Random Reboot on WDR36000

[smiller]

Hi,

I've been looking for info on what I think is the same problem.

I have had a WDR3600 running Gargoyle for quite a while. (1.7 something, I think)
Recently we had some bandwidth issues at home so I implemented Quotas - something like the "teenager scenario" described here https://www.gargoyle-router.com/wiki/do ... -teenagers

Sometime after this, I started seeing the wifi go up and down. It finally annoyed me enough to look closely at it and I found my router was rebooting spontaneously as often as every few minutes.

First, I upgraded to 1.9.x. (Mar 21) Same problem. So, since I'd had the router a few years, and there were no problems indicated in the logs, I started to think it was a hardware problem. I went out and got a new Archer C7. Flashed it to 1.9.x. Everything good? Nope. Same problem.

Turn quotas off? Problem solved.
I do not have QoS configured at all - just quotas. From other topics I see these are related.

This is easily reproducible for me.

Any ideas?

[Lantis]

Are you able to get a logread at the time it happens?
I know that is difficult given the router restarts on you.

We had identified the issue and put in a patch that should have fixed it.

[smiller]

I'd been running logread -f (IIRC - like "tail -f") over an ssh shell and saw nothing. It just went away. There is probably a 1 second poll interval there? I was also running "dmesg -c; sleep 1; " in a loop - nothing.

I saw some other threads about bandwidth usage log files being corrupted. I do occasionally see logs about timestamp skew. Could NTP be involved?

[Lantis]

Possibly.
You can purge that file and see if the router recovers.
The other threads had instructions I believe.

[Spine]

Code: Select all

[  108.230000] ipt_bandwidth: timezone shift of 180 minutes detected, adjusting
[  108.230000]                old minutes west=0, new minutes west=-180
[  118.390000] Kernel bug detected[#1]:
[  118.390000] CPU: 0 PID: 10281 Comm: tor Not tainted 3.18.23 #5
[  118.390000] task: 8613d360 ti: 85820000 task.ti: 85820000
[  118.390000] $ 0   : 00000000 00000001 810c4e00 00000001
[  118.390000] $ 4   : 00000000 00000001 86270800 00000002
[  118.390000] $ 8   : 0003002a 86906e40 00000015 0003002a
[  118.390000] $12   : 7f9b4828 77662210 77615280 0056c600
[  118.390000] $16   : 869e0540 0001df3f 00000000 8618eb04
[  118.390000] $20   : 878133d0 87ae1a08 00000000 00000000
[  118.390000] $24   : 00000000 775fcc38
[  118.390000] $28   : 85820000 85821e38 7fd33ac8 8022482c
[  118.390000] Hi    : 00000011
[  118.390000] Lo    : 00000000
[  118.390000] epc   : 800f7e7c kfree+0x64/0x11c
[  118.390000]     Not tainted
[  118.390000] ra    : 8022482c __kfree_skb+0x14/0xd0
[  118.390000] Status: 1100dc03 KERNEL EXL IE
[  118.390000] Cause : 00800034
[  118.390000] PrId  : 0001974c (MIPS 74Kc)


[   41.510000] pppoe-wan: renamed from ppp0
[   41.670000] br-lan: port 1(eth0.1) entered forwarding state
[   43.440000] device wlan0 entered promiscuous mode
[   46.140000] br-lan: port 2(wlan0) entered forwarding state
[   46.150000] br-lan: port 2(wlan0) entered forwarding state
[   48.150000] br-lan: port 2(wlan0) entered forwarding state
[   68.220000] device wlan0.sta2 entered promiscuous mode
[   68.220000] br-lan: port 3(wlan0.sta2) entered forwarding state
[   68.230000] br-lan: port 3(wlan0.sta2) entered forwarding state
[   70.230000] br-lan: port 3(wlan0.sta2) entered forwarding state
[  111.390000] ipt_bandwidth: timezone shift of 180 minutes detected, adjusting
[  111.390000]                old minutes west=0, new minutes west=-180
[  120.110000] Kernel bug detected[#1]:
[  120.110000] CPU: 0 PID: 10327 Comm: tor Not tainted 3.18.23 #5
[  120.110000] task: 878d9360 ti: 86ab0000 task.ti: 86ab0000
[  120.110000] $ 0   : 00000000 d3e86ad9 810cf8c0 00000001
[  120.110000] $ 4   : 00000000 00000001 867c6000 00000000
[  120.110000] $ 8   : 00000075 3a2ecf50 e7943e39 80360000
[  120.110000] $12   : 7fe84808 867c6640 00000000 0074b9c8
[  120.110000] $16   : 86946300 0000035d 86946300 00000c8f
[  120.110000] $20   : 00000000 00000000 00000237 865cadac
[  120.110000] $24   : 00000001 77760ef8


