kellytk: Is anyone aware of a reason why my ISP-provided wireless router (192.168.254.254) would be showing up in the HTTP access log of a computer on the LAN? The request to / is every 1-2 minutes mercutio: hmm irc is lagging a LOT
and then it's fine again
http://hackaday.com/2015/08/31/fcc-introduces-rules-banning-wifi-router-firmware-modification/
i wonder how that will play out
it may mean linux is illegal on routers?
well until firmware blobs are built into wireless cards.
oh, it's still proposal
that made it look final
https://apps.fcc.gov/kdb/GetAttachment.html?id=1UiSJRK869RsyQddPi5hpw%3D%3D&desc=594280%20D02%20U-NII%20Device%20Security%20v01r02&tracking_number=39498
erk sounds like it is law already milki: no, i think its still in proposal phase, comments link to a request for comments on the "proposal"
https://www.federalregister.gov/articles/2015/08/06/2015-18402/equipment-authorization-and-electronic-labeling-for-wireless-devices
This document has a comment period that ends in 7 days (09/08/2015) m0unds: well, it seems like they're not banning linux on routers but banning modification of wlan radios
makes sense considering how often people use other countries' regulatory domain to configure power levels beyond allowable limits
oh well
can't be bothered to care about it brycec: Right it doesn't ban Linux, but does make open-source drivers more difficult. We'll see a return to the Atheros "HAL" style I guess. m0unds: yep, probably mercutio: milki: yeh it got extended it looks like
cos deadlines forcomments were 16th august
and there are comments since then
i wonder what fsf has tos ay about that heh
i don't really see how controlling this stuff is meanet to work
what if i take my cellphone to the US?
what about laptop or old device
it seems better to me if they just be proactive with dealing with infringers. ***: pjs has quit IRC (*.net *.split)
eryc has quit IRC (*.net *.split)
tellnes has quit IRC (*.net *.split)
mjp_ has quit IRC (*.net *.split)
trobotham has quit IRC (*.net *.split)
tellnes has joined #arpnetworks
mjp_ has joined #arpnetworks
pjs has joined #arpnetworks
eryc has joined #arpnetworks kellytk: Does anyone have experience with a DSL router making HTTP requests for "/" to IPs on the LAN every 1-2 minutes?
I've only noticed it after setting up a web server on the private LAN ***: trobotham has joined #arpnetworks m0unds: mercutio: how on earth would they address people misusing low power unlicensed radio spectrum? mercutio: m0unds: the same way they do now m0unds: haha, okay mercutio: they'd listen for radio interference and move closer to the source m0unds: right, and it would require them to sit in a truck looking for people infringing on unlicensed stuff
vs just making it harder to abuse it
if it was licensed stuff and high power transmission, that'd be one thing mercutio: yeah well usually it's in response to complaints from things like weather radar stuff m0unds: but when the unlicensed bands include things like xbox controllers, wireless mice, etc, it's a little harder to deal with mercutio: that stuff is 2.4 ghz normally ***: pjs has quit IRC (*.net *.split)
eryc has quit IRC (*.net *.split) mercutio: it's 5ish ghz stuff they're cracking down on m0unds: right, but it's still a cesspool mercutio: i'm not sure it's indoors stuff they're relaly trying to deal with though m0unds: and there still was nowhere near enough spectrum allocated for unlicensed stuff mercutio: so much as stuff that goes outdoors too
new zealand's just allocated more unlicensed spectrum m0unds: yes, lots of shitty WISPs use unlicensed 5ghz
and it's a mess mercutio: like 3 ghz or something
yeah wisp's will be able to use 3 ghz here now too
i can't remember the exact frequency
maybe it is 3ghz
cos fcc say they may add 3 ghz too
oh fcc say they may hadd 3.65 to 3.7 ghz
for wisp type stuff i think licensed spectrum is a good idea
and that the cost of doing licensed should be minimised. ***: pjs has joined #arpnetworks
eryc has joined #arpnetworks m0unds: so long as the licensed gear is more expensive and the licensing cost is >$0, most of the shitty wisps i'm thinking of would probably stick to unlicensed stuff kellytk: So a call to the ISP who has provided the DSL/wifi router yielded no clue unsurprisingly
Time to write a custom pf ruleset for LAN usage! mercutio: i've heard of attacks on routers from web sites that try to pull stuff off local network to exploit m0unds: kellytk: what's the mfgr of the router/modem thing? kellytk: Netgear
mercutio: I had two ideas, either the router was configured to monitor customers to prevent services being ran (unlikely and better controlled from the upstream) and secondly some shenanigans afoot
Either way, it's a gross display mercutio: windows often shows names of routers, maybe related. kellytk: Thank you m0unds: yea, was thinking discovery of some sort mercutio: http://blog.multipath-tcp.org/blog/html/2015/07/24/korea.html
that's actually kind of cool kellytk: It reminds me of modem shotgunning from days of yore
Nice link, thanks mercutio mercutio: now i am curious to try this on wireless heh
it's a pity have to use two wireless cards to hit two routers. ***: jpalmer has joined #arpnetworks
jpalmer_ has joined #arpnetworks grody: mercutio, someone else told me it bodged their bootloader back to windows too mercutio: i am using uefi grody: maybe 'cause these last two were windows default (one 7 one 8), then resized and linux..... ah
one is UEFI type (does secure boot etc), other is just BIOS
hmmm...
