44net

44net@mailman.ampr.org

3196 discussions

Network Routers for Harsh Environments
by Jason McCormick 28 Jul '19

28 Jul '19

Hello all, I am looking for some recommendations on network routers for harsh environments. By harsh, I mean hot but not particularly dirty. I know there's a number of people on this list who use Mikrotik and Ubiquit EdgeRouters for your deployments and I am looking for thoughts on those product lines, as well as any other similar products. I am looking for a reasonably low-cost device. The key items I am looking for are: - Wide operating temperatures. Some of the sites are tower buildings that don't have A/C and during last week's heatwave the in-room temperature in one of the buildings was 105F-110F (killed an SRX power supply). None of the rooms will freeze in the winter but the temp will get quite low. - RELIABLE -- some of our sites have limited access and frequent power cycles, port failures, etc. are show stoppers - IPv6 support for basic routing and switching functions - Support for BGP and OSPFv3 (routing tables < 100 items) - VLAN tagging - IPSec or OpenVPN tunnels as clients - No firewalling or the ability to pass traffic without firewalls - note I don't mean a "permit any/any", I mean literally the device operates as a router or traffic can bypass a firewall-like function. - The device is a pre-packaged commercial system; I don't want to have to mess around with custom Linux/BSD installs, etc. - A device that can operate as both a router and switch would be nice, but not required As background, our club has a wide-area network around southwest Akron, Ohio-area stretching out to Rittman, Ohio and southeast to Alliance, Ohio with legs planned in other directions. This network supports a number of ham repeaters as well as other related services for the repeaters plus some of the site operations for the tower locations. Most of the gear is showing its age and we're going to start a wholesale replacement of most of it later this summer/autumn. All of the sites are connected with narrow-beam WiFi using point-to-point dishes, currently N but upgrading to AC AirMAX when we do the hardware refresh. Each site has a router or switch or both (depending on size and needs). The gear is a mix of Juniper SSG5s, Juniper SRXs, HP ProCurve switches, and some other random items. As part of the work, we're going to do a complete renumbering of the networks (which have grown up ad-hoc over the years), incorporate better use of 44Net space, and overlay IPv6 across the whole network. The reason for not wanting firewalling between the sites is we've run into problems with link reliability over UDP (losing a few packets) and also certain network device behaviors that don't survive multiple hops through stateful firewalls for a variety of reasons. Any thoughts in this community? I've been looking at the Mikrotik hEX S and the Ubiquiti EdgeRouter X as example devices. I don't have experience with Mikrotik gear although I hear it mentioned here a lot. Thanks Jason N8EI

4 3

Re: [44net] Is this a worry ?
by Rob Janssen 25 Jul '19

25 Jul '19

> < IP 0.0.0.0.5678 > 255.255.255.255.5678: UDP, length 106 > > They say a little knowledge is dangerous ! > At the moment at my tunl0 ( remote location ) I am getting groups of five of the above line every minute in my recently installed gateway. > I have tried without success to stop it appearing using iptables with < -A INPUT -p udp -i tunl0 --sport 5678 -j DROP >. Well, there is no need to do that really. Those packets are not dangerous, they sometimes may be even useful. This is a router discovery protocol that will inform you about manufacturer, version number of software, router name etc of the device sending it. When you have a device that supports it (sends it), you will also have some overview table in the user interface there which shows the decoded info received in the last 2 minutes or so. At first I also disabled this everywhere I saw it, and I am still not sending it on IPIP tunnels (where it actually is a bit troublesome for some cases because every minute such a packets would go out on all 500 tunnels at the same time and would result in a traffic burst that may overflow some queues somewhere and would result in packet loss of other usage). But on the VPN links that we use with our new style connections, I now keep it enabled. It provides a nice overview about who is connected and what software they are running, and I occasionally mail people that never upgrade their software to urge them to do so. As the mechanism for enabling/disabling these broadcasts has changed a little in the recent past in MikroTik firmware, and people not always update their software, some people that run an older YO2LOJ (Marius) script version may unknowingly be sending these things. But they are not dangerous and should not cause any noticable traffic volume. Rob

