Thanks On Monday, December 24, 2018, Dave Taht <[email protected]> wrote:
> On Mon, Dec 24, 2018 at 11:54 AM Emeka <[email protected]> wrote: > > > > > > Dave Taht (1): > > net: Allow class-e address assignment via ifconfig ioctl > > > > Dave , please through more light on the above. > > That's the kernel, not babel patch. > > I've been engaged for several months now on a project to make class-e > routable with a few other folk. > > The linux kernel has been "class-e ready" since about 2008, capable of > both accepting and routing these unicast addresses, with the one > exception of an ancient bit of pre-cidr (1992!) code invoked by > early-boot and the ioctl invoked by the version of ifconfig used by > busybox and toybox among others. Modern distributions using the > iproute2 equivalent or netlink directly had no such limitations. Linux > 4.21 (backported to openwrt also) makes it possible to fully use > class-e under all circumstances. > > We've tested openbsd, freebsd, android, ios and OSX, and they all work > today (well, there's *one* teeny patch for ping on freebsd). Junos > requires a config option, some cisco routers work, some don't, all the > switches we've tried work, multiple iot devices work ,nat works, every > vpn we've tried works, and we know how to go about making windows work > in the long run (it can neither accept or route to 240/4 presently)... > and nothing thus far, crashes. > > The hope is to run a global test next year (with the help of caida, > iana, irtf and others) to further study how this space can be used to > keep ipv4 crutching along in the face of ever more extreme address > depletion problems. It would be a long slog (years!), and many rfc > updates, and infrastructure (as one example, we'd need > 240.in-addr.arpa) to make them fully usable, but just as we made > 1.0.0.0/8 and 8.0.0.0/8 work eventually, *so far* we see no technical > limits. > > In the interim, allowing the capability of 240/4 addresses in more > routing daemons and moving the policy into config files, rather than > hardcoding things, seemed like a good idea, and there are patches > outstanding for both bird and FRR, also. Please give 'em a shot, and > then ponder what would be the best possible use for the last available > 268m ipv4 addresses in the world.... > > > > > > On Saturday, December 22, 2018, Juliusz Chroboczek <[email protected]> wrote: > >> > >> > 'Cause I'd sent you a patch earlier for just e (240/4) and you didn't > >> > apply it? :) > >> > >> I'm a bad man. Please double-check commit 19a442ba. > >> > >> -- Juliusz > >> > >> _______________________________________________ > >> Babel-users mailing list > >> [email protected] > >> https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/babel-users > > > > > > > > -- > > P.S Please join our groups: [email protected] > > or [email protected] These are platforms for > learning and sharing of knowledge. > www.satajanus.com | > Satajanus Nig. Ltd > > > > > > > > > -- > > Dave Täht > CTO, TekLibre, LLC > http://www.teklibre.com > Tel: 1-831-205-9740 > -- P.S Please join our groups*: *[email protected] * or *[email protected] These are platforms for learning and sharing of knowledge. www.satajanus.com | *Satajanus Nig. Ltd*
_______________________________________________ Babel-users mailing list [email protected] https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/babel-users
