It is working!!!
You're my hero
Please remind me how to make scripts executable.
Wireguard Session Manager - Discussion (2nd) thread
- Thread starter Martineau
- Start date
You're my hero
Please remind me how to make scripts executable.
ZebMcKayhan
Very Senior Member
Finally! Some progress at last!!!It is working!!!
You're my hero
Please remind me how to make scripts executable.
Code:
chmod +x /jffs/ ...etc....@Martineau perhaps 6-in-4 ipv6 should also enable rules and general wgm ipv6... it's just ppp tunnel much as WireGuard is. Basic setup works for static ips as shown by @pkoci
ZebMcKayhan
Very Senior Member
Looks like the package passes PREROUTING and never reaches FORWARD chain. Guess what happens in between: routing...
So ip route drops the package, altough there is a perfectly good route.
Must be some interface setting switch in the router somewhere. I'll dig and see what I can find during the evening.
Feel free to setup your wg clients, there should not be any problem forwarding packages to them, as this is your ultimate goal.
I'll let you know if I find anything.
This happens with every wgm restart. Does this mean something bad?
Btw. I have to run command "fc disable" (Broadcom Packet Flow Cache learning via BLOG disabled) to avoid terrible throughput and extreme CPU usage. Do I miss anything?
What speeds are expected on RT-AX56U?
Code:
Requesting WireGuard VPN Peer restart (wg21)
Restarting Wireguard 'server' Peer (wg21)
ip6tables: No chain/target/match by that name.
ip6tables: No chain/target/match by that name.
ip6tables: No chain/target/match by that name.
ip6tables: No chain/target/match by that name.
ip6tables: Bad rule (does a matching rule exist in that chain?).
ip6tables: Bad rule (does a matching rule exist in that chain?).
ip6tables: Bad rule (does a matching rule exist in that chain?).
ip6tables: Bad rule (does a matching rule exist in that chain?).
wireguard-server1: Wireguard VPN 'Server' Peer (wg21) on 10.50.1.0:11501 (# RT-AX56U (IPv4/IPv6) Server 1) TerminatedBtw. I have to run command "fc disable" (Broadcom Packet Flow Cache learning via BLOG disabled) to avoid terrible throughput and extreme CPU usage. Do I miss anything?
What speeds are expected on RT-AX56U?
ZebMcKayhan
Very Senior Member
No... it means that ip6tables rules are not found when trying to delete them. It means wgm is probably deleting 8 rules before your wg21-down.sh is executed so you could probably rename wg21-down as it is not needed. That would remove the errors.
Wgm does this for select models where reported needed... guess @Martineau will need to add AX-56U to the list.
No idea, I would expect around 450Mbit/s something...
ZebMcKayhan
Very Senior Member
you could try to add a new route, in case there is some flag in the routing entry that prevents this routing from ULA, this rule should take priority over your default route as it is more specific:
Code:
ip -6 route add 2::/3 via fe80::4255:82ff:febd:9c3 dev eth0Found some info here:
https://openwrt.org/docs/guide-user/network/ipv6/ipv6.nat6
now, I have no idea if it is relevant, but perhaps worth a try? They seem keen on setting up wan for temporary addresses by:
Code:
echo 2 > "/proc/sys/net/ipv6/conf/eth0/accept_ra"
echo 2 > "/proc/sys/net/ipv6/conf/eth0/use_tempaddr"It should do the same on asuswrt and I checked, I have these variables. Whatever set here is reset at reboot so no worries about that. But before doing so, perhaps check what they are before changing so you could set them back if it don't work:
Code:
cat /proc/sys/net/ipv6/conf/eth0/accept_ra
cat /proc/sys/net/ipv6/conf/eth0/use_tempaddrAnother interesting thing to try:
Wonder what happens if you replace your ula fd/fc prefix with something else, like fa or aa or just whatever.
Last edited:
Martineau
Part of the Furniture
I have uploaded
wireguard_manager Beta v4.16b3Flow Cache will be now be checked for model RT-AX56U and auto DISABLE Flow Cache
To apply the patch use
Code:
e = Exit Script [?]
