After the mystery reboot 4 days ago, the IP assignment issue cleared itself, I'd guess it's a boot sequencing issue and depends on which services get started first.
As well as the R7000, I have an AT&T Mobley hotspot in my car outside the house. The Mobley WiFi signal is 'fair' inside the house, whereas the R7000 signal is 'strong'. Two times over the last 4 days, I noticed my PC had switched from R7000 WiFi to the Mobley WiFi. Most likely the R7000 WiFi signal turned off for a short period and recovered? Also yesterday, the R7000 admin page wouldn't load (192.168.0.1). I telnet'ed in and checked, httpd was running. Tried to load the web page again and it loaded really slowly and sort of quit, not everything loaded. Tried again and this time the full page loaded normally. I can't say if these are new issues or not, as I only recently bought the Mobley (unlimited AT&T LTE for $20/month, speed reduced after 22GB reached).
Router: Netgear R8000
Firmware: DD-WRT v3.0-r32170M kongac (06/11/17)
Kernel: Linux 4.4.70 #340 SMP Sun Jun 11 12:38:55 CEST 2017 armv7l
Status: Working
Reset: No, used wget, write, reboot
Previous: DD-WRT v3.0-r32170M kongac (06/03/17) and DD-WRT v3.0-r31980M kongac (05/11/17)
Errors: random rebooting
CPU Clock: 1200 MHz
The router now randomly reboots. First, it happened after 15 hours uptime. Then 12-15 hours. Today it happened twice with one reboot in 7 hours. No power issues (and everything is on USP power backup) I wrote a script to maintain system logs on the USB stick to see if there was an error message or signal to reboot. Nothing.
Code:
Jun 20 19:13:51 NG_R8000 daemon.info dnsmasq-dhcp[1996]: DHCPDISCOVER(br0) ec:1a:....
Jun 20 19:13:51 NG_R8000 daemon.info dnsmasq-dhcp[1996]: DHCPOFFER(br0) 192.168.....
Jun 20 19:13:54 NG_R8000 daemon.info dnsmasq-dhcp[1996]: DHCPDISCOVER(br0) 14:91:....
Jun 20 19:13:54 NG_R8000 daemon.info dnsmasq-dhcp[1996]: DHCPOFFER(br0) 192.168.....
Dec 31 19:00:14 NG_R8000 syslog.info syslogd started: BusyBox v1.26.2
Dec 31 19:00:14 NG_R8000 user.info : cron : cron daemon successfully started
Dec 31 19:00:14 NG_R8000 cron.info cron[1033]: (CRON) STARTUP (fork ok)
And this is the same for every reboot. No message, no indication. Just a power up. It seems to happen at times when no one is streaming or doing heavy usage -- 6 am (sleeping), 2 pm (lunch), 7 pm (dinner).
CPU temperatures are still around mid-60s. Nothing else out of the ordinary.
The only changes to config since previous version is addition of CRL to VPN server, turning on SSH with certificate logon, and soft linking the system log to usb stick for persistence. And I am NOT using WDS/Connection watchdog yet. _________________ LinkSys E1000 ...... DD-WRT
LinkSys E2000 ...... DD-WRT
Netgear R8000 ...... Finally DD-WRT
TP-Link WR710N ..... Stock (considering DD-WRT)
Last edited by neerav on Wed Jun 21, 2017 17:10; edited 1 time in total
Joined: 09 Nov 2014 Posts: 314 Location: Bakersfield, CA
Posted: Wed Jun 21, 2017 12:36 Post subject:
This build has been running great on my two R7000's, no reboots or OpenVPN server crashing. USB auto mount still works, and overall seems like a pretty stable build for the R7000. _________________ Deployed Routers:
Netgear R7800 - 1x build 46979
- Gateway (USB /w Entware, CAKE QoS)
Netgear R7000 - 3x build 46979
Router: Netgear R8000
Firmware: DD-WRT v3.0-r32170M kongac (06/11/17)
Kernel: Linux 4.4.70 #340 SMP Sun Jun 11 12:38:55 CEST 2017 armv7l
Status: Working
Reset: No, used wget, write, reboot
Previous: DD-WRT v3.0-r32170M kongac (06/03/17) and DD-WRT v3.0-r31980M kongac (05/11/17)
Errors: random rebooting
CPU Clock: 1200 MHz
The router now randomly reboots. First, it happened after 15 hours uptime. Then 12-15 hours. Today it happened twice with one reboot in 7 hours. No power issues (and everything is on USP power backup) I wrote a script to maintain system logs on the USB stick to see if there was an error message or signal to reboot. Nothing.
Code:
Jun 20 19:13:51 NG_R8000 daemon.info dnsmasq-dhcp[1996]: DHCPDISCOVER(br0) ec:1a:....
Jun 20 19:13:51 NG_R8000 daemon.info dnsmasq-dhcp[1996]: DHCPOFFER(br0) 192.168.....
Jun 20 19:13:54 NG_R8000 daemon.info dnsmasq-dhcp[1996]: DHCPDISCOVER(br0) 14:91:....
