians believe in this like we believe in the theory of gravity. It's cultural.
I recall a Paypal dispute of a past provider I worked for because the customer didn't get to choose the IPs from their /29 addons seeking that 88 (has a meaning with Neo-Nazis) magic number
Neonazis, magical symbols and asians... What a blast from the past
@doughmanes must secure a future for clean ips. (I hate that I've been on the internet to remember this much.)