E:Option ==> uf devMartineau
Part of the Furniture
I have uploaded
wireguard_manager Beta v4.16b3Update wg_manager.sh · MartineauUK/wireguard@426e8ee
CHANGE: Allow NVRAM 'ipv6_service' values '6in4' ('6to4' & '6rd' also added for completeness!) to be considered true IPv6 environments for ip6tables rule creatio...
github.com
Code:
e = Exit Script [?]
E:Option ==> uf devI Have tried each of the above
Code:
ip -6 route add 2::/3 via fe80::4255:82ff:febd:9c3 dev eth0
Code:
echo 2 > "/proc/sys/net/ipv6/conf/eth0/accept_ra"
echo 2 > "/proc/sys/net/ipv6/conf/eth0/use_tempaddr"None of these worked and the only solution that does work is using wan IPv6 address.
From a personal perspective this is not an issue as (a) my 'dynamic' IPV6 is not at all dynamic and (b) creating a new server + device + adding it to the phone takes no more than a couple of minutes. However, I am happy to keeping helping you find a solution for actual dynamic IPv6 if you are prepared to pursue this.
ZebMcKayhan
Very Senior Member
crap... I dont really find anything that would do this, and without having this myself, it is extremely hard to debug.
The only thing left could be to find the scope of the problem. meaning how much of you prefix could you change. while you accidently encountered the bug, you used a /48 prefix instead of your /64, so you changed 4 more numbers than your prefix allowed. this package got forwarded to eth0 successfully. so could you change more, like
2a02:c7f:: -->?
2a02:: -->?
to speed up testing, just check if the packages arrives out on eth0 (using tcpdump), but dont expect any reply. to create an ULA with 2 as prefix is risky as it could cause conflict with ips on the internet. but A.F.A.I.K only ips starting with 2 is used today, so if we could just change something:
2 = 0b0010 /4
so if the limit is /3 then perhaps 3 could be used as well as:
3 = 0b0011 /3
or can we change just one more bit:
4 = 0b0000 /2
best case, maybee even another:
6 = 0b0110 /1
finally:
a = 0b1010 (already checked this to not work)
so the full list to test would be something like:
2a02:c7f:: -->?
2a02:: -->? (/16)
2a00:: -->? (/12)
2af0:: --> ? (/8)
2000:: -->? (/4)
3000:: -->? (/3)
4000:: --> ? (/2)
6000:: --> ? (/1)
a000:: --> dont work /0
I know it is a lot of testing and will probably take like half an hour to go through. on the other hand, you could do by successive approximation, start in the middle (2000:
and if it works, continue down if not continue up. Hopefully we could find a prefix that we could use that is not used on the internet.whenever you found the smallest amount, try to use this with your masquarading rule (change the -s address accordingly) and hopefully it should work so you get a reply.
To finally resolve this, we probably need someone who has deeper knowledge on kernel routing than me.
Martineau
Part of the Furniture
So just to recap...
Your 'client' Peer is the Android SGS20?...if it is ..... is the Endpoint in the SGS20 defined as the explicit IPv6 address ofthe router, or are you using a DDNS name?
Given my post regarding Android using/preferring IPv4, can/have you tried say a Win laptop as the IPv6 client?
I have uploadedwireguard_managerBeta v4.16b3
To apply the patch use
Update wg_manager.sh · MartineauUK/wireguard@426e8ee
CHANGE: Allow NVRAM 'ipv6_service' values '6in4' ('6to4' & '6rd' also added for completeness!) to be considered true IPv6 environments for ip6tables rule creatio...
Code:e = Exit Script [?] E:Option ==> uf dev
Thank you for your work.
If I run fc disable, my internet speed is a way slower (around 300mbit down/100mbit up) than should be (1g down/100mbit up). Is this normal behavior? I've found that "fc disable" is somehow related to hw acceleration. Correct?