Jun 20 19:13:54 NG_R8000 daemon.info dnsmasq-dhcp[1996]: DHCPOFFER(br0) 192.168.....
Dec 31 19:00:14 NG_R8000 syslog.info syslogd started: BusyBox v1.26.2
Dec 31 19:00:14 NG_R8000 user.info : cron : cron daemon successfully started
Dec 31 19:00:14 NG_R8000 cron.info cron[1033]: (CRON) STARTUP (fork ok)
And this is the same for every reboot. No message, no indication. Just a power up. It seems to happen at times when no one is streaming or doing heavy usage -- 6 am (sleeping), 2 pm (lunch), 7 pm (dinner).
CPU temperatures are still around mid-60s. Nothing else out of the ordinary.
The only changes to config since previous version is addition of CRL to VPN server, turning on SSH with certificate logon, and soft linking the system log to usb stick for persistence. And I am NOT using WDS/Connection watchdog yet.
I removed overclocking as I know most units do not run stable with the higher clocks. You may have already damaged your unit through overclocking. _________________ KONG PB's: http://www.desipro.de/ddwrt/
KONG Info: http://tips.desipro.de/
I removed overclocking as I know most units do not run stable with the higher clocks. You may have already damaged your unit through overclocking.
Ooops. My bad. Good thing I included that in my posts. After many years, I ventured into overclocking for the first time 20 days ago... using your blog post. Back to stock standard 1000.
Is it possible to assess for damage other than hoping the router doesn't reboot again randomly?
Fingers crossed that I haven't busted such an expensive router. But that might mean I have an excuse to buy a "newer" or "better" router! _________________ LinkSys E1000 ...... DD-WRT
LinkSys E2000 ...... DD-WRT
Netgear R8000 ...... Finally DD-WRT
TP-Link WR710N ..... Stock (considering DD-WRT)
Router NameR7000
Router ModelNetgear R7000
Firmware Version DD-WRT v3.0-r32170M kongac (06/11/17)
Kernel VersionLinux 4.4.70 #340 SMP Sun Jun 11 12:38:55 CEST 2017 armv7l
I'm facing a very strange behaviour: every night the router freezes and the only way to recover it is to turn it off and on again.
I went from a old mid-2016 build and everything was fine (over 60 day of uptime). I erased NVRAM and configured it from the default values. I don't use IPv6 or anything else...just DHCP, USB Storage and DDNS. Anyone else?
Router: 3x R7000
Firmware: 31870M -> 32170M
Kernel: Linux Linux 4.4.61 #300 SMP -> Linux 4.4.70 #332 SMP Sat Jun 3 11:22:27 CEST 2017 armv7l
Status: All 3 Working & Monitoring
Reset: Erase NVRAM & Manual Configure
Errors: None so far
UpTime: +3H
#1 R7000 (WAN + OpenVPN Client#1)
- IPv4 Only
- USB Custom Startup Script for Route Tables, OpenVPN Client & Cron Scripts
- GUI Firewall & Startup Scripts
- WiFi Disabled
- SPI with non-filtered Multicast
- WAN port -> Cable Modem
- DNSCrpyt with Cisco OpenDNS
- NTP with Static IP
- Cron Job - Monitor WAN & OpenVPN Connections
- QoS HFSC/FQ_CODEL
- DDNS Custom with DNSoMatic
- 6 BRs, 8 VLANS & 6 Assignments
- SSH (WAN Access), Telnet & Syslog
- WAN Traffic Disabled
- USB Storage with Auto Mount
- YaMON 3.1
#2 R7000 (WiFi + OpenVPN Client#2)
- IPv4 Only
- USB Custom Startup Script for OpenVPN Client
- GUI Firewall Script
- 2.4Ghz Radio with 3 SSIDs (BW 40Mhz)
- 5Ghz Radio with 2 SSIDs (BW 80Mhz)
- SPI with non-filtered Multicast
- WAN Port (static IP) Assignment to vlan8 -> Vlan Switch
- DNSSEC via GUI
- NTP with Static IP
- Cron Job - Monitor OpenVPN Connection
- 5 BRs, 7 VLANS & 11 Assignments
- SSH (local only), Telnet & Syslog
- WAN Traffic Disabled
- USB Storage with Auto Mount
#3 R7000 (HE IPv6)
- IPv6 Only (with Limited IPv4)
- 6in4 Static Tunnel with HE
- IPv6 via DNSMasq
- GUI Firewall & Startup Scripts
- WiFi Disabled
- SPI with non-filtered Multicast
- WAN Port -> #1 R7000 LAN Port
- DNSCrypt with Cisco OpenDNS IPv6
- 4 BRs, 4 VLANS & 4 Assignments
- SSH (local only), Telnet & Syslog
- WAN Traffic Disabled
UPDATE....
Quickly Rolled Back to my last working configuration with Build 31870M, No Time to troubleshoot....
