Build 382.1-B4, both CPUs at 100%- might know why.. The Russians

[jazzmonger]

Russia may actually be the cause....seriously.

UPDATE: more testing reveals the cause - there must be a bug in the core wireless update routine with repeating same letters in the SSID's. These SSIDs all cause the problem:

Russia Russia
RussiaRussia
Russia-Russia
MerlinMerlin
Merlin Merlin
Merlin-Merlin

These dont:
Merlinmerlin
Russiarussia
RussiaRu5siaRuss1a


Build 382 beta 4. 2 brand new RT-AC88U units, both do the same exact thing. Setup/config went fine. Named the Sid’s PV and PV5. No issues for a week. Worked great.

Then I tried paying a joke on my Uber liberal mom at Thanksgiving... changing my ssid from PV to “Russia Russia Russia” password “Maga2017”. Yes, that’s the kind of sick humor that gives me perverse pleasure at Thanksgiving with my family, especially when they have no choice but to connect AND type that password to satiate their FB, Snapchat and CNN addictions because I have spotty cell Svc at my house. Your mileage may vary....

Upon submit with the new awesome-sure to create fireworks-SSID, the wireless system shuts down completely, CPU’s bounce between each other at 100% and you have to physically power cycle the box. Then it comes back. There is no other recovery. Trying reboot from the web interface does nothing and any wireless changes can’t be made.

When it happened last week, I ordered a replacement from Amazon as I was sure it had to be hardware related as it also happens with the firmware shipped with box . The New unit does the same thing.

During my reasearch trying to fix it I SSH’d into the router and noticed the wireless daemon (acsd) was pegging each cpu in succession at about 2-3 second intervals. I killed that process ID and was able to soft reboot the router back or just wait and it comes back into operation.

Load average: 1.24 0.89 0.41 4/121 1423
PID PPID USER STAT VSZ %VSZ CPU %CPU COMMAND
1032 1 admin R 1420 0.2 1 48.4 /usr/sbin/acsd
311 1 admin R 5276 1.0 1 0.7 nt_center
291 1 admin R 4896 0.9 1 0.7 /sbin/netool
351 1 admin S 4876 0.9 1 0.5 sw_devled
346 1 admin S 5488 1.0 0 0.4 httpd -i br0
859 854 admin S 9252 1.7 1 0.3 aaews --sdk_log_dir=/tmp
290 1 admin S 2200 0.4 0 0.3 protect_srv
371 1 admin S 5168 1.0 1 0.1 mastiff
1028 1 admin S 1344 0.2 1 0.1 /bin/eapd
826 1 admin S 9252 1.7 0 0.0 aaews --sdk_log_dir=/tmp
861 854 admin S 9252 1.7 0 0.0 aaews --sdk_log_dir=/tmp
858 854 admin S 9252 1.7 1 0.0 aaews --sdk_log_dir=/tmp
855 854 admin S 9252 1.7 0 0.0 aaews --sdk_log_dir=/tmp
854 826 admin S 9252 1.7 0 0.0 aaews --sdk_log_dir=/tmp
860 854 admin S 9252 1.7 1 0.0 aaews --sdk_log_dir=/tmp
646 1 admin S 6068 1.1 1 0.0 /usr/sbin/smbd -D -s /etc/smb.conf
666 1 admin S 5928 1.1 0 0.0 minidlna -f /etc/minidlna.conf -r
675 674 admin S N 5928 1.1 0 0.0 minidlna -f /etc/minidlna.conf -r
674 666 admin S 5928 1.1 0 0.0 minidlna -f /etc/minidlna.conf -r
644 1 admin S 5904 1.1 1 0.0 nmbd -D -s /etc/smb.conf

Here’s the last part of the system log. The last few lines tell the story. Restart_wireless fails no matter what:

Nov 29 04:50:51 afpd[1142]: WARNING: signature string RT-AC88U-1C08.local is very long !