As noted above if the server is configured using assigned WAN address (the 2a02::...) address then IPv6 works fine. This issue is that it stops working with a ULA (non 2...) address. As my IPv6 is dynamic, rather then static, it would be preferable find how to use a ULA address.
The endpoint on the SGS20 is the DDNS_name: port. My challenge for (all suggestions welcome) is how best to test on a laptop. I mainly work from home and while I can add a client to a laptop, I don't want to be on the home LAN when testing - at least with the phone I can just disable wi-fi.
I could try tethering the laptop and phone (something new to learn), but that adds an extra layer of routing. I will give a try later.
Last edited:
ZebMcKayhan
Very Senior Member
Agreed, but you dont HAVE to use an ULA. You only need something that you know never will conflict with anything online. So in theory you could borrow a portion of someone else static ipv6 that you know you will never contact online. You will anyway masquarade it so it wont be used online.
The problem is that if I.e. Google start using the same address as you choose, any lan clients wanting to visit google sites will end up at wg21, were no such services exist.
If you could find an address space not used today it would be the second best thing.
Martineau
Part of the Furniture
I would hope that forcing the Endpoint on the SGS20 to the actual IPv6 2a02:: address, should then allow ULA addresses to be statically used by the 'server' Peer and its 'client' Peers...albeit with the ugly IPv6 NAT rule.
As currently enabled (and I have no idea if it will change) this is normal. As I understand it, If you want to run the more secure VPN options on the router (not PP2P or Instant Guard) then you have a trade off between more resource heavy solution such as OpenVPN which can work with hardware acceleration and resource light WireGuard that cannot. @RMerlin is the best person to advise here.
Which will give the best throughput is down to the router. For me I would like to work with WireGuard as it seems to (should) deal with IPv6 better than OpenVPN. If you want to run either VPN 'flat out' with a 1g link without running it on your PCs / laptops, etc, then you will need to move it onto a device with a lot more raw power.
As I have Cake enabled (which also disables HW acceleration) and a much slower ISP (80/20) I had not checked whether wgm was looking at runner/flow cache. I have just switched off Cake and restarted a server instance and noted that both (runner/flow cache) are showing as enabled in Sysinfo. While the memory usage seems higher, all CPU stats are still very low, so it doesn't seem to be an issue (yet).
Martineau
Part of the Furniture
Already has.... December 2021
and I can personally verify that when I loaded ASUS 386 Beta RC3-3 Public Beta onto my RT-AX86U it was indeed disabled when WireGuard was configured, but cannot vouch for other AX models such as your RT-AX88U
Last edited:
Martineau
Part of the Furniture
ZebMcKayhan
Very Senior Member
@Odkrys Who made the initial compilation put in xmark rules in the firewall to bypass hardware acceleration only for wireguard packages. These seem to work on ac86u and ax88u so we can run with hw acceleration enabled. But they don't seem to work well on newer models. I don't know where/how @Odkrys got these marks and if there are other marks that works for the newer models but none that this community seems to be aware of.
Guess it would be possible to setup a script to brute-force test through marks and write in a log which seems to atleast get rid of syslog events but who knows what some marks do...
ZebMcKayhan
Very Senior Member
I clearly meant a friend whom you could ask... you will only be using it internally in your network, not externally.
Or ask/pay for a static IP from your ISP?
Not all ISP offer static ipv6, sadly.
Similar threads
Similar threads
Similar threads
-
New Addon - ChatAI - have a chat session with Googles Gemini AI Bot- Started by JGrana
- Replies: 2
-
connmon Making config changes invokes password manager in MacOS
- Started by Jeff-
- Replies: 4
-
VPNMON VPNMON-R3 - Early, early alpha discussion, progress & screenshots- Started by Viktor Jaep
- Replies: 17
Latest threads
-
-
-
RT-AC87U 5GHz frequently dropping after latest signature update
- Started by bibbis
- Replies: 0
-
-
Yet another question about VLAN w/ RT-AX88U Pro in AP mode.
- Started by mbze430
- Replies: 1
Sign Up For SNBForums Daily Digest
Get an update of what's new every day delivered to your mailbox. Sign up here!