Very unstable WiFi

jada4p · January 4, 2017, 10:04pm

From the very beginning of Turris Omnia installation I encounter problem with Wifi At first I thought it is fault of my ISP provider, thus I developed script testing WAN connection… Of course there were sometimes some connectivity problems, but not so big as my loosing connection again and again… Slowly, after reading TO forum, I realised that I have the same problem that a lot of another TO user have: WiFi is unfortunately unstable, loosing connection again and again I wonder if and when will the team be able to solve this problem. It’s ridiculous - it should be working well from the very beginning… it’s substantial part of router AFAIK I’m not happy with this but I do hope that the team will find a solution soon… all I can do is to keep my fingers crossed and wishing them good luck…

Michal_Makovec · January 4, 2017, 10:46pm

Some of these bugs affects all users, they promised update before the end of January.
(https://twitter.com/turris_cz/status/816683425927229440) in Czech only

mackrauss · January 5, 2017, 3:10pm

Would be nice if the team would post more in English since the user base got very international and not a lot of those new users speak Czech

Maybe the Czech users would be OK with having more English documentation and communication?

jhuebner · January 5, 2017, 7:47pm

I’d generally wish there was more communication towards the users, or maybe I’m just looking at the wrong places. The wifi issue is getting more and more annoying from day to day.

turrisuser · January 5, 2017, 8:38pm

This is so frustrating, I am seriously considering to replace the buggy ath10k card with a Broadcom/Realtek 802.11ac one from this list:
https://wikidevi.com/wiki/List_of_802.11ac_Hardware

epic9x · January 5, 2017, 10:06pm

TL;DR - I think there’s at least two problems, the 2.4GHz/wlan1 drops/deauths clients - though not linux ones, so it would be cool if others on the thread could help corroborate that for testing. Something definitely also got worse after the last one or two updates for the 5Ghz band. Curious if anyone else takes anything from this or has any suggestions for anything to try for the de-auth problem.

Speaking specifically to the 2.4GHz drops some are seeing, I’m suspecting this is maybe the ath9k driver being buggy somehow, I ran into these issues the last time I was running an OpenWRT router and they were never resolved (wish I’d read that Turris was using atheros, might’ve pushed me over to Ubiquity )

I’m suspected drops on the 2.4GHz band, I swapped the 5Ghz to another SSID to confirm and ended up with very unhappy roomies who were previously doing okay on the 5Ghz band (despite some not great performance from the Turris, which looks like it’s maybe being addressed?)

Looking into it, I found this was happening as often as every ten minutes for a single device:

2017-01-05T14:14:23-05:00 info hostapd[]: wlan1: STA f4:5c:89:aa:69:37 IEEE 802.11: deauthenticated due to local deauth request
2017-01-05T14:24:23-05:00 info hostapd[]: wlan1: STA f4:5c:89:aa:69:37 IEEE 802.11: deauthenticated due to local deauth request
2017-01-05T14:34:23-05:00 info hostapd[]: wlan1: STA f4:5c:89:aa:69:37 IEEE 802.11: deauthenticated due to local deauth request
2017-01-05T14:44:23-05:00 info hostapd[]: wlan1: STA f4:5c:89:aa:69:37 IEEE 802.11: deauthenticated due to local deauth request
2017-01-05T14:54:23-05:00 info hostapd[]: wlan1: STA f4:5c:89:aa:69:37 IEEE 802.11: deauthenticated due to local deauth request
2017-01-05T15:04:23-05:00 info hostapd[]: wlan1: STA f4:5c:89:aa:69:37 IEEE 802.11: deauthenticated due to local deauth request
2017-01-05T15:14:23-05:00 info hostapd[]: wlan1: STA f4:5c:89:aa:69:37 IEEE 802.11: deauthenticated due to local deauth request
2017-01-05T15:24:23-05:00 info hostapd[]: wlan1: STA f4:5c:89:aa:69:37 IEEE 802.11: deauthenticated due to local deauth request

To quite a few devices - mostly on wlan1 (2.4Ghz)

grep "deauthenticated due to local deauth request" /var/log/messages  | awk '{print $4,$6}'  |sort | uniq -c
      2 wlan0: a8:66:7f:22:85:3b
      8 wlan1: 0c:47:c9:ab:15:75
      4 wlan1: 2c:f0:a2:11:01:df
      5 wlan1: 48:d7:05:50:c6:0a
     46 wlan1: 64:20:0c:ad:30:f3
     16 wlan1: 64:bc:0c:43:30:d7
     15 wlan1: a8:66:7f:22:85:3b
     62 wlan1: ac:22:0b:49:4d:7d
     50 wlan1: f4:5c:89:aa:69:37

Looked at the mac addresses for each of these based off what was in /tmp/dhcp.leases and each was either an apple product or android, but no linux machines on the network were impacted (a couple RPIs sitting around all have only 2GHz).

