Didn't make it to 3 days. This morning the Windows PC's connected through the VPN had no internet. Their little WiFi status indicators in the task bar said they had internet connection, but nothing could be browsed. The Windows PC's that are outside the VPN (policy based routing) could get to google.com to do searches, but every other web site I tried could not be browsed. The router has Keep Alive enabled and set to 8.8.8.8 i.e google, the router did not automatically reset.
Did a manual router reset through its web interface, now all PC's can access the internet again.
EDIT - back to square 1. The erase nvram and manual reconfigure were a waste of time
The VPN is hosed again, TUN/TAP write bytes are 0. PC's routed through the VPN have no internet. PC's not routed through the VPN are fine. Router can ping 8.8.8.8, so the Keep Alive is not triggering.
Router: Netgear AC1450
Firmware: 31520M
Kernel: Linux DD-WRT 4.4.52 #255 SMP Wed Mar 1 23:58:41 CET 2017 armv7l DD-WRT
Status: Works with restore / no reset
Errors: With 30-30-30 reset and manual setup, several issues arise
When I upgraded from 31205 (which was upgraded from 30700 etc down the line), things work. However I have effectively no space left for nvram and so, I tried resetting and entering my settings manually with the webgui in this build. That's where things go bad. 3 issues:
1) Something simple: OpenVPN server doesn't recognize that I've entered a certificate / doesn't add the required option to the config file. This is easily worked around by adding this to the custom configuration:
cert /tmp/openvpn/cert.pem
2) It appears the PPPoE dual is not working correctly, as there is no route to the outside router (on 192.168.5.0/24), and it is indeed unreachable. It seems to be forgetting the 'dual' part of PPPoE dual.
3) Last issue I have no idea what's wrong:
I have a slightly tricky setup where I have a VAP (br1, 192.168.9.0/24) over over PPP (router is a gateway using PPPoE dual). While on the default network (br0, 192.168.10.0/24) routes all traffic across openvpn (client). This allows network clients to change WAN IP by simply choosing a different wifi network (although both networks can talk to each other).
Sadly, with my rebuilt setup, this fails, even though I can't find anything wrong with the settings. While the default wifi network can connect to the internet over the VPN, clients on the VAP CANNOT. The router has full connectivity and I can ping out over the (default) ppp interface from the router. When I restore from the older backups, things work again. I've tried disabling the firewall to rule that out; no change.
Here are the pertinent settings:
Code:
root@DD-WRT:~# ip rule show
0: from all lookup local
2000: from 192.168.10.0/24 lookup 200
2001: from 192.168.10.0/24 lookup 201
32766: from all lookup main
32767: from all lookup default
root@DD-WRT:~# ip route show table 200
default via 10.8.0.4 dev tun1
192.168.9.0/24 dev br1 scope link
192.168.10.0/24 dev br0 scope link
192.168.12.0/24 dev vlan2 scope link
root@DD-WRT:~# ip route show table 201
unreachable default
root@DD-WRT:~# ip route show table main
default via 109.xxx.xxx.1 dev ppp0
10.7.0.0/24 dev tun0 proto kernel scope link src 10.7.0.1
10.8.0.0/24 dev tun1 proto kernel scope link src 10.8.0.4
109.xxx.xxx.1 dev ppp0 proto kernel scope link src 109.xxx.xxx.59
127.0.0.0/8 dev lo scope link
169.254.0.0/16 dev br0 proto kernel scope link src 169.254.255.1
192.168.9.0/24 dev br1 proto kernel scope link src 192.168.9.1
192.168.10.0/24 dev br0 proto kernel scope link src 192.168.10.1
10.7.0.1 is the openvpn server tunnel
10.8.0.1 is the openvpn client tunnel (it uses route-nopull to avoid importing routing)
table 201 is there to ensure there are no leaks in case tun1 (the vpn client tunnel) goes down.
In short, wifi clients on br1 192.168.9.x cannot ping the ppp gateway 109.xxx.xxx.1, even though the router can.
Any help appreciated (also, the main reason I've been upgrading is I get wireless dropouts for a couple minutes every few hours since I added the VAPs, and not sure of the cause, my latest guess is wifi key renegotiation).
Router: Buffalo WXR-1900DHP
Firmware: v3.0-r31520M kongac (03/01/17)
Kernel: Linux 4.4.52 #255 SMP Wed Mar 1 23:58:41 CET 2017 armv7l
Status: online
Upgraded: From r31205M via ddup --flash-latest
Reset: No
Errors: Yes, just like last update had.
Firmware update fails to flash the second image.
Multiple processes attempting to operate wireless signals on both bands.