2 1

Re: [44net] 44 net connectivity
by Rob Janssen 25 Jul '19

25 Jul '19

> But, afer almost 20 years playing around witht his topic, I think the lack of specific network services on 44net is the actual element preventing the spread of its usage. > The original goal was conectivity, global if possible. > But there simply is no incentive for using it anymore. Dx clusters are reachable on the regular internet, repeaters and reflectors too. Specialized sites dedicated to contesting are homebrew stuff, or whatever, use regular internet access. They can not rely on a network where only 10k users are reachable. And BBS systems are outdated and replaced by forums, mailing lists and discussion groups on various platforms. > Unless we can find desirable services that really need a "private" globally routeable network for them to work, there will be no real new users influx beyond the occasional network enthusiast. I agree with that, it does not make much sense to run it as an isolated network. However, the 44net address space has enabled us to build a radio linked network and have infrastucture like repeaters, WebSDR, VoIP exchanges etc attached to it and also connect interested users to the same network. Without NAT issues. For that network to be useful though, it needs to be connected to internet in a reasonably efficient way. That is why we announced our country network on internet, and now we can use the services both on the radio network and from internet. (of course as far as firewall rules permit, you really don't want to route all traffic from internet to your radio network...) Still it is difficult to interest new users. However, when we get an interested user and he wants to setup some new service or experiment with what is there, shouldn't we make access to the network easy for new users, instead of making them fight with their router settings? The 44net also often makes things easier by having transparent routing. E.g. to setup a system with echolink usually is a nightmare because of the port forwarding requirements. By connecting to 44net and putting the echolink system on a 44net address, the filtering in the ISP router is eliminated and it is much easier to make it working. Rob

7 8

Re: [44net] 44 net connectivity
by Rob Janssen 25 Jul '19

25 Jul '19

> So that we don't have to have the same conversations twice -- let's move > this thread to 44ngn at mailman.ampr.org <https://mailman.ampr.org/mailman/listinfo/44net> I was against the creation of that second list, but I was too late to respond: it already was created. This inevitably creates such problems. I vote for deleting it. Now I have to find some way to use that new list, while using this one already is so difficult. With two lists it is becoming even more problematic... Rob

1 0

Re: [44net] 44.192.0.0/10 sale
by Rob Janssen 24 Jul '19

24 Jul '19

> >/Because we are trying to draft a new solution that would not work only for /> >/you, but also for others. You do not seem to be interested in that. / > Please quote me showing me saying I'm not interested in something new. > If there's something I *could* do where I don't have to increase my cost > not even $0.01/yr that's well documented I'd be quite happy to try > something out. Not once did I say I'm not interested, I have however > said don't scrap what's there that's working for me. The problem is that the new proposal will not work as smoothly and will be much more difficult to implement when the old system is to be left in place. That is because the new system is to be using automatic routing (with BGP or maybe OSPF when others can convince my that would work better), while the old system uses manual static routing managed by the portal. > >/Come on, it costs like $5-$10 per month per location to host such a service. / > Even this is still $5-10 per month more than I'm willing to spend. *you* are not supposed to be spending that money. It is to be spent by those that deploy the VPN servers. Hopefully ARDC. > >/And that is only when it is paid for. Last time I asked here for volunteers /> >/to host an echolink proxy farm, there were like 10 volunteers that would /> >/do (and did) it for free. It is likely that they would add such a VPN /> >/server feature to their already existing hosted system, if we would kindly /> >/ask it to them. / > Again, you're in the Netherlands, I am not. You most likely use 220v a/c > @50hz where we use 120v a/c at 60hz. Things are not the same here as > they are where you are and doubtful in other parts of the world as well. But NONE of those volunteers were from the Netherlands!! (they did not need to be, because I already host such a proxy farm in the Netherlands and the call for volunteers was to get more of them running across the world) The spread of volunteers was not as good as it could have been, but there were several from California, from Canada, from Australia, etc. Probably there are not so affected by backward internet connectivity (or lack of money) as you are. I don't know if they end up spending the $5/month for the benefit of amateur radio or if they got their hosting for free as part of some other deal. But they do offer the service! Voluntarily, without complaining. > Why would it be a waste of money? > I've wanted to see actual drafts and test environments where something new works. Because I think it is a waste of money (or a waste of time) to do a lot of research do potentially get a new product or another innovation started, for a community that is not interested in innovation and will make up any motivation to resist it. (including age, expense, "lack of reliability of the new system", and so on) You very well know that a system like the proposed one is operating in many other places, and the only proposed change is to interconnect those systems and make more of them, to replace the old and static system with something more flexible and more usable. > Why do I need internet connectivity from 44-net systems? It's a bonus > sure but I don't *need* it. Maybe you are mostly stuck in the usage scenario where AMPRnet is mostly used to emulate the old packet radio system, e.g. for BBS forwarding, DX cluster, telnet chat, etc. However, most new users who already know the internet want to do other things, like WebSDR, audio and video streaming (e.g. to YouTube), running amateur-oriented web services like a Mattermost server, etc. Internet connectivity is a must for most of these. > Often we forget many factors, some which we don't necessarily physically > see. No matter the solution, there will always be a very large amount of > points of failure. There's nothing you can do about that. But you can at least try! I remember seeing the questions here about having an IPIP gateway with more than one external address (e.g. using two different ISP links) to have redundancy at that level. Not possible, Sir! An IPIP tunnel endpoint has no way of seeing that the other side is down. With the newly proposed system this function would be no problem. BGP sees that a link is down and switches over to the alternate. Even within a second, when you configure BFD with it. But as long as the IPIP advocates get it their way, this will not happen. > One argument against IPIP however is it's deployment in the home. > Has anyone tried to simplify this? I have, and I've updated my system to > reflect the 2 subnets. All you do is set your device as the DMZ of your > router and run the install script which will ask for your 44-net info. I could deploy it to my home, but I have no need to do that. I already have a 5GHz WiFi link to our nearby (5 miles) access point so I get my AMPRnet over radio, thank you very much. And I have a VPN as a backup in case it somehow fails. In fact I have backup in 2 directions: when my VDSL connection fails, I can still surf the web via my amateur radio connection. If only to check the ISP website to see if there are any known interruptions. And there are several other stations here with the same setup. Not only do they have the above advantages, their systems also provide backup to the links between the access points. When the interlink to my accesspoint fails (another radio link), it would be automatically backed up via my own VPN and radio link. BGP does all that. The IPIP system can't. Rob