On the 5Ghz performance and some suspected (but not really confirmed) drops - I’m not really sure what’s going on there, but I definitely have a worse signal since early december. I do see these things popping up in the logs now - not sure if they’re “normal” - I’d be curious if anyone else has seen this type of thing:

[38148.848037] ath10k_pci 0000:02:00.0: ath10k_pci ATH10K_DBG_BUFFER:
[38148.848047] ath10k: [0000]: 02542E6F 14004C4F 000001CF 0041E75C 00437D70 00000000 00000000 02542E6F
[38148.848052] ath10k: [0008]: 14004C56 00437D70 009BD018 00000011 00000000 00003D20 02542E6F 14004C56
[38148.848057] ath10k: [0016]: 00437D70 0043C2C8 00000000 0000000C 00001D00 02542E6F 14004C4A F2032421
[38148.848062] ath10k: [0024]: F0040000 0000F203 2421F004 0000010C 02542E6F 14005C03 00000000 430CBC64
[38148.848066] ath10k: [0032]: 0000D730 009BA7FC 00437D70 02542E71 14004C52 00437D70 00000080 00000001
[38148.848070] ath10k: [0040]: 00000002 00000208 02542E71 17FC2801 71100B07 03080003 02019003 0000FFFA
[38148.848075] ath10k: [0048]: 0000FFFF 02542E71 17FC2801 71108880 00000008 004481E0 0096003F 00000FF0
[38148.848079] ath10k: [0056]: 02542E71 17FC2801 71108880 0001000A 004481E0 0096003F 0000F81F 02542E71
[38148.848084] ath10k: [0064]: 17FC2801 71108880 0002000A 004481E0 0096003F 0000F81F 02542E71 17FC2801
[38148.848088] ath10k: [0072]: 71108880 0003000B 004481E0 0096003F 0007E01F 02542E71 17FC2801 71108880
[38148.848092] ath10k: [0080]: 0004000C 004481E0 0096003F 000FE01F 02542E71 17FC2801 71108880 0005000C
[38148.848097] ath10k: [0088]: 004481E0 0096003F 000FE01F 02542E71 14004C56 00437D70 009BD060 00000006
[38148.848101] ath10k: [0096]: 0000000C 00003D00 02542EE1 10004C1C 00437D70 00000000 00000006 000000C0
[38148.848105] ath10k: [0104]: 02542EE1 14004C52 00437D70 000000E0 00000001 00000002 00004208
[38148.848110] ath10k_pci 0000:02:00.0: ATH10K_END

Random Info:

root@turris:~# ethtool -i wlan0
driver: ath10k_pci
version: 4.4.38-efe609c5e5f25db4116d6912
firmware-version: 10.1.467-ct-_fW-018-3d8c23b
expansion-rom-version:
bus-info: 0000:02:00.0
supports-statistics: yes
supports-test: no
supports-eeprom-access: no
supports-register-dump: no
supports-priv-flags: no
root@turris:~# ethtool -i wlan1
driver: ath9k
version: 4.4.38-efe609c5e5f25db4116d6912
firmware-version: N/A
expansion-rom-version:
bus-info: 0000:01:00.0
supports-statistics: yes
supports-test: no
supports-eeprom-access: no
supports-register-dump: no
supports-priv-flags: no

iwinfo output - note the “Generic MAC80211” for wlan1, which seemed like a flag:

