What's new

Repeated kernel snap shots on RT-AC86U with Merlin 384.18 + Cake-QOS

Morris

Regular Contributor
I've been seeing this snap in my log and it happens every few minutes. Can anyone identify what this is?

Aug 11 22:28:39 kernel: pgd = ffffffc011f25000
Aug 11 22:28:39 kernel: [00000000] *pgd=000000000b70d003, *pud=000000000b70d003, *pmd=0000000011fb0003, *pte=0000000000000000
Aug 11 22:28:39 kernel: CPU: 1 PID: 17413 Comm: dcd Tainted: P O 4.1.27 #2
Aug 11 22:28:39 kernel: Hardware name: Broadcom-v8A (DT)
Aug 11 22:28:39 kernel: task: ffffffc01e049500 ti: ffffffc011008000 task.ti: ffffffc011008000
Aug 11 22:28:39 kernel: PC is at 0xf6e55f44
Aug 11 22:28:39 kernel: LR is at 0x1dc74
Aug 11 22:28:39 kernel: pc : [<00000000f6e55f44>] lr : [<000000000001dc74>] pstate: 600e0010
Aug 11 22:28:39 kernel: sp : 00000000ffe40a38
Aug 11 22:28:39 kernel: x12: 000000000009ff08
Aug 11 22:28:39 kernel: x11: 00000000f60ff024 x10: 00000000000a02ac
Aug 11 22:28:39 kernel: x9 : 00000000f60ffd3c x8 : 00000000000a0764
Aug 11 22:28:39 kernel: x7 : 00000000f60ffd6c x6 : 00000000000a075e
Aug 11 22:28:39 kernel: x5 : 0000000000000000 x4 : 00000000f60ffd20
Aug 11 22:28:39 kernel: x3 : 0000000000000000 x2 : 0000000000000000
Aug 11 22:28:39 kernel: x1 : 000000000007c66c x0 : 0000000000000000

Thank you,

Morris
 
Last edited:

Smokey613

Senior Member
I get these if I enable TM stuff and it is a known ”issue”. I think Asus says it is of no concern but for me it is reason enough to keep all those TM services disabled.
 

Morris

Regular Contributor
I get these if I enable TM stuff and it is a known ”issue”. I think Asus says it is of no concern but for me it is reason enough to keep all those TM services disabled.
Thank you,

Interesting, I've been using AI protection for many years and have only seen this a few times a year till a few days ago. I guess it could be a bad definition causing it.

Morris
 

faux123

Regular Contributor
DCD == data collection daemon.. this is the engine TrendMicro uses to collect statistics on your router. This is a completely "closed" source software, so there's NOTHING anyone can do to fix this except for TrendMicro/Asus themselves.

I was testing between TrendMicro AdaptiveQoS and AiProtect (closed sourced, proprietary) vs CakeQoS+Suricata solution (both open source), I now prefer the later because I can tweak my settings better open source software (nothing to do with data collection paranoia discussed on this forum before).

If only TrendMicro would open source their solutions, people could have contributed to their software like myself to make it better, bug free, etc etc.
 

waeking

Regular Contributor
you have cakeqos running on a RT-68U? I thought this was impossible atm. How did you accomplish this?
 

Morris

Regular Contributor
you have cakeqos running on a RT-68U? I thought this was impossible atm. How did you accomplish this?
Type: it's an 86U

Morris
 

Morris

Regular Contributor
Thank you to all. I've disabled AI Protection and will monitor. Open DNS will provide the protection so if I need to keep it off it's fine. I'll report what the results of the change are.

Morris
 

Wade Coxon

Regular Contributor
Thank you to all. I've disabled AI Protection and will monitor. Open DNS will provide the protection so if I need to keep it off it's fine. I'll report what the results of the change are.
Best to also revoke permissions to the Trend Micro services as well. That will prevent any of them from even being loaded, regardless if they are disabled or not. I got back a huge chunk of memory after doing that too which was a bonus.
 

Morris

Regular Contributor
I turned everything TM related off and withdrew perdition. issue persisted. Stopped cake-qos. No snaps in over an hour and they were every 5 to 15 minutes. Posting to the cake-qos thread.
 

faux123

Regular Contributor
Best to also revoke permissions to the Trend Micro services as well. That will prevent any of them from even being loaded, regardless if they are disabled or not. I got back a huge chunk of memory after doing that too which was a bonus.
Can you share some tips on disabling all TM services? I remember someone stated that TM services are needed for the correct operations of the router even if you never enabled them...
 

