View previous topic :: View next topic |
Author |
Message |
bluephoenix Tux's lil' helper

Joined: 01 Nov 2008 Posts: 94
|
Posted: Mon Apr 21, 2014 11:07 am Post subject: Should I use wlan0 or wlp0s26u1u1? |
|
|
Dears,
I am using LINKSYS Compact wireless-G USB adapter and the model number is WUSB54GC. I am using systemd and wpa_supplicant.
I ever used to use wlan0 when I deploy my wifi connection. Everything is OK at that time. But as I updated new kernel (now is 3.12.13) some days ago, I remember some documents suggest me to use wlp0sxxxxxx this kind of interface presentation as a trend. So I disabled all of the valid configuration in "/etc/udev/rules.d/70-persistent-net.rules". So the system gave me "wlp0s26u1u1" as the new wlan interface. And then the problem comes.
When I boot or reboot my computer, the system cannot correctly activate the wlan interface. I mean if I use "ifconfig" command, the wlan interface cannot be displayed, whatever "wlan0" or "wlp0s26u1u1". But if I use "lsusb" command, the dongle can be displayed. But if I plug and insert the USB dongle manually, everything becomes OK. I mean if I use "ifconfig" to show, the "wlp0s26u1u1" interface could be displayed, the IP address could be given by the wifi router and I can surf the internet also.
1. I checked the "dmesg" when I initially reboot my computer with the dongle inserted but not insert the dongle after booting as below:
Code: | [ 1.385892] usb 1-1.1: New USB device found, idVendor=13b1, idProduct=0020
[ 1.385894] usb 1-1.1: New USB device strings: Mfr=1, Product=2, SerialNumber=0
[ 1.385896] usb 1-1.1: Product: Compact Wireless-G USB Adapter
[ 1.385911] usb 1-1.1: Manufacturer: Cisco-Linksys
[ 1.460008] usb 1-1.1: reset high-speed USB device number 3 using ehci-pci
[ 1.722284] [drm] Enabling RC6 states: RC6 on, RC6p on, RC6pp off
[ 1.877010] ieee80211 phy0: rt2x00_set_chip: Info - Chipset detected - rt: 2573, rf: 0002, rev: 000a
[ 1.877688] ieee80211 phy0: Selected rate control algorithm 'minstrel_ht'
[ 2.241672] ieee80211 phy0: rt2x00lib_request_firmware: Info - Loading firmware file 'rt73.bin'
[ 2.244437] ieee80211 phy0: rt2x00lib_request_firmware: Info - Firmware detected - version: 1.7 |
2. I checked the "dmesg" after I plug and insert the USB dongle manually as below:
Code: | [ 948.846265] usb 1-1.1: USB disconnect, device number 3
[ 950.302706] usb 1-1.1: new high-speed USB device number 6 using ehci-pci
[ 950.576223] usb 1-1.1: New USB device found, idVendor=13b1, idProduct=0020
[ 950.576227] usb 1-1.1: New USB device strings: Mfr=1, Product=2, SerialNumber=0
[ 950.576229] usb 1-1.1: Product: Compact Wireless-G USB Adapter
[ 950.576231] usb 1-1.1: Manufacturer: Cisco-Linksys
[ 950.650694] usb 1-1.1: reset high-speed USB device number 6 using ehci-pci
[ 951.068079] ieee80211 phy1: rt2x00_set_chip: Info - Chipset detected - rt: 2573, rf: 0002, rev: 000a
[ 951.068628] ieee80211 phy1: Selected rate control algorithm 'minstrel_ht'
[ 951.076607] systemd-udevd[1992]: renamed network interface wlan0 to wlp0s26u1u1
[ 951.078745] ieee80211 phy1: rt2x00lib_request_firmware: Info - Loading firmware file 'rt73.bin'
[ 951.078765] ieee80211 phy1: rt2x00lib_request_firmware: Info - Firmware detected - version: 1.7
[ 952.522826] wlp0s26u1u1: authenticate with 08:10:77:a9:e7:ce
[ 952.580470] wlp0s26u1u1: send auth to 08:10:77:a9:e7:ce (try 1/3)
[ 952.604513] wlp0s26u1u1: authenticated
[ 952.604531] rt73usb 1-1.1:1.0 wlp0s26u1u1: disabling HT/VHT due to WEP/TKIP use
[ 952.605567] wlp0s26u1u1: associate with 08:10:77:a9:e7:ce (try 1/3)
[ 952.608036] wlp0s26u1u1: RX AssocResp from 08:10:77:a9:e7:ce (capab=0x11 status=0 aid=3)
[ 952.618002] wlp0s26u1u1: associated |
My question is why everything is OK when I use old "wlan0" presentation while system cannot load the USB dongle automatically when I try to use the new style "wlp0s26u1u1". Is there anything wrong with my configuration? |
|
Back to top |
|
 |
bluephoenix Tux's lil' helper

