[iwlagn] Wireless stops working for seconds at a time

Bug #384924 reported by Paul Jones
14
This bug affects 2 people
Affects Status Importance Assigned to Milestone
network-manager (Ubuntu)
Expired
Undecided
Unassigned
Nominated for Intrepid by Emmanuel Stapf [ES]
Nominated for Jaunty by Emmanuel Stapf [ES]
Nominated for Karmic by Emmanuel Stapf [ES]

Bug Description

Binary package hint: network-manager

I'm using Intrepid and my wireless connection (iwlagn) seems to fail intermittently for 15-30s at a time as frequently as every couple of minutes. NetworkManager doesn't report that the connection has failed and there doesn't seem to be anything of note in syslog or wpa_supplicant.log.

I have no idea where this bug could be located and I would like some help with debugging. I have experienced the bug using three difference AP. I haven't eliminated a hardware issue however, so I've only run this laptop using Intrepid. But hardware seems unlikely as the problem is alleviated by rebooting. I have experienced the bug with every Intrepid kernel that has been released thus far. The bug seems to occur mostly when the wireless network is under load. For instance, when I'm streaming videos off of Hulu.

I've looked at various bug reports (303802, 295414 - this bug seems the closest to my problem, but I'm not sure if it's the same bug).

The following displays the lag that the network experiences while watching hulu. The ping returns to normal when hulu is closed.
======================================
> ping yahoo.com
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=91 ttl=55 time=37.4 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=92 ttl=55 time=40.4 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=93 ttl=55 time=119 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=94 ttl=55 time=99.6 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=95 ttl=55 time=97.7 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=96 ttl=55 time=11037 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=97 ttl=55 time=10037 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=98 ttl=55 time=9037 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=99 ttl=55 time=8036 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=100 ttl=55 time=7037 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=101 ttl=55 time=6037 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=102 ttl=55 time=5037 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=103 ttl=55 time=4037 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=104 ttl=55 time=3037 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=105 ttl=55 time=2037 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=106 ttl=55 time=1036 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=107 ttl=55 time=37.7 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=108 ttl=55 time=45.0 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=109 ttl=55 time=79.6 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=110 ttl=55 time=12693 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=111 ttl=55 time=11694 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=112 ttl=55 time=10694 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=113 ttl=55 time=9694 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=114 ttl=55 time=8694 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=115 ttl=55 time=7694 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=116 ttl=55 time=6694 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=117 ttl=55 time=5694 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=118 ttl=55 time=4694 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=119 ttl=55 time=3694 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=120 ttl=55 time=2694 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=121 ttl=55 time=1694 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=122 ttl=55 time=694 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=123 ttl=55 time=36.4 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=124 ttl=55 time=76.5 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=125 ttl=55 time=103 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=126 ttl=55 time=37.1 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=127 ttl=55 time=37.1 ms
64 bytes from b1.www.vip.sp1.yahoo.com (209.131.36.159): icmp_seq=128 ttl=55 time=38.2 ms

ProblemType: Bug
Architecture: amd64
DistroRelease: Ubuntu 8.10
ExecutablePath: /usr/sbin/NetworkManager
Gconf:

IfupdownConfig:
 auto lo
 iface lo inet loopback
Package: network-manager 0.7~~svn20081018t105859-0ubuntu1.8.10.2
ProcAttrCurrent: unconfined
ProcEnviron: PATH=/usr/local/sbin:/usr/local/bin:/sbin:/bin:/usr/sbin:/usr/bin
SourcePackage: network-manager
Syslog:
 Jun 8 11:01:01 Athena NetworkManager: <info> (wlan0): supplicant connection state change: 7 -> 6
 Jun 8 11:01:01 Athena NetworkManager: <info> (wlan0): supplicant connection state change: 6 -> 7
 Jun 8 12:01:00 Athena NetworkManager: <info> (wlan0): supplicant connection state change: 7 -> 6
 Jun 8 12:01:00 Athena NetworkManager: <info> (wlan0): supplicant connection state change: 6 -> 7
Uname: Linux 2.6.27-11-generic x86_64

Tags: apport-bug
Revision history for this message
Paul Jones (pajones) wrote :
Revision history for this message
Jerone Young (jerone) wrote :

Try installing package "linux-backport-modules-intrepid" .. I think that's the name. Then reboot.
Also try out 9.04 live cd and see if you still see the same issue.