Details:
Code:
root@WXR-1900DHP:~# ddup --flash-latest
New release 31520 for Buffalo WXR-1900DHP available.
Do you want to update (y/n) [default=n]:
y
Connecting to www.desipro.de (212.227.247.157:80)
fw.bin 100% |**********************************************| 22596k 0:00:00 ETA
Connecting to www.desipro.de (212.227.247.157:80)
fw.bin.sig 100% |*********************************************| 256 0:00:00 ETA
Signature for http://www.desipro.de/ddwrt/K3-AC-Arm/dd-wrt.v24-K3_AC_ARM_STD.bin ok.
Are you sure you want to proceeed (y/n) [default=n]:
y
Closing ssh connections now. Flashing now...
--no connection after a couple minutes of waiting--
--connected serial cable--
Code:
Device eth0: hwaddr DE-FB-02-29-AB-E8, ipaddr 192.168.11.1, mask 255.255.255.0
gateway not set, nameserver not set
============================================
tmpr current 58 (289)
tmpr normal_thresh=98 bootable_thresh=115
current 58 < thresh 116, booting FW...
============================================
FW-A / FW-B header are different
Checking recover image
check_trx: start nflash1.trx_recover
check_trx: exit nflash1.trx_recover
Loader:raw Filesys:raw Dev:nflash0.os_recover File: Options:(null)
Loading: .. 1632728 bytes read
Entry at 0x00008000
Closing network.
Starting program at 0x00008000
Uncompressing Linux... done, booting the kernel.
--Panic a bit later and reboots--
Code:
Kernel panic - not syncing: Fatal exception
CPU1: stopping
CPU: 1 PID: 625 Comm: startservice Tainted: P D 4.4.45 #248
Hardware name: Northstar Prototype
Fixed by flashing recovery manually to match main:
Code:
CFE> flash nflash1.trx nflash1.trx_recovery
Reading nflash1.trx: Done. 32376832 bytes read
Programming...done. 32376832 bytes written
*** command status = 0
CFE> go
============================================
tmpr current 60 (286)
tmpr normal_thresh=98 bootable_thresh=115
current 60 < thresh 116, booting FW...
============================================
FW-A / FW-B header are same
Checking normal image
check_trx: start nflash1.trx
check_trx: exit nflash1.trx
Loader:raw Filesys:raw Dev:nflash0.os File: Options:(null)
Loading: .. 1632992 bytes read
Entry at 0x00008000
Closing network.
Starting program at 0x00008000
Uncompressing Linux... done, booting the kernel.
After boot you see this in ps output, which also causes WiFi to go up/down constantly as the computer seems to switch between the 2 processes for the same SSID, the second process has no BSSID (based on what OSX sees) so I could see BSSID going to all zeros and back to a real value until I killed the second pair of processes. I have modified the ps output to remove my password and actual wifi name. This happens at every reboot.
output from PS shows multiple instances of WiFi running ??
Code:
1559 root 1508 S nas -P /tmp/nas.wl0lan.pid -H 34954 -l br0 -i eth1 -A -m 128 -k hahahahahah -s Cool SSID Bro - 2.4G -w 4 -g 3600
1562 root 1508 S nas -P /tmp/nas.wl1lan.pid -H 34954 -l br0 -i eth2 -A -m 128 -k hahahahahah -s Cool SSID Bro - 5G -w 4 -g 3600
1629 root 1508 S nas -P /tmp/nas.wl0lan.pid -H 34954 -l br0 -i eth1 -A -m 128 -k hahahahahah -s Cool SSID Bro - 2.4G -w 4 -g 3600
1652 root 1508 S nas -P /tmp/nas.wl1lan.pid -H 34954 -l br0 -i eth2 -A -m 128 -k hahahahahah -s Cool SSID Bro - 5G -w 4 -g 3600
Checking the pid files shows that 2 of these aren't supposed to be running....
After killing the second errant processes the WiFi works fine
Code:
root@WXR-1900DHP:~# cat /tmp/nas.wl0lan.pid
1629
root@WXR-1900DHP:~# cat /tmp/nas.wl1lan.pid
1652
root@WXR-1900DHP:~# kill 1559
root@WXR-1900DHP:~# kill 1559
sh: can't kill pid 1559: No such process
root@WXR-1900DHP:~# kill 1562
root@WXR-1900DHP:~# kill 1562
root@WXR-1900DHP:~# kill -9 1562
root@WXR-1900DHP:~# kill -9 1562
sh: can't kill pid 1562: No such process
root@WXR-1900DHP:~# cat /tmp/nas.wl1lan.pid
1652
_________________ Routers:
WXR-1900DHP - Active (main) - v3.0-r36070M kongac (05/31/18 )
WZR-N600DHP - Wired AP - v3.0-r33679 BS (11/04/17)
WNDR-3400 - retired to its box for several years
good to know... I was trying to avoid it due to my complicated setup, I'll report back on whether it did the trick
I would be interesting to be able to import and export settings from previous versions... of course not everything would work but if is reported what could be imported it would tell the user what to configure again... but i guess it's a lot of work for something that is only used occasionally
The scripts only work if they are updated as well all the time... which they are not, oh well...