Morris

Regular Contributor
Can you share some tips on disabling all TM services? I remember someone stated that TM services are needed for the correct operations of the router even if you never enabled them...
Left side tab: AiProtection
- Click Network Protection, then turn off "Enabled AiProtection"
- Click the top tab "Parental Controls" and then set"Web & Apps Filters" to off

Left side tab: Adaptive QoS:
- On the Bandwidth Monitor tab "Apps analysis" turn off
- Classification Tab set Enable QoS" to off
- Web History Tab set "Web History" to off

I believe that's it

Morris
 

Morris

Regular Contributor
I turned everything TM related off and withdrew perdition. issue persisted. Stopped cake-qos. No snaps in over an hour and they were every 5 to 15 minutes. Posting to the cake-qos thread.
I'm in agreement that it's Trend Micro that's causing the snaps. I enabled it again and did not have cake enabled. Good buy Trend Micro on my routers.

Morris
 

brummygit

Senior Member
Can you share some tips on disabling all TM services? I remember someone stated that TM services are needed for the correct operations of the router even if you never enabled them...
Also

Administration -> Privacy tab -> Withdraw Consent from Trend Micro
 

Morris

Regular Contributor
What puzzles me is that I used Trend Micro AI protect and Traffic Analyzer - Statistic for years without an issue. Then a few days ago I started getting the snaps.

Is this what others have experienced?

Is this something like a bad definition update and the next one will clear the issue?

Thank you,

Morris

Ps. It started about the same time as I installed Cake QoS. That made me suspect Cake yet I've clearly proven that the snaps are caused by Trend Micro.
 

dave14305

Part of the Furniture
What puzzles me is that I used Trend Micro AI protect and Traffic Analyzer - Statistic for years without an issue. Then a few days ago I started getting the snaps.

Is this what others have experienced?

Is this something like a bad definition update and the next one will clear the issue?

Thank you,

Morris

Ps. It started about the same time as I installed Cake QoS. That made me suspect Cake yet I've clearly proven that the snaps are caused by Trend Micro.
I’ve never experienced this because I have an AC68U which isn’t affected, but there was a decent theory put forward last year by @Delusion about additional interfaces causing dcd to crash. See post #15 in the thread below.


The cake script adds an additional interface:
Bash:
ifconfig ifb9eth0 up
Could be completely coincidental, but it’s intriguing nonetheless.
 

faux123

Regular Contributor
Also

Administration -> Privacy tab -> Withdraw Consent from Trend Micro
haha, I was way overthinking this.. I was thinking about scripts and tons of mods to the firmware to disable this. Thank you.
 

Morris

Regular Contributor
I’ve never experienced this because I have an AC68U which isn’t affected, but there was a decent theory put forward last year by @Delusion about additional interfaces causing dcd to crash. See post #15 in the thread below.


The cake script adds an additional interface:
Bash:
ifconfig ifb9eth0 up
Could be completely coincidental, but it’s intriguing nonetheless.
Thank you Dave,

There seems to be a lot in alignment. I'm continuing to look at this as I miss "Traffic Analyzer - Statistic". Lots of useful information when something dose not work and to watch out for running out of bandwidth. I have a RT-AC68U with a bad 5-GHz radio that might let me peek at things.

Morris
 

dave14305

Part of the Furniture
Thank you Dave,

There seems to be a lot in alignment. I'm continuing to look at this as I miss "Traffic Analyzer - Statistic". Lots of useful information when something dose not work and to watch out for running out of bandwidth. I have a RT-AC68U with a bad 5-GHz radio that might let me peek at things.

Morris
You can get a debug log by running this:
Code:
killall dcd
dcd -i 3600 -p 43200 -b -l 8 -d /tmp/bwdpi/ &
Check your system log. There will be a few dcd messages every minute, but near the beginning will be a line to tell you when to expect the crash. It will be random.
Code:
Aug 13 22:20:06 dcd[7832]: [main(1089)] to delay 359 sec ...
Sure enough for me, 6 minutes later the data collection started and completed. In your case, I expect it to crash and whatever is the last dcd line will be interesting.
Bash:
grep -F " dcd[" /tmp/syslog.log
Once you have the output, turn off the logging with:
Bash:
killall dcd
service restart_wrs
Of course, if it crashes, there’s nothing to kill. :rolleyes:

Anyone with dcd crashes can test this. Hoping for common denominators.
 
Last edited:

Latest threads

Sign Up For SNBForums Daily Digest

Get an update of what's new every day delivered to your mailbox. Sign up here!
Top