Quick-up update: I was having intermittent Internet connect issues which became a big problem at home. I know that OpenVPN video streaming would cause pausing (not buffering) every couple of minutes. Also problems with DNSCypt which caused sites to be down.
Hi Mac 913,
I was reading the comments in [Kong's new build DD-WRT v3.0-r32170M kongac (06/03/17)] and your configuration drew my attention; - 2.4Ghz Radio with 3 SSIDs (BW 40Mhz)
I have two r7000 routers and on both I'm not able to create more than 2 SSID on 2.4Ghz without one disappearing. I saw that you have 3 SSID up and running. And I wanted know to whether you're facing the same problem and otherwise how you did it.
I already made a topic and a bugreport for it where you can find the exact problem I have:
Router: Buffalo WZR-1750DHP
Firmware: DD-WRT v3.0-r32170M kongac (06/03/17)
Previous: DD-WRT v3.0-r31980M kongac (05/11/17)
Kernel: Linux 4.4.70 #332 SMP Sat Jun 3 11:22:27 CEST 2017 armv7l
Status: Working aside from intermittent latency spikes
Reset: No
Errors: None in logs
Uptime: 1 Day
Setup:
- WAN: DHCP
- IPv4 only
- DDNS (no-ip.org)
- Wireless: Regulatory Domain = UNITED_STATES; wl0, AP, AC/N Mixed, VHT80, lower/lower. Implicit Beamforming; wl1, AP, N-Only, 20MHz
- Port Forwarding (only for ARD at this point); netmask forwarding through iptables
- QoS (a few ports, mostly for the 2 netmasks specified in iptables)
- Westwood TCP congestion control.
- Syslogd/klogd with remote server on a laptop on the LAN to monitor in real-time.
Very basic setup. I'm still getting lag spikes over ethernet while playing *any* game. The issue is isolated to this router; if I plug the computer directly into the modem I experience no lag spikes in-game. The lag spikes last for seconds at a time and are upward of 999ms. Syslog doesn't show anything out of the ordinary when the lag spikes occur.
The OOM error the kernel was throwing seems to have disappeared again. It wasn't present in 31830M, appeared again in 31980M. I'll keep this post updated because it seems that it only appears after a few days of uptime.
EDIT:
I'm also concerned that the last couple builds, including this one, seem to maintain high load averages. While network traffic seems consistently low and stable, the load average fluctuates between .10 to .33 or so.
And the temperature, again, without very much load on the device, hovers around 78C. Looking through this thread, the average temp of an R7000 seems to be 55C or so. That's a HUGE difference in temp.
I just upgraded to the stable 32170M build and now the CPU is running hotter (between 80-86C) and showing almost constant .10-.18 load while only 1 computer is connected and sending traffic; incidentally, that computer is doing nothing but using Chrome to access this forum.
Is Buffalo WZR-1750DHP still supported by Kong? It seems as though the majority of people posting are using Netgear R7000 routers. I searched for this router specifically to run DD-WRT, but I haven't been able to get decent performance out of it to date. The lag spikes during gaming exist only when using the router (not if I connect directly to the modem) and it just *seems* slower than my old Netgear WNDR3700. Yes, it's anecdotal, but I didn't have these persistent issues on the WNDR3700.
^---- genuine question. I'm not trying to be snarky or belligerent.
After 4 days of uptime the router suddenly started refusing all connections. Received no kernel OOM error. The only thing visible in syslog was:
"dnsmasq[3276]: failed to send packet: Resource temporarily unavailable."
No extraneous load on the network, just researching info on the web.
The reason why Netgear R7000's chipset temperature is much lower than Buffalo WZR-1750DHP's is due to the heat-sink design difference. On R7000, the bottom cover seems a alumium heat-spreader. But it is a plastic cover on 1750DHP.
My unit also runs at 86 ~ 90 degrees during the summer time w/o overheated issue while the room temperature is around 30 ~ 35 degrees. It should be still within the chipset's maximum thermal design like 100 ~ 105 degrees.
[R7000]
[1750DHP]
My unit has been run with KongAC build DD-WRT FW since 2013 ~ 2014. The following recorded versions were ever installed. The older versions had been removed
Thanks for the info. What is your config? I've only had 1 KongAC build that didn't seem to randomly throw an OOM error and I've had the same setup for ages. And even that build saw huge lag spikes during gaming every few minutes. Those lag spikes are completely absent if I connect the PC directly to the modem.
I do notice that a 'erase nvram && reboot' doesn't clear WiFi passwords.
'erase nvram' only clears your custom settings to reset to default firmware settings, which in this latest Kong build -the default firmware wifi password is the factory wifi pw that came with your router.
On R8000 shows as: "passwordèòv/dev/mtdblock/4"
Also, "password" is missing two characters at end. So if router default wifi password were "supersilly123", this ddwrt build would set it as: "supersilly1èòv/dev/mtdblock/4"
My bad. Good thing I included that in my posts. After many years, I ventured into overclocking for the first time 20 days ago... using your blog post. Back to stock standard 1000.
