| View previous topic :: View next topic |
| Author |
Message |
parsim
Guru
Joined: 12 Aug 2004
Posts: 347
Location: Melbourne, Australia
|
 Posted: Tue Mar 08, 2005 1:22 am Post subject: Why am I falling off the network? |
 |
|
All of a sudden my headless PC has decided to periodically drop off the network. I'll be in the middle of an ssh session, usually doing something with emerge, and suddenly the connection hangs. I can no longer ping the box, but when I go downstairs and look at the LCD screen, it's scrolling along as normal -- so the PC itself is still running.
When I press the reset button and ssh back in, this is what I find was accumulating in the system log:
| Code: | Mar 8 11:58:18 [kernel] NETDEV WATCHDOG: wlan0: transmit timed out
Mar 8 11:58:18 [kernel] Polling for an IRQ FAILED with 0, cmd_status 0, irqs_active 1, irq_status 0. Bailing.
Mar 8 11:58:18 [kernel] [<d08c538f>] acx_issue_cmd+0x3df/0x4d0 [acx_pci]
Mar 8 11:58:18 [kernel] [<d08b671d>] acx111_recalib_radio+0x4d/0x60 [acx_pci]
Mar 8 11:58:18 [kernel] Polling for an IRQ FAILED with 0, cmd_status 0, irqs_active 1, irq_status 0. Bailing.
- Last output repeated 4 times -
Mar 8 11:58:22 [kernel] NETDEV WATCHDOG: wlan0: transmit timed out
Mar 8 11:58:22 [kernel] Polling for an IRQ FAILED with 0, cmd_status 0, irqs_active 1, irq_status 0. Bailing.
Mar 8 11:58:22 [kernel] [<d08c538f>] acx_issue_cmd+0x3df/0x4d0 [acx_pci]
Mar 8 11:58:22 [kernel] Polling for an IRQ FAILED with 0, cmd_status 0, irqs_active 1, irq_status 0. Bailing.
- Last output repeated 4 times -
Mar 8 11:58:26 [kernel] NETDEV WATCHDOG: wlan0: transmit timed out
Mar 8 11:58:26 [kernel] Polling for an IRQ FAILED with 0, cmd_status 0, irqs_active 1, irq_status 0. Bailing.
Mar 8 11:58:26 [kernel] [<d08c538f>] acx_issue_cmd+0x3df/0x4d0 [acx_pci]
Mar 8 11:58:26 [kernel] Polling for an IRQ FAILED with 0, cmd_status 0, irqs_active 1, irq_status 0. Bailing.
- Last output repeated 4 times - |
I'm using a Netgear WG311v2 PCI card (Texas Instruments ACX111 chipset), the acx100 driver, and kernel 2.6.10-gentoo-r6.
If it's relevant, I have frequently noticed lines like this in the syslog ever since I first installed it:
| Code: | tx: error 0x20! (excessive Tx retries due to either distance too high or unable to Tx or Tx frame error - try changing 'iwconfig txpower XXX' or 'sens'itivity or 'retry')
tx: error 0x20! (excessive Tx retries due to either distance too high or unable to Tx or Tx frame error - try changing 'iwconfig txpower XXX' or 'sens'itivity or 'retry')
several excessive Tx retry errors occurred, attempting to recalibrate the radio!! This radio drift *might* be due to increasing card temperature, so you may want to verify proper card temperature, since recalibration might delay card over-temperature failure until it's too late (final fatal card damage). Just a (over?)cautious warning... |
Any ideas? |
|
| Back to top |
|
 |
Tahoe_Strider
Apprentice
Joined: 06 Jul 2003
Posts: 152
Location: Amador County, Awarded 9th Best Place to live in Rural America
|
| Posted: Tue Mar 08, 2005 2:50 am Post subject: |
|
|
Do you have other devices in the same proximity that connect(and remain nailed up) fine? How long was it running fine until these symptoms occurred? I've deployed numerous wireless networks for my customers and the "environment" can change. I know this may not be helpful, but a bit more info might give myself and others some insight. Frankly, it appears as if you may be getting some interference from some radio source and that's why I ask about other devices in the area.
_________________
"It does not require a majority to prevail, but rather an irate, tireless minority keen to set brush fires in people's minds." |
|
| Back to top |
|
|
parsim
Guru
Joined: 12 Aug 2004
Posts: 347
Location: Melbourne, Australia
|
| Posted: Tue Mar 08, 2005 3:27 am Post subject: |
|
|
It's the only wireless PC on this network, and my AP is the only one the card finds when it scans. The trouble started yesterday. Before then it was running fine under ndiswrapper for a month or two, then suddenly it stopped working for no reason I could figure out, and I switched to the acx_pci module. It's been fine on that for a few weeks.
The only other oddity I can think of is that when the machine boots up, it occasionally fails to receive an IP address from the router, and this pops up in the syslog:
| Code: | | Mar 8 12:08:13 [dhcpcd] infinite IP address lease time. Exiting_ |
This has happened every time today. I have a script that checks every few minutes to see if the network connection is working and brings it up if it isn't, without which I wouldn't have been able to log in at all.