i've tried forcing androids (back in 2.x days) to try stay on 3G when on WIFI and load balance m0unds: i think the radio status change on android causes it to wipe out your default route for data
at least that was how it was the last time i looked at it
wifi takes preference over viable carrier data
i think adb dumpstate connectivity will give you an idea of how the device sees it mercutio: http://arstechnica.co.uk/information-technology/2015/09/microsoft-accused-of-adding-spy-features-to-windows-7-8/
the plot thickens. m0unds: i'm kind of over it tbh mercutio: i might try going back to linux fill time m0unds: go for it mercutio: err full
had video driver issues in the past
but maybe it's better now m0unds: doubt it, but that's the pessimist in me mercutio: heh
does radeon do good at spinning fans down?
that was my main annoyance m0unds: no idea mercutio: my full time linux box has fanless video card to fix it :) m0unds: i don't use linux on the desktop because i don't have patience to fix stuff mercutio: for me it's mostly cos games and web browsing work better on windows brycec: It did not botch my MBR dual-boot, fwiw.
(re: Win10) mercutio: yeah they seemed to get better with mbr brycec: All my UEFI installs of Win10 were standalone, so I can't comment on that :p mercutio: well it just removes the uefi boot option
you can add the uefi back in. with windows it seems that it can be detected?
with linux it seems you have to add. so it may be i'm doing something wrong.
like - if i stick a ssd in a different computer with uefi it won't just boot
i need to add boot option in
have you done uefi installs with arch? brycec: Right. But that's by design
Yes I have,several times mercutio: and it goes away if you move ssd to a different computer giht?
right
and you need to boot off stick and put it back in brycec: Some OS work around that "limitation" by installing their bootloader to the default shellx64.efi location mercutio: oh
i suppose i could install shellx64.efi brycec: It's a dirty hack, but an effective one all the same. mercutio: and get a shell then
that's good enough to fix it :) brycec: You could. or just copy grub.efi to shellx64.efi :P mercutio: heh brycec: (There's a grub-install switch to install as shellx64.efi in fact) mercutio: sweet
yeh could do that
i'm wondering if i should even still use grub
but i suppose it works
thinking about trying the linux create uefi stub thingy
and avoid the whole bootloader thing completely
and just select boot device by pressing f12 or whatever brycec: Personally I recommend refind
mercutio: https://wiki.archlinux.org/index.php/GRUB#UEFI_firmware_workaround it's bootx64 not shellx64 apparently mercutio: yeh i probably should have checked arch linux wiki
i keep forgetting that it's good brycec: lolol mercutio: been using linux so long i'm used to finding it hard to find good answers brycec: (no that doesn't mention the grub-install switch. And maybe I was even thinking of refind having such a switch) mercutio: oh that's stated as a workaround
not to fix the disappearing if moving to different computer issue
"grub-install automatically tries to create a menu entry in the boot manager."
so yeah that's the primary issue brycec: i.e. it calls to efibootmgr mercutio: because that's per system not per hard-disk
and i like the freedom to move ssd/hard-disk between machines
(and for windows to not screw with the boot list)
hmm maybe i can install shell straight from windows brycec: shellx64? You should be able to, just copy into place... mercutio: not on windows
windows doesn't mount it as a drive brycec: More of an issue getting to it in the first place
So... mount it as a drive? (Assign it a letter) mercutio: i dunno if that works
yeah you can't
oh looks like there may be a way
woot, mountvol b: /s mnathani_: are we mounting zfs partitions on windows as a drive? mercutio: how would you do that? mnathani_: I didnt think it was possible mercutio: i'm trying to boot efi shell :( mnathani_: oh ok mercutio: "launch efi shell from filesystem device" - "not found"
i gave up and used usb stick
seem to be capped at 24 gigabit for network speed for some reason
still a lot better than the 8 gigabit i'm getting out of windows
even linux to linux i'm still getting synergy issues in one direction
maybe it was linux that was to blame
https://github.com/synergy/synergy/issues/4735
it's not just me, apparently lots of combinations can't paste from client to server mnathani_: On October 15, SoftLayer will change the reverse DNS addresses for servers on our network from xxx.reverse.softlayer.com to xxx.softlayer-reverse.com. If you have hard-coded your servers' default SoftLayer reverse DNS entries in any of your applications or systems, your code will need to be updated to reflect the new reverse DNS entries.