Anyway I erased nvram and ran into the following issues:
- Running OpenVPN Daemon now requires to add the server cert into the config file (previously it was auto generated)
- 2.4Ghz Radio still always occupies 2 channels (regardless of the 20Mhz setting)
- openvpn complains about group accessibility on ta.key (i think it has always done this)
I will monitor Wifi stability for a while (I don't have too much faith in my laptops wifi card either so it will be hard to debug)
Edit: I'm using a guest network as well on the 2.4 band could that be the reason for multiple channels?
wifi instability seems to be strongly related to the 5Ghz network... which explains why i dont experience it on other networks.
hey, what happened? is my poor hdd dying?^^ I just copied a 30gb file over samba...
asus ac68u DD-WRT v3.0-r31520M
I got the similar issue with this DD-WRT version on my 120GB USB 3.0 pen drive formatted as an exFAT partition. But this is the first time I start to use the USB3 port to attach a faster drive for samba / DLNA sharing.
It could be kernel xHCI driver stability issue when an IO error happens. Somehow the device won't be reconnected until a system reboot action is performed.
Another USB2 port is attached with a 16GB pen drive basically for some customized setting storage for a long time. So it never does so huge amount of burst IO request at such short period.
--EDITED--
It can be easily duplicated again. The kernel log shows following messages:
Code:
xhci_hcd 0000:00:0b.0: Timeout while waiting for setup device command
xhci_hcd 0000:00:0b.0: Timeout while waiting for setup device command
usb 4-1: device not accepting address 2, error -62
xhci_hcd 0000:00:0b.0: Timeout while waiting for stop endpoint command
xhci_hcd 0000:00:0b.0: Cannot set link state.
usb usb4-port1: cannot disable (err = -32)
xhci_hcd 0000:00:0b.0: Cannot set link state.
usb usb4-port1: cannot disable (err = -32)
usb 4-1: USB disconnect, device number 2
sd 1:0:0:0: [sdb] tag#0 UNKNOWN(0x2003) Result: hostbyte=0x01 driverbyte=0x00
sd 1:0:0:0: [sdb] tag#0 CDB: opcode=0x2a 2a 00 01 65 dc 40 00 00 40 00
blk_update_request: I/O error, dev sdb, sector 23452736
sd 1:0:0:0: [sdb] tag#0 UNKNOWN(0x2003) Result: hostbyte=0x01 driverbyte=0x00
sd 1:0:0:0: [sdb] tag#0 CDB: opcode=0x2a 2a 00 01 65 dd 00 00 03 c0 00
blk_update_request: I/O error, dev sdb, sector 23452928
hotplug (18510): drop_caches: 3
[EXFAT] trying to unmount...
[EXFAT] unmounted successfully
VFS: Dirty inode writeback failed for block device sdb (err=-5).
xhci_hcd 0000:00:0b.0: Command completion event does not match command
_________________ Buffalo WZR-1750DHP
Last edited by pie1394 on Mon Mar 13, 2017 11:29; edited 1 time in total
The scripts only work if they are updated as well all the time... which they are not, oh well...
Anyway I erased nvram and ran into the following issues:
- Running OpenVPN Daemon now requires to add the server cert into the config file (previously it was auto generated)
- 2.4Ghz Radio still always occupies 2 channels (regardless of the 20Mhz setting)
- openvpn complains about group accessibility on ta.key (i think it has always done this)
I will monitor Wifi stability for a while (I don't have too much faith in my laptops wifi card either so it will be hard to debug)
Edit: I'm using a guest network as well on the 2.4 band could that be the reason for multiple channels?
wifi instability seems to be strongly related to the 5Ghz network... which explains why i dont experience it on other networks.