Joined: 01 Nov 2008 Posts: 94
|
Posted: Wed Apr 23, 2014 12:45 am Post subject: |
|
|
Today I use command "journalctl -r" to show the relative log as below when I try to use wlp0s26u1u1:
Code: | 4月 23 08:12:06 BPLinuxDesk dhcpcd[1767]: wlp0s26u1u1: stopping wpa_supplicant
4月 23 08:12:06 BPLinuxDesk kernel: Switched to clocksource tsc
4月 23 08:12:06 BPLinuxDesk dhcpcd[1639]: wlp0s26u1u1: removing interface
4月 23 08:12:06 BPLinuxDesk dhcpcd[1639]: wlp0s26u1u1: waiting for carrier
4月 23 08:12:06 BPLinuxDesk dhcpcd[1639]: wlp0s26u1u1: waiting for carrier
4月 23 08:12:06 BPLinuxDesk dhcpcd[1724]: wlp0s26u1u1: starting wpa_supplicant
4月 23 08:12:05 BPLinuxDesk kernel: ieee80211 phy0: rt2x00lib_request_firmware: Info - Firmware detected - version: 1.7
4月 23 08:12:05 BPLinuxDesk kernel: ieee80211 phy0: rt2x00lib_request_firmware: Info - Loading firmware file 'rt73.bin'
4月 23 08:12:05 BPLinuxDesk dhcpcd[1639]: enp2s0: waiting for carrier
4月 23 08:12:05 BPLinuxDesk dhcpcd[1639]: no interfaces have a carrier
4月 23 08:12:05 BPLinuxDesk systemd-sysctl[1714]: Overwriting earlier assignment of kernel/sysrq in file '/usr/lib64/sysctl.d/60-gentoo.conf'.
4月 23 08:12:05 BPLinuxDesk systemd-udevd[1487]: renamed network interface wlan0 to wlp0s26u1u1
4月 23 08:12:05 BPLinuxDesk kernel: ieee80211 phy0: Selected rate control algorithm 'minstrel_ht'
4月 23 08:12:05 BPLinuxDesk kernel: ieee80211 phy0: rt2x00_set_chip: Info - Chipset detected - rt: 2573, rf: 0002, rev: 000a
4月 23 08:12:05 BPLinuxDesk kernel: [drm] Enabling RC6 states: RC6 on, RC6p on, RC6pp off
4月 23 08:12:05 BPLinuxDesk kernel: usb 1-1.1: reset high-speed USB device number 3 using ehci-pci
4月 23 08:12:05 BPLinuxDesk kernel: usb 1-1.1: Manufacturer: Cisco-Linksys
4月 23 08:12:05 BPLinuxDesk kernel: usb 1-1.1: Product: Compact Wireless-G USB Adapter
4月 23 08:12:05 BPLinuxDesk kernel: usb 1-1.1: New USB device strings: Mfr=1, Product=2, SerialNumber=0
4月 23 08:12:05 BPLinuxDesk kernel: usb 1-1.1: New USB device found, idVendor=13b1, idProduct=0020
4月 23 08:12:05 BPLinuxDesk kernel: tsc: Refined TSC clocksource calibration: 3292.522 MHz
4月 23 08:12:05 BPLinuxDesk kernel: usb 1-1.1: new high-speed USB device number 3 using ehci-pci |
And when I try to use wlan0 instead of wlp0s26u1u1, I get the following log:
Code: | 4月 23 08:24:52 BPLinuxDesk dhcpcd[1637]: wlan0: adding default route via 192.168.1.253
4月 23 08:24:52 BPLinuxDesk dhcpcd[1637]: wlan0: adding route to 192.168.1.0/24
4月 23 08:24:52 BPLinuxDesk dhcpcd[1637]: wlan0: leased 192.168.1.10 for 86400 seconds
4月 23 08:24:43 BPLinuxDesk dhcpcd[1637]: wlan0: offered 192.168.1.10 from 192.168.1.253
4月 23 08:24:39 BPLinuxDesk dhcpcd[1637]: wlan0: soliciting a DHCP lease
4月 23 08:24:39 BPLinuxDesk dhcpcd[1637]: wlan0: DHCP lease expired
4月 23 08:24:34 BPLinuxDesk dhcpcd[1637]: wlan0: rebinding lease of 192.168.1.10
4月 23 08:24:34 BPLinuxDesk dhcpcd[1637]: wlan0: carrier acquired