root@turris:~# iwinfo
wlan0     ESSID: "NOPE"
          Access Point: NOPE
          Mode: Master  Channel: 56 (5.280 GHz)
          Tx-Power: 23 dBm  Link Quality: 51/70
          Signal: -59 dBm  Noise: -102 dBm
          Bit Rate: 97.5 MBit/s
          Encryption: WPA2 PSK (TKIP, CCMP)
          Type: nl80211  HW Mode(s): 802.11bgnac
          Hardware: 168C:003C 0000:0000 [Qualcomm Atheros QCA9880]
          TX power offset: none
          Frequency offset: none
          Supports VAPs: yes  PHY name: phy0

wlan1     ESSID: "NOPE"
          Access Point: NOPE
          Mode: Master  Channel: 11 (2.462 GHz)
          Tx-Power: 19 dBm  Link Quality: 39/70
          Signal: -71 dBm  Noise: -95 dBm
          Bit Rate: 29.4 MBit/s
          Encryption: WPA2 PSK (TKIP, CCMP)
          Type: nl80211  HW Mode(s): 802.11bgn
          Hardware: 168C:002E 168C:30A4 [Generic MAC80211]
          TX power offset: unknown
          Frequency offset: unknown
          Supports VAPs: yes  PHY name: phy1

Any thoughts, or anyone want to compare notes?

ubermutant · January 5, 2017, 10:20pm

some hints are up above. what helped me with the 5GHz was to switch to the master nightly build, which has a different driver than the stable.

Leonardo · January 5, 2017, 11:18pm

I have the issue with only one client (Ecobee3) and another user confirmed he have the same issue.
The de-authorizations for us are more a consequence.
The device can be only be pinged every minute, and all the replies from the ping come in bulk.
I don’t have problems with another router running Open-WRT, so the issue is on the ath9k driver.
I don’t know if the Turris team applied all the patches for it yet.Some they need to be backported from later kernels.

Can you ping the device that de-authorize itself?

I tried static IP, G instead of N, no security, and the ping issue is still there.

Darei · January 6, 2017, 12:15am

I found one strange thing. I have 3 devices on wlan1 (2,4GHz), WiFi module on Sony TV (2014), Android tablet (Prestigio) and SE Xperia Neo. When I turn off phone (Xperia Neo), then I don’t have in Syslog on wlan1 “deauthenticated due to local deauth request”, and others 2,4GHz devices worked like a charm. But if I turn on WiFi on Xperia Neo, then appears “deauthenticated due to local deauth request” for this devices, and others 2,4GHz devices stopped working without any notification in Syslog, or else (I cannot confirm TV for now).

Darei · January 6, 2017, 12:32am

And I found some kind of errors when router starts, I don’t know, if this somehow helps guys.