- When setting the 2.4Ghz Radio to 40 Mhz it starts using 4 channels... not what i expected but at least it reacts to settings now, maybe i was just wrong about what 20 mhz means :p
- I disabled beamforming, airtime fairness and ran only one wlan at the time... and now everything seems stable (including the 5ghz)
I'll start increasing the complexity of the wifi setup until it starts failing again, hopefully it will reveal the issue
2) It appears the PPPoE dual is not working correctly, as there is no route to the outside router (on 192.168.5.0/24), and it is indeed unreachable. It seems to be forgetting the 'dual' part of PPPoE dual.
It turns out that this is because in my rebuilt setup, I chose to use DHCP (the default setting), but this does not work with PPPoE. In my original working setup (before I rebuilt by hand), this was the case also, but in fact DD-WRT was using the static IP that I must have entered at some point in the past (even though DHCP was selected).
The short story is that since at least 30700, DHCP with PPPoE dual seems not to be working, and you have to enter a static IP/netmask/gateway to be sure the routing will work properly. Your static IP info is used even if you (re-)enable DHCP. (ie, these are bugs to my mind)
jtbr wrote:
3) Last issue I have no idea what's wrong:
I have a slightly tricky setup where I have a VAP (br1, 192.168.9.0/24) over over PPP (router is a gateway using PPPoE dual). While on the default network (br0, 192.168.10.0/24) routes all traffic across openvpn (client). This allows network clients to change WAN IP by simply choosing a different wifi network (although both networks can talk to each other).
Sadly, with my rebuilt setup, this fails.
[details omitted...]
In short, wifi clients on br1 192.168.9.x cannot ping the ppp gateway 109.xxx.xxx.1, even though the router can, and they use the main routing table.
After rebuilding by hand with 30700, and correcting the above issue with PPPoE (which should not have affected this so far as I can tell), it works. So it seems to be some issue that is new since 30700, the version I have reverted to. I don't seem to recall being asked the IP address/netmask when I created br1 in the web gui in 31520. So if that is absent and thus causing the bridge to not operate correctly, that would explain the problem.
No issues on two of my R7000's after 9 days. _________________ 2 R7800's (Kong OpenWRT 23.05) as access points
X86_64 build PC (DDWRT Build r55630)
3 R7000's (DDWRT build r53843)using one as a bridge currently
Netgear WNDR 3400 (DDWRT r55630)
Netgear DGND 3700 (with OpenWRT)
R7000: PC's on 2.4GHz wireless. Some PC's are accessing internet through VPN (policy based routing), some are not. All was running ok since the last reboot, a couple of days ago.
This morning, a PC on wireless with IP outside the VPN e.g. 192.168.0.100, can ping a PC on wireless with IP inside the VPN using the local IP e.g. 192.168.0.200, but can't browse files or connect VNC to the PC on 192.168.0.200. That was working last night. Probably a reboot will fix it, but as I am not home, I'll wait until I am home in case the R7000 does not recover.
Joined: 16 Nov 2015 Posts: 6440 Location: UK, London, just across the river..
Posted: Mon Mar 13, 2017 22:30 Post subject:
R7000 after 10 days it locked in, no GUI no SSh no internet,
and no connection to WI-FI...on LAN side no internet and connection to the router....
I had to restart it, but even thou same condition...
than erased nvram and reboot & rebuild manually all its working. _________________ Atheros
TP-Link WR740Nv1 ---DD-WRT 55630 WAP
TP-Link WR1043NDv2 -DD-WRT 55723 Gateway/DoT,Forced DNS,Ad-Block,Firewall,x4VLAN,VPN
TP-Link WR1043NDv2 -Gargoyle OS 1.15.x AP,DNS,QoS,Quotas
Qualcomm-Atheros
Netgear XR500 --DD-WRT 55779 Gateway/DoH,Forced DNS,AP Isolation,4VLAN,Ad-Block,Firewall,Vanilla
Netgear R7800 --DD-WRT 55819 Gateway/DoT,AD-Block,Forced DNS,AP&Net Isolation,x3VLAN,Firewall,Vanilla
Netgear R9000 --DD-WRT 55779 Gateway/DoT,AD-Block,AP Isolation,Firewall,Forced DNS,x2VLAN,Vanilla
Broadcom
Netgear R7000 --DD-WRT 55460 Gateway/SmartDNS/DoH,AD-Block,Firewall,Forced DNS,x3VLAN,VPN
NOT USING 5Ghz ANYWHERE
------------------------------------------------------
Stubby DNS over TLS I DNSCrypt v2 by mac913
I have radio scheduled to turn off radio at night. Today it didnt turn on at 6am as scheduled, did anyone face this issue on this build?
No, I have both 2.4 and 5 scheduled to turn off and on at the same time.
I faced a problem when the schedule was not the same for both radios. _________________ Netgear R9000 main router
RAX80 as AP