4月 23 08:24:34 BPLinuxDesk kernel: wlan0: associated
4月 23 08:24:34 BPLinuxDesk kernel: wlan0: RX AssocResp from 08:10:77:a9:e7:ce (capab=0x11 status=0 aid=3)
4月 23 08:24:34 BPLinuxDesk kernel: wlan0: associate with 08:10:77:a9:e7:ce (try 1/3)
4月 23 08:24:34 BPLinuxDesk kernel: rt73usb 1-1.1:1.0 wlan0: disabling HT/VHT due to WEP/TKIP use
4月 23 08:24:34 BPLinuxDesk kernel: wlan0: authenticated
4月 23 08:24:34 BPLinuxDesk kernel: wlan0: send auth to 08:10:77:a9:e7:ce (try 1/3)
4月 23 08:24:34 BPLinuxDesk kernel: wlan0: authenticate with 08:10:77:a9:e7:ce
4月 23 08:24:33 BPLinuxDesk dhcpcd[1637]: wlan0: waiting for carrier
4月 23 08:24:33 BPLinuxDesk dhcpcd[1637]: wlan0: carrier lost
4月 23 08:24:33 BPLinuxDesk dhcpcd[1637]: wlan0: carrier acquired
4月 23 08:24:33 BPLinuxDesk dhcpcd[1637]: wlan0: waiting for carrier
4月 23 08:24:33 BPLinuxDesk dhcpcd[1720]: wlan0: starting wpa_supplicant
4月 23 08:24:32 BPLinuxDesk kernel: ieee80211 phy0: rt2x00lib_request_firmware: Info - Firmware detected - version: 1.7
4月 23 08:24:32 BPLinuxDesk kernel: ieee80211 phy0: rt2x00lib_request_firmware: Info - Loading firmware file 'rt73.bin'
4月 23 08:24:32 BPLinuxDesk dhcpcd[1637]: enp2s0: waiting for carrier
4月 23 08:24:32 BPLinuxDesk dhcpcd[1637]: no interfaces have a carrier
4月 23 08:24:32 BPLinuxDesk systemd-sysctl[1710]: Overwriting earlier assignment of kernel/sysrq in file '/usr/lib64/sysctl.d/60-gentoo.conf'.
4月 23 08:24:32 BPLinuxDesk /etc/init.d/net.wlan0[1709]: net.wlan0: not allowed to be hotplugged
4月 23 08:24:32 BPLinuxDesk kernel: ieee80211 phy0: Selected rate control algorithm 'minstrel_ht'
4月 23 08:24:32 BPLinuxDesk kernel: ieee80211 phy0: rt2x00_set_chip: Info - Chipset detected - rt: 2573, rf: 0002, rev: 000a
4月 23 08:24:32 BPLinuxDesk kernel: [drm] Enabling RC6 states: RC6 on, RC6p on, RC6pp off
4月 23 08:24:32 BPLinuxDesk kernel: usb 1-1.1: reset high-speed USB device number 3 using ehci-pci
4月 23 08:24:32 BPLinuxDesk kernel: usb 1-1.1: Manufacturer: Cisco-Linksys
4月 23 08:24:32 BPLinuxDesk kernel: usb 1-1.1: Product: Compact Wireless-G USB Adapter
4月 23 08:24:32 BPLinuxDesk kernel: usb 1-1.1: New USB device strings: Mfr=1, Product=2, SerialNumber=0
4月 23 08:24:32 BPLinuxDesk kernel: usb 1-1.1: New USB device found, idVendor=13b1, idProduct=0020
4月 23 08:24:32 BPLinuxDesk kernel: tsc: Refined TSC clocksource calibration: 3292.520 MHz
4月 23 08:24:32 BPLinuxDesk kernel: usb 1-1.1: new high-speed USB device number 3 using ehci-pci |
When I use wlan0, I can find that interface is up after booting, but when I use wlp0s26u1u1 that interface is down.
I doubt if I have some missing configuration cause the booting order is wrong, which makes the wpa_supplicant works after dhcpcd. So the process cannot go ahead correctly?
BTW: I found my ethernet interface is also up even there is no cable connecting. I think it wired!
Still make effort to solve this issue. |
|
Back to top |
|
 |
bluephoenix Tux's lil' helper