2017-01-06T01:23:32+01:00 info kernel: [ 18.301638] ath10k driver, optimized for CT firmware, probing pci.
2017-01-06T01:23:32+01:00 info kernel: [ 18.308146] pci 0000:00:02.0: enabling device (0140 → 0142)
2017-01-06T01:23:32+01:00 info kernel: [ 18.314042] ath10k_pci 0000:02:00.0: pci irq msi oper_irq_mode 2 irq_mode 0 reset_mode 0
2017-01-06T01:23:32+01:00 warning kernel: [ 18.493286] ath10k_pci 0000:02:00.0: Direct firmware load for ath10k/pre-cal-pci-0000:02:00.0.bin failed with error -2
2017-01-06T01:23:32+01:00 warning kernel: [ 18.504021] ath10k_pci 0000:02:00.0: Falling back to user helper
2017-01-06T01:23:32+01:00 err kernel: [ 18.513815] firmware ath10k!pre-cal-pci-0000:02:00.0.bin: firmware_loading_store: map pages failed
2017-01-06T00:23:37+01:00 notice netifd: radio1 (1928): handle_probe_req: send failed
2017-01-06T01:23:37+01:00 warning kernel: [ 18.522918] ath10k_pci 0000:02:00.0: Direct firmware load for ath10k/cal-pci-0000:02:00.0.bin failed with error -2
2017-01-06T01:23:37+01:00 warning kernel: [ 18.533303] ath10k_pci 0000:02:00.0: Falling back to user helper
2017-01-06T01:23:37+01:00 err kernel: [ 18.543034] firmware ath10k!cal-pci-0000:02:00.0.bin: firmware_loading_store: map pages failed
2017-01-06T01:23:37+01:00 warning kernel: [ 18.551804] ath10k_pci 0000:02:00.0: Direct firmware load for ath10k/fwcfg-pci-0000:02:00.0.txt failed with error -2
2017-01-06T01:23:37+01:00 warning kernel: [ 18.562362] ath10k_pci 0000:02:00.0: Falling back to user helper
2017-01-06T01:23:37+01:00 err kernel: [ 18.572049] firmware ath10k!fwcfg-pci-0000:02:00.0.txt: firmware_loading_store: map pages failed
2017-01-06T01:23:37+01:00 warning kernel: [ 18.581044] ath10k_pci 0000:02:00.0: Direct firmware load for ath10k/QCA988X/hw2.0/firmware-5.bin failed with error -2
2017-01-06T01:23:37+01:00 warning kernel: [ 18.591777] ath10k_pci 0000:02:00.0: Falling back to user helper
2017-01-06T01:23:37+01:00 err kernel: [ 18.601485] firmware ath10k!QCA988X!hw2.0!firmware-5.bin: firmware_loading_store: map pages failed
2017-01-06T01:23:37+01:00 err kernel: [ 18.610524] ath10k_pci 0000:02:00.0: could not fetch firmware file ‘ath10k/QCA988X/hw2.0/firmware-5.bin’: -11
2017-01-06T00:23:37+01:00 notice netifd: radio1 (1928): handle_probe_req: send failed
2017-01-06T01:23:37+01:00 notice netifd: Last message ‘radio1 (1928): handl’ repeated 3 times, supressed by syslog-ng on DareiRouter
2017-01-06T01:23:37+01:00 warning kernel: [ 18.620519] ath10k_pci 0000:02:00.0: Direct firmware load for ath10k/QCA988X/hw2.0/firmware-4.bin failed with error -2
2017-01-06T01:23:37+01:00 warning kernel: [ 18.631250] ath10k_pci 0000:02:00.0: Falling back to user helper
2017-01-06T01:23:37+01:00 err kernel: [ 18.640936] firmware ath10k!QCA988X!hw2.0!firmware-4.bin: firmware_loading_store: map pages failed
2017-01-06T01:23:37+01:00 err kernel: [ 18.649974] ath10k_pci 0000:02:00.0: could not fetch firmware file ‘ath10k/QCA988X/hw2.0/firmware-4.bin’: -11
2017-01-06T01:23:37+01:00 warning kernel: [ 18.659961] ath10k_pci 0000:02:00.0: Direct firmware load for ath10k/QCA988X/hw2.0/firmware-3.bin failed with error -2
2017-01-06T01:23:37+01:00 warning kernel: [ 18.670689] ath10k_pci 0000:02:00.0: Falling back to user helper
2017-01-06T01:23:37+01:00 err kernel: [ 18.680321] firmware ath10k!QCA988X!hw2.0!firmware-3.bin: firmware_loading_store: map pages failed
2017-01-06T01:23:37+01:00 err kernel: [ 18.689400] ath10k_pci 0000:02:00.0: could not fetch firmware file ‘ath10k/QCA988X/hw2.0/firmware-3.bin’: -11
2017-01-06T01:23:37+01:00 info kernel: [ 18.706404] ath10k_pci 0000:02:00.0: qca988x hw2.0 target 0x4100016c chip_id 0x043202ff sub 0000:0000
2017-01-06T01:23:37+01:00 info kernel: [ 18.715675] ath10k_pci 0000:02:00.0: kconfig debug 0 debugfs 0 tracing 0 dfs 1 testmode 0
2017-01-06T01:23:37+01:00 info kernel: [ 18.725037] ath10k_pci 0000:02:00.0: firmware ver 10.1.467-ct-_fW-018-3d8c23b api 2 features wmi-10.x,has-wmi-mgmt-tx,txstatus-noack,wmi-10.x-CT,ratemask-CT crc32 0920f3f9
2017-01-06T01:23:37+01:00 warning kernel: [ 18.767979] ath10k_pci 0000:02:00.0: Direct firmware load for ath10k/QCA988X/hw2.0/board-2.bin failed with error -2
2017-01-06T01:23:37+01:00 warning kernel: [ 18.778448] ath10k_pci 0000:02:00.0: Falling back to user helper
2017-01-06T01:23:37+01:00 err kernel: [ 18.788014] firmware ath10k!QCA988X!hw2.0!board-2.bin: firmware_loading_store: map pages failed
2017-01-06T01:23:37+01:00 info kernel: [ 18.796943] ath10k_pci 0000:02:00.0: board_file api 1 bmi_id N/A crc32 bebc7c08
2017-01-06T01:23:37+01:00 info kernel: [ 19.697456] ath10k_pci 0000:02:00.0: wmi print ‘P 128 V 8 T 410’
2017-01-06T01:23:37+01:00 info kernel: [ 19.703510] ath10k_pci 0000:02:00.0: wmi print ‘msdu-desc: 1424 sw-crypt: 0’
2017-01-06T01:23:37+01:00 info kernel: [ 19.710670] ath10k_pci 0000:02:00.0: wmi print ‘alloc rem: 27208 iram: 27008’
2017-01-06T01:23:37+01:00 info kernel: [ 19.762772] ath10k_pci 0000:02:00.0: htt-ver 2.1 wmi-op 2 htt-op 2 cal otp max-sta 128 raw 0 hwcrypto 1