Thu Apr  7 10:38:03 2016 daemon.notice Tor[10327]: Bootstrapped 50%: Loading relay descriptors
Thu Apr  7 10:38:03 2016 daemon.notice Tor[10327]: Bootstrapped 50%: Loading relay descriptors
Thu Apr  7 10:38:05 2016 kern.warn kernel: [  120.110000] Kernel bug detected[#1]:
Thu Apr  7 10:38:05 2016 kern.warn kernel: [  120.110000] CPU: 0 PID: 10327 Comm: tor Not tainted 3.18.23 #5
Thu Apr  7 10:38:05 2016 kern.warn kernel: [  120.110000] task: 878d9360 ti: 86ab0000 task.ti: 86ab0000
Thu Apr  7 10:38:05 2016 kern.warn kernel: [  120.110000] $ 0   : 00000000 d3e86ad9 810cf8c0 00000001
Thu Apr  7 10:38:05 2016 kern.warn kernel: [  120.110000] $ 4   : 00000000 00000001 867c6000 00000000
Thu Apr  7 10:38:05 2016 kern.warn kernel: [  120.110000] $ 8   : 00000075 3a2ecf50 e7943e39 80360000
Thu Apr  7 10:38:05 2016 kern.warn kernel: [  120.110000] $12   : 7fe84808 867c6640 00000000 0074b9c8
Thu Apr  7 10:38:05 2016 kern.warn kernel: [  120.110000] $16   : 86946300 0000035d 86946300 00000c8f
Thu Apr  7 10:38:05 2016 kern.warn kernel: [  120.110000] $20   : 00000000 00000000 00000237 865cadac
Thu Apr  7 10:38:05 2016 kern.warn kernel: [  120.110000] $24   : 00000001 77760ef8
Thu Apr  7 10:38:05 2016 kern.warn kernel: [  120.110000] $28   : 86ab0000 86ab1d50 86ab1e30 8022482c
Thu Apr  7 10:38:05 2016 kern.warn kernel: [  120.110000] Hi    : 00000000
Thu Apr  7 10:38:05 2016 kern.warn kernel: [  120.110000] Lo    : 02699afa
Thu Apr  7 10:38:05 2016 kern.warn kernel: [  120.110000] epc   : 800f7e7c kfree+0x64/0x11c
Thu Apr  7 10:38:05 2016 kern.warn kernel: [  120.110000]     Not tainted
Thu Apr  7 10:38:05 2016 kern.warn kernel: [  120.110000] ra    : 8022482c __kfree_skb+0x14/0xd0
Thu Apr  7 10:38:05 2016 kern.warn kernel: [  120.110000] Status: 1100dc03      KERNEL EXL IE
Thu Apr  7 10:38:05 2016 kern.warn kernel: [  120.110000] Cause : 00800034
Thu Apr  7 10:38:05 2016 kern.warn kernel: [  120.110000] PrId  : 0001974c (MIPS 74Kc)

restart stopped if you turn off TOR
1.9.X (Built 20160321-2244 git@a36337c)
TP-Link TL-WDR4300 v1

[smiller]

Possibly.
You can purge that file and see if the router recovers.
The other threads had instructions I believe.

I have done so in the past. It seemed to stop the problem (long enough for me to declare victory) but then it started up again after a few hours IIRC.

[smiller]

Router was up without incident for four days. (This is the new C7)
I re-enabled the throttling rules. Crashed within seconds.

I enabled one rule; a single ten-ip range with one current user. Router promptly crashed.
I changed that one rule to stop traffic instead of throttle. No crash.

I re-enabled a throttle rule for my guest ip address range. No problem... but there are not currently any guests attached. I enabled all the other rules as shutdown instead of throttle. All good.

The rule that crashed the box was set up that it would start throttling almost right away.
One of my kids is on youtube, the rule was designed to throttle after a few MB so he couldn't do video.

So a single throttling rule *that is actively throttling* is enough to crash, but shutdown rules are fine.

[smiller]

I should probably start a new QoS topic. This is no longer random or about WDR3600...