16 21

Re: [44net] Reverse DNS broken
by Cory (NQ1E) 24 Jul '19

24 Jul '19

Indeed, rdns for HamWAN is still broken (44.25.0.0/16). It appears to be because second level delegations were made on ARIN's DNS servers for every "44.x" label. Since that's at the same level as our delegation, DNS servers see it as an invalid "horizontal" delegation. Can you please see that the "25.44.in-addr.arpa." record is updated in ARIN's DNS servers to point to: a.ns.hamwan.net. b.ns.hamwan.net. Thank you, -Cory, NQ1E HamWAN $ dig +trace 25.44.in-addr.arpa. <snip> 44.in-addr.arpa. 86400 IN NS z.arin.net. 44.in-addr.arpa. 86400 IN NS x.arin.net. 44.in-addr.arpa. 86400 IN NS r.arin.net. ;; Received 92 bytes from 200.10.60.53#53(200.10.60.53) in 232 ms 25.44.in-addr.arpa. 86400 IN NS ns2.threshinc.com. 25.44.in-addr.arpa. 86400 IN NS ampr-dns.in-berlin.de. 25.44.in-addr.arpa. 86400 IN NS a.coreservers.uk. 25.44.in-addr.arpa. 86400 IN NS ampr.org. 25.44.in-addr.arpa. 86400 IN NS munnari.oz.au. ;; Received 181 bytes from 199.180.180.63#53(199.180.180.63) in 498 ms 25.44.in-addr.arpa. 3600 IN NS a.ns.hamwan.net. 25.44.in-addr.arpa. 3600 IN NS c.ns.hamwan.net. 25.44.in-addr.arpa. 3600 IN NS b.ns.hamwan.net. ;; BAD (HORIZONTAL) REFERRAL ;; Received 97 bytes from 192.109.42.4#53(192.109.42.4) in 269 ms 25.44.in-addr.arpa. 3600 IN SOA a.ns.hamwan.net. hostmaster.hamwan.net. 2019071706 900 180 604800 900 ;; Received 98 bytes from 44.24.245.2#53(44.24.245.2) in 11 ms On Fri, Jul 19, 2019 at 7:40 AM Tom Hayward via 44Net < 44net(a)mailman.ampr.org> wrote: > > > > ---------- Forwarded message ---------- > From: Tom Hayward <esarfl(a)gmail.com> > To: AMPRNet working group <44net(a)mailman.ampr.org> > Cc: > Bcc: > Date: Fri, 19 Jul 2019 07:38:57 -0700 > Subject: Re: [44net] Reverse DNS broken > Seems 44.25.0.0/16 is still broken. It was previously delegated to [abc]. > ns.hamwan.net. > > Tom KD7LXL > > On Fri, Jul 19, 2019, 05:07 Job Snijders <job(a)ntt.net> wrote: > > > I think this is resolved now. > > > > On Fri, Jul 19, 2019 at 02:37 Job Snijders <job(a)ntt.net> wrote: > > > > > Some good debugging information was provided to me offlist , I've now > > > been in touch with ARIN staff and some changes were made. It may take a > > > few hours for the changes to be visible on the Internet, new zone file > > > needs to be generated & pushed out. > > > > > > My initial assessment is that the delegations for the remaining /10 and > > > /9 towards AMPRnet DNS servers didnt exist. I don't know whose > > > responsiblity it would've been to ensure that existed. I can't guess > why > > > they were missing, perhaps a coordination issue in the transfer process > > > from the previous state to the current state. > > > > > > The AMPRnet reverse DNS administators may want to verify that the > > > authoritative dns servers are configured not for for 44.in-addr.arpa, > > > but for the individual /16s within 44/8 that still are AMPRnet. I don't > > > know who manages that so I hope this message finds them. > > > > > > I'm running on fumes and only have 4 hours of sleep opportunity ahead > of > > > me - so signing off, I think things will probably restore in the next > > > few hours. > > > > > > Good luck! > > > > > > Job > > > > > > On Fri, Jul 19, 2019 at 05:27:27AM +0000, Job Snijders wrote: > > > > To help speed up the resolution process, it would be beneficial if > > > > someone provides me with data what exactly is broken, and what the > > > > values / parameters previously were. > > > > > > > > was 44.in-addr.arpa. previously delegated to z.arin.net, x.arin.net, > > and > > > > r.arin.net? If not, where was it delegated? Where should it be > > > > delegated? > > > > > > > > Any tangible data about what the state currently is, and what it > should > > > > be; will help speed up recovery. > > > > > > > > Kind regards, > > > > > > > > Job > > > > > _________________________________________ > > 44Net mailing list > > 44Net(a)mailman.ampr.org > > https://mailman.ampr.org/mailman/listinfo/44net > > > > > > ---------- Forwarded message ---------- > From: Tom Hayward via 44Net <44net(a)mailman.ampr.org> > To: AMPRNet working group <44net(a)mailman.ampr.org> > Cc: Tom Hayward <esarfl(a)gmail.com> > Bcc: > Date: Fri, 19 Jul 2019 07:38:57 -0700 > Subject: Re: [44net] Reverse DNS broken > _________________________________________ > 44Net mailing list > 44Net(a)mailman.ampr.org > https://mailman.ampr.org/mailman/listinfo/44net >