Is there a way to check for radio interference? The router is in a room with me, my wireless phone, my mobile phone, and my wireless keyboard and mouse. Over the last month or two I've noticed that occasionally the keyboard seems to go a bit crazy: one second it refuses to register any keystrokes, then it acts as if I'm holding down one of the keys. It often does this just before my mobile phone rings. |
|
| Back to top |
|
|
Tahoe_Strider
Apprentice
Joined: 06 Jul 2003
Posts: 152
Location: Amador County, Awarded 9th Best Place to live in Rural America
|
| Posted: Tue Mar 08, 2005 3:44 am Post subject: |
|
|
Is your wireless phone a 2.4Ghz unit?
_________________
"It does not require a majority to prevail, but rather an irate, tireless minority keen to set brush fires in people's minds." |
|
| Back to top |
|
|
parsim
Guru
Joined: 12 Aug 2004
Posts: 347
Location: Melbourne, Australia
|
| Posted: Tue Mar 08, 2005 3:57 am Post subject: |
|
|
| I don't think so; it's this ("30-39 Mhz Cordless Telephone"). |
|
| Back to top |
|
|
Timitsch
n00b
Joined: 20 Mar 2003
Posts: 19
Location: Zug, Switzerland
|
| Posted: Wed Apr 06, 2005 12:08 am Post subject: |
|
|
i too am dropping off my wireless network. i've experienced the problem for about 1-2 months now on my laptop using an atheros madwifi driver. it gets an IP and works fine for a while and then it losses the connection. works fine however in windows. if i use a cisco pcmcia card however the connection is fine.
i just now installed a acx111 based card on my desktop and am experiencing the same problem. it works fine for a few minutes and then just drops off |
|
| Back to top |
|
|
mirko_3
l33t
Joined: 02 Nov 2003
Posts: 605
Location: Birreria
|
| Posted: Tue Apr 12, 2005 4:07 pm Post subject: |
|
|
Happens here with an acx111 card as well... WEP enabled, thanks to the latest release...
_________________
Non fa male! Non fa male! |
|
| Back to top |
|
|
parsim
Guru
Joined: 12 Aug 2004
Posts: 347
Location: Melbourne, Australia
|
| Posted: Wed Apr 20, 2005 6:33 am Post subject: |
|
|
This problem went away for me for a while, but today I switched a few things (-r4 of acx100 instead of -r3, enabled WEP, and changed channel) and it's back! Dropped silently off the network after a few hours, then when I tried to reconnect, it generated the same syslog errors about being unable to poll for an IRQ.
I'm going to try rolling back the things I did one by one until it stops happening. |
|
| Back to top |
|
|
mirko_3
l33t
Joined: 02 Nov 2003
Posts: 605
Location: Birreria
|
| Posted: Wed Apr 20, 2005 4:50 pm Post subject: |
|
|
I fixed it by changing firmware file...
_________________
Non fa male! Non fa male! |
|
| Back to top |
|
|
parsim
Guru
Joined: 12 Aug 2004
Posts: 347
Location: Melbourne, Australia
|
| Posted: Thu Apr 28, 2005 12:30 pm Post subject: |
|
|
I'm still having this very annoying error... it's so intermittant and (apparently) random that I'm having trouble identifying any pattern.
I did notice today that the network connection restored itself some hours later, and this appeared in the system log:
| Code: | Apr 28 19:38:54 [kernel] NETDEV WATCHDOG: wlan0: transmit timed out
Apr 28 19:38:54 [kernel] successfully recalibrated radio
Apr 28 19:38:54 [kernel] DEAUTHEN <4>00:09:5b:ba:f1:af <4>00:09:5b:ba:f1:af <4>00:0f:b5:50:1c:d2 <4>00:0f:b5:50:1c:d2 <4>00:0f:b5:50:1c:d2
Apr 28 19:38:56 [kernel] acx_timer: status = 2
Apr 28 19:38:56 [kernel] 00:09:5b:ba:f1:af <4>00:09:5b:ba:f1:af <4>00:0f:b5:50:1c:d2 <4>00:0f:b5:50:1c:d2 <4>00:0f:b5:50:1c:d2
Apr 28 19:38:56 [kernel] acx_set_status: Setting status = 4 (ASSOCIATED)
Apr 28 19:38:57 [kernel] acx_timer: status = 4 |
|
|
| Back to top |
|
|
mirko_3
l33t
Joined: 02 Nov 2003
Posts: 605
Location: Birreria
|
| Posted: Thu Apr 28, 2005 12:39 pm Post subject: |
|
|
Have you tried different firmwares?
_________________
Non fa male! Non fa male! |
|
| Back to top |
|
|
parsim
Guru
Joined: 12 Aug 2004
Posts: 347
Location: Melbourne, Australia
|
| Posted: Thu Apr 28, 2005 10:07 pm Post subject: |
|
|
No, I just use the firmware that comes with the acx100 package. I'm not sure how I would change to different firmware. What did you change from and to? And were you experiencing the same "Polling for IRQ" errors in the system log?
Incidentally I didn't seem to have this problem with -r3, and I think that was the same firmware. |
|
| Back to top |
|
|
mirko_3
l33t
Joined: 02 Nov 2003
Posts: 605
Location: Birreria
|
| Posted: Mon May 02, 2005 3:34 pm Post subject: |
|
|
I don't really remember, but I guess it's worth a try... The readme at http://rhlx01.fht-esslingen.de/~andi/acx100/ expalins where to get different firmware files and how to test them better than I could explain it
_________________
Non fa male! Non fa male! |
|
| Back to top |
|
|
|
Networking & Security