is it just me or does softlayer-reverse.com not seem to be a registered domain
lol : YES! YOUR DOMAIN IS AVAILABLE. BUY IT BEFORE SOMEONE ELSE DOES.
up_the_irons: I was wondering when the invoice would show up mercutio: wow
i think you're right mnathani
although i still think it's pretty rude ;)
you could just email them abou it up_the_irons: lol let's register it ;) mike-burns: Haa! mnathani_: :-) m0unds: hahaha
resisting the urge kellytk: "block in on em0: 199.249.120.1 > 192.168.254.100: ip-proto-17"
What is "ip-proto-17"? m0unds: @google protocol 17 BryceBot: 32,600,000 total results returned for 'protocol 17', here's 3
List of IP protocol numbers - Wikipedia, the free encyclopedia (https://en.wikipedia.org/wiki/List_of_IP_protocol_numbers) 23, 0x17, TRUNK-1, Trunk-1. 24, 0x18, TRUNK-2, Trunk-2. 25, 0x19, LEAF-1, Leaf-1. 26, 0x1A, LEAF-2, Leaf-2. 27, 0x1B, RDP, Reliable Datagram Protocol ...
Protocol Numbers (http://www.iana.org/assignments/protocol-numbers) Aug 25, 2015 ... In the Internet Protocol version 4 (IPv4) [RFC791] there is a field called "Protocol" to identify ... 17, UDP, User Datagram, [RFC768][Jon_Postel].
User Datagram Protocol - Wikipedia, the free encyclopedia (https://en.wikipedia.org/wiki/User_Datagram_Protocol) The User Datagram Protocol (UDP) is one of the core members of the Internet .... The protocol is that for UDP (see List of IP protocol numbers): 17 (0x11). kellytk: I searched before asking of course. There's surprisingly little info. It seems related to fragmented packets/DDoS m0unds: udp. kellytk: It's coming from only one, two max hosts though m0unds: ip protocol is udp
what udp traffic is it
er 17
figure out what traffic it is using wireshark or something and maybe you can figure out what specifically it is
otherwise, it's just...udp mike-burns: UDP is a refreshingly short RFC. m0unds: yes
i like udp mercutio: receiving random udp packets is pretty common kellytk: http://pastebin.com/thYvsH6H
30 and 258 m0unds: uhh
that's dns
at least the udp i can be bothered to look at
haha, friend sent me a link to some "avoid people" site that looks for places on foursquare or whatever that nobody checks into
most of the ones within 10 mins of my house are other people's houses. guess they added them to check in to and then stopped or never did kellytk: Nice app idea mercutio: terrible app idea i reckon :)
it promotes less popular places to make them popular
even if that's some random guys house.
are people generally using gpt with freebsd these days?
as opposed to bsd disk slices
with zfs i suppose you don't even notice which it is kellytk: Well, yes I see your point. What would cause a ping to the router to result in "64 bytes from 192.168.254.254: icmp_seq=5 ttl=64 time=2.213 ms (DUP!)"? I've not seen DUP! before mercutio: i've seen dup heaps of times BryceBot: That's what she said!! kellytk: I use GPT however I only run VPSs at present, so ZFS isn't an option mercutio: it means you're receiving the same ping data back in a packet that's already been received kellytk: What would cause that? mercutio: switch loop is an easy way to create it
but with your router it may be something silly like pinging broadcast address kellytk: One virtual server can ping the router IP with no issues, the other gives the DUP! message. They are both running the same pf.conf ruleset mercutio: it could also be due to inprecise clock
it may be that the mac is the same on both the vm's
and it's sending data two both vm's
or it may be because you're pinging both at once kellytk: I've verified their MACs differ mercutio: try pinging the host then kellytk: What?
This is such a strange issue. I thought it was caused by bringing up pf, but it's not
Even when I stop pf, the DUP! issue remains
What's strange is the guest which gets the DUP! only gets that for pinging the router, none of the other computers on the LAN
Interesting, DUP! is also caused by pinging google.com
It may have started when I changed my host system to use a static IP vs the router's DHCP
Can anyone see any obvious problems with the pf.conf ruleset http://pastebin.com/TkYyTrMc freebsd-update fetch is failing
With the pf service stopped, the fetch functions correctly
I assumed allowing out on 80 would allow in for the same connection, apparently both out and in 80 needed to be allowed
I figured out the DUP! issue mercutio. I need to restart Parallels after changing my workstation's IP to static from DHCP