Nov 29 04:50:51 afpd[1142]: Netatalk AFP/TCP listening on 192.168.3.106:548

Nov 29 04:50:51 Timemachine: daemon is started

Nov 29 04:52:08 rc_service: httpd 336:notify_rc restart_wireless

Nov 29 04:52:10 kernel: dhd_detach(): thread:dhd_watchdog_thread:7a terminated OK

Nov 29 04:52:10 kernel: dhd_detach(): thread:dhd_watchdog_thread:76 terminated OK

Nov 29 04:52:14 kernel: PCI_PROBE: bus 1, slot 0,vendor 14E4, device 4365(good PCI location)

Nov 29 04:52:15 kernel: dhd_attach(): thread:dhd_watchdog_thread:4cf started

Nov 29 04:52:15 kernel: Dongle Host Driver, version 1.363.2 (r665954)

Nov 29 04:52:15 kernel: Compiled in drivers/net/wireless/bcmdhd on Oct 11 2017 at 23:25:58

Nov 29 04:52:15 kernel: Register interface [eth1] MAC: 10:7b:44:ae:1c:08

Nov 29 04:52:15 kernel: PCI_PROBE: bus 1, slot 0,vendor 14E4, device 4365(good PCI location)

Nov 29 04:52:16 kernel: dhd_attach(): thread:dhd_watchdog_thread:4d3 started

Nov 29 04:52:16 kernel: Dongle Host Driver, version 1.363.2 (r665954)

Nov 29 04:52:16 kernel: Compiled in drivers/net/wireless/bcmdhd on Oct 11 2017 at 23:25:58

Nov 29 04:52:16 kernel: Register interface [eth2] MAC: 10:7b:44:ae:1c:0c

Nov 29 04:52:18 kernel: dhd_module_init dhd24 in

Nov 29 04:52:18 kernel: chipnum = 0xaa90

Nov 29 04:52:18 kernel: chiprev = 0x4

Nov 29 04:52:18 kernel: dhd_module_init: dhd24 don't support chip 0xaa90 rev 0x4

Nov 29 04:53:35 rc_service: httpd 336:notify_rc restart_diskmon

Nov 29 04:53:35 rc_service: waitting "restart_wireless" via httpd ...

Nov 29 04:53:50 rc_service: skip the event: restart_diskmon.

Nov 29 04:55:31 rc_service: httpd 336:notify_rc reboot

Nov 29 04:55:31 rc_service: waitting "restart_wireless" via httpd ...

Nov 29 04:55:46 rc_service: skip the event: reboot.

Nov 29 04:55:55 rc_service: httpd 336:notify_rc reboot

Nov 29 04:55:55 rc_service: waitting "restart_wireless" via httpd ...

Nov 29 04:56:10 rc_service: skip the event: reboot.

Nov 29 04:57:04 rc_service: httpd 336:notify_rc reboot

Nov 29 04:57:04 rc_service: waitting "restart_wireless" via httpd ...

Nov 29 04:57:19 rc_service: skip the event: reboot

So, it appers the Russians are just as adept at shutting down this router with no path to recovery as they were at shutting down Hillary Clinton (sorry had to go there).

I’d be curious if others can reproduce since I can on 2 different boxes and if there’s any fix for this before I send them both back. Tormenting my family during the holidays is more important than what router I use. And I like this router sans this problem- the wireless coverage is great.

Jeff

 

[elorimer]

Its the spaces in the SSID

 

[jazzmonger]

Nope. I tried it without the first time.

 

[skeal]

You can try to ssh into the router and issue the erase nvram or nvram erase not sure which one it may work. Be careful erasing the wrong partition can brick your router.

 

[RMerlin]

Build 382 beta 4.

Update to the final release. The acsd issue was only occurring with one of the betas, it was caused by Asus providing the wrong prebuild components for the RT-AC88U.

 

[heysoundude]

Russia may actually be the cause....seriously.

<snip>

Then I tried paying a joke on my Uber liberal mom at Thanksgiving... changing my ssid from PV to “Russia Russia Russia” password “Maga2017”. Yes, that’s the kind of sick humor that gives me perverse pleasure at Thanksgiving with my family, especially when they have no choice but to connect AND type that password to satiate their FB, Snapchat and CNN addictions because I have spotty cell Svc at my house. Your mileage may vary....

I like the cut of your jib. Me, I'm going to re-do my dad's the next time I feel the need to tweak him with POTUSTrump and 4m0r3yE@r5