3 4

Is this a worry ?
by gm4upx＠gb7jd.co.uk 24 Jul '19

24 Jul '19

< IP 0.0.0.0.5678 > 255.255.255.255.5678: UDP, length 106 > Hello, They say a little knowledge is dangerous ! At the moment at my tunl0 ( remote location ) I am getting groups of five of the above line every minute in my recently installed gateway. I have tried without success to stop it appearing using iptables with < -A INPUT -p udp -i tunl0 --sport 5678 -j DROP >. It shows as dropped when I monitor iptables but still appears when using tcpdump at the same time. It can be stopped by removing my ipencap entry but that stops my ampr-ripd reception was well. Is this just something I have to accept or is there a solution ? Regards, Ian..

2 3

New mailing list '44NGN' - was '44 net connectivity'
by Brian Kantor 23 Jul '19

23 Jul '19

As a few people have requested, I've created a new mailing list for discussions of the architecture of the Next Generation Network for AMPRNet, such as have been recently discussed here under the subject line of '44 net connectivity'. Feel free to join, and to invite your colleagues to join at <https://mailman.ampr.org/mailman/listinfo/44ngn> Welcome! - Brian

1 0

Re: [44net] 44.192.0.0/10 sale
by Rob Janssen 23 Jul '19

23 Jul '19

> >/And these addresses are normally included with your internet service for free. / > One thing to note however is that in many cases, any ports or protocols > blocked by the ISP on IPv4 are also blocked on IPv6. This is what makes > getting a block from a tunnel broker beneficial. Of course you will not need to have some unusual protocol like IPIP operational over that IPv6 connection, and usually the plain TCP and UDP ports can be opened without problem so it should be possible to run the usual services. A tunnel broker introduces an additional dependency and additional weird routing, so when people complain that using some VPN servers across the world instead of a tunnel mesh is objectionable, they certainly should not use an IPv6 tunnel broker! Rob

4 3

Re: [44net] 44 net connectivity
by Rob Janssen 23 Jul '19

23 Jul '19

> That is true, but only if we are the only ones using private ASNs. > But if happens that some people already uses private ASN for other > purpose, they would be in trouble connecting to 44net... because there > would be collisions. > I see no point of using public IP addresses and route them using private > ASNs. It may be that I do not understand BGP well. It is important to note that the routing in this overlay network, let's call it BGP44 from now, is running as a separate BGP instance that is not combined with BGP announcement to internet or private use of BGP. The BGP44 instance will only distribuite the routing info for AMPRnet subnets, similar to what is now in the IPIP routing table. Also, even though there could theoretically be collisions, in practice there is not much risk because there are 94,967,295 available AS numbers in the space that we use. That is 8 times more than the number of IP addresses we have left. Rob

3 2

Jump to page:

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

44net