Revision history for this message
Alexander Sack (asac) wrote :

besides from checking the backport modules, please see if you can reproduce the network hickup by running "sudo iwlist scan" (e.g. scanning from command line).

Changed in network-manager (Ubuntu):
status: New → Incomplete
Revision history for this message
Paul Jones (pajones) wrote :

@Jerone I have linux-backport-modules-intrepid installed. I will try out the 9.04 live cd and report back. I haven't upgraded my system due to an intel graphics driver bug.

@Alexander I've tried the command "sudo iwlist scan", but it doesn't seem to trigger then bug. I say "doesn't seem" because I haven't found a way to consistently reproduce the bug.

Revision history for this message
Emmanuel Stapf [ES] (xumoqkn5d6or9iyw) wrote :

I'm having the exact same problem on Ubuntu 9.04. In a different mailing list, Paul Jones got the following response:

<<
I'd suspect that the lack of scanning patches in your kernel is the problem.

commit a9a6ffffd05f97e6acbdeafc595e269855829751
Author: Kalle Valo <kalle.valo@xxxxxxxxx>
Date: Wed Mar 18 14:06:44 2009 +0200

You'll also want to try turning on software scanning instead of the hardware scanning, which can help in some cases with older kernels (ie, before 2.6.30). That can be accomplished by passing the driver "disable_hw_scan=1" when modprobing it, usually via file dropped into /etc/modprobe.d/ that contains:

options iwlagn disable_hw_scan=1
>>

but it did not do much to improve the situation.

Any idea what is wrong?

Revision history for this message
Emmanuel Stapf [ES] (xumoqkn5d6or9iyw) wrote : apport-collect data

Architecture: amd64
DistroRelease: Ubuntu 9.04
IfupdownConfig:
 auto lo
 iface lo inet loopback
NonfreeKernelModules: fglrx
Package: network-manager 0.7.1~rc4.1.cf199a964-0ubuntu2
PackageArchitecture: amd64
ProcEnviron:
 PATH=(custom, user)
 SHELL=/usr/bin/tcsh
 LANG=C
Uname: Linux 2.6.28-15-generic x86_64
UserGroups: adm admin cdrom dialout lpadmin plugdev sambashare

Revision history for this message
Emmanuel Stapf [ES] (xumoqkn5d6or9iyw) wrote :

I cannot add more via apport-collect as I'm getting the following error:
Elba [Manu] : rm -rf ~/.cache/apport/
Elba [Manu] : apport-collect 384924
Logging into Launchpad...
The authorization page (https://edge.launchpad.net/+authorize-token?oauth_token=2V0fHRz2HWS2Vf98g4Gx) should be opening in your browser. After you have authorized this program to access Launchpad on your behalf you should come back here and press <Enter> to finish the authentication process.
Xlib: extension "RANDR" missing on display ":0.0".

Downloading bug...
Bug title: [iwlagn] Wireless stops working for seconds at a time
Collecting apport information for source package network-manager...
Uploading additional information to Launchpad bug...
   short text data...
   attachment: Dependencies.txt...
Error connecting to Launchpad: HTTP Error 500: Internal Server Error
You have to allow "Change anything" privileges.
Elba [Manu] :

Revision history for this message
Emmanuel Stapf [ES] (xumoqkn5d6or9iyw) wrote :

I've followed the suggestion from:
http://www.backports.ubuntuforums.org/showthread.php?p=7931639

and although it does not fix the problem, it certainly improved the situation by reducing the number of loss packets.

Revision history for this message
Kenyon Ralph (kralph) wrote :

I noticed that on Debian Squeeze with iwlagn (Intel WiFi Link 5100 in a Dell Latitude 2100) and power management enabled, I get very poor network performance: high packet loss and large latencies. The laptop-mode-tools package by default enables power management for iwlagn via /etc/laptop-mode/conf.d/wireless-iwl-power.conf when the laptop is on battery power. Ubuntu may not work the same way, but use iwconfig to see if your wireless device has power management enabled when your wireless connection goes bad.

Revision history for this message
Launchpad Janitor (janitor) wrote :

[Expired for network-manager (Ubuntu) because there has been no activity for 60 days.]

Changed in network-manager (Ubuntu):
status: Incomplete → Expired
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.