dakaix · January 7, 2017, 12:27pm

As of writing I am on the latest nightly build (and Turris OS 3.5); and can reproduce the disconnects consistently.

For me I only need to put the connection under load, copying from my NAS to my MacBook (connected via the 5GHz Ath10k), and by the time I get 4-5GB copied I’ll lose connection. Only way to get it back is to turn the WiFi off and on again within OSX.

I’ve reproduced this 8-10 times now, using both SMB and AFP as the file transfer mechanism.

fraunhofer · January 7, 2017, 4:47pm

Is it just me, or is 5 GHz wifi virtually unusable now? Mine stops connecting to all devices after 12 hours and requires a reboot of the router to restore functionality. It’s quite annoying and makes me want to demand a refund somehow…

nijel · January 7, 2017, 4:59pm

What did help me is using different firmware:

opkg update; opkg remove ath10k-firmware-qca988x-ct ; opkg install ath10k-firmware-qca988x; reboot

After reboot the ath10k works way better. AFAIK this change should be in upcoming update as well.

Nick87 · January 7, 2017, 5:01pm

I’ve put both the 5 and 2.4 ghz network on channel select “auto” and don’t have the problem anymore. I presume it had something to do with the “radar” problem.

thefoxhole · January 9, 2017, 12:15am

Is changing the cards easy or does it require knowledge of loading firmware and such.

Twinkie · January 9, 2017, 5:53am

It does require little bit of linux knowledge, but you can follow guides in documentation or google openwrt config

Simir · January 9, 2017, 1:58pm

After removing ath10k-firmware-qca988x-ct I can not install ath10k-firmware-qca988x:

root@turris:~# opkg install ath10k-firmware-qca988x
Unknown package ‘ath10k-firmware-qca988x’.
Collected errors:

opkg_install_cmd: Cannot install package ath10k-firmware-qca988x.
root@turris:~# opkg install ath10k-firmware-qca988x-ct
Unknown package ‘ath10k-firmware-qca988x-ct’.
Collected errors:

opkg_install_cmd: Cannot install package ath10k-firmware-qca988x-ct.

So now I have only Wifi 2

JamesT42 · January 9, 2017, 2:48pm

do a “opkg update” first

hrbac · January 10, 2017, 8:24am

On Turris HW v1 I am getting simiar error.

At least 2 phones (BBry classic and BBry Priv) a Lenovo yXXX NB get randomly disconnected, not at the same time. Also their response time (kind of slow dns resolve) is sometimes fine, sometimes slow.
Also it is associated with firewall not working (Firewall: 13 hodin, uCollect: 0 hodin, Celkový součet doby výpadků je nyní 129 h (17,92 % povoleného maxima)) Firewall error appears now every day.

The same connection works fine, when same phones are connected to asus router working as bridge and connected via cable. ipv6 is disabled.

Daniel

Nick87 · January 11, 2017, 2:00pm

Hmm, after 7 days of no problems the 5GHZ dropped again.

Restarted the router and everything ok. I’ve manually put the channel on 36 and 80mhz wideband, maybe this works.

It’s indeed probably the driver as I still get errors in the system log.
I’ll wait for the official system update 3.5 to arrive.
Does anyone have an idea about the ETA of 3.?