Joined: 01 Nov 2008 Posts: 94
|
Posted: Wed Apr 23, 2014 1:46 am Post subject: |
|
|
I think I found the issue.
if I disable dhcpcd.service from the booting time, but start it after the booting. Everything will be OK.
So I believe as some clue listed in Archlinux website, dhcpcd does't works very smoothly with wifi interface, not like the status with a wire-ethernet interface.
Just till now, I don't know how to make a clean solution.If I can load the dhcpcd at last during the booting time, I think this issue can be solved. |
|
Back to top |
|
 |
saellaven l33t

Joined: 23 Jul 2006 Posts: 655
|
Posted: Wed Apr 23, 2014 3:38 am Post subject: |
|
|
Keep in mind that, if you use the new naming scheme, should you plug your adapter into any other USB port or possibly even add a USB hub/other devices, you're going to get a different device node and have to change your settings accordingly.
IMO the new naming scheme is extremely broken in related to its stated purposes, but that's a giant can of worms that's already been discussed elsewhere. If the old naming scheme is working for you, there's little harm in sticking with it. |
|
Back to top |
|
 |
Aiken Apprentice

Joined: 22 Jan 2003 Posts: 240 Location: Toowoomba/Australia
|
Posted: Wed Apr 23, 2014 10:07 pm Post subject: |
|
|
saellaven wrote: | Keep in mind that, if you use the new naming scheme, should you plug your adapter into any other USB port or possibly even add a USB hub/other devices, you're going to get a different device node and have to change your settings accordingly.
|
Been bitten by that. One of the computers I look after has 8 or 10 usb ports, provides an access point and has a couple of firewall rules. It gets moved sometimes. The 4 usb cables that are plugged into it are not always put in exactly the same usb port they came from.
The moment the wifi adaptor is plugged into a different port hostapd.conf and the firewall rules are broken. So much for being able to plug in a usb device and have it 'just work'. My preference is to disable the renaming unless actually needed. I still believe it should have been that way instead of default. My dual nic machines are well behaved and do not mysteriously swap names and I do not see why renaming would be needed for eth0 or wlan0 or eth0 + wlan0 machines. _________________ Beware the grue. |
|
Back to top |
|
 |
bluephoenix Tux's lil' helper

Joined: 01 Nov 2008 Posts: 94
|
Posted: Fri Apr 25, 2014 3:01 am Post subject: |
|
|
Thanks, you guys!
I don't know why new naming scheme is developed. But I think there must be some reason. There is a Chinese interesting sentence: " if I don't know why, I feel he is very hard."
Although, I stick to use wlan0 - the old name to keep the stability, I guess if I use systemd-tools to replace udev, maybe it can help something.
I will try it later. |
|
Back to top |
|
 |
saellaven l33t

Joined: 23 Jul 2006 Posts: 655
|
Posted: Sat Apr 26, 2014 10:08 pm Post subject: |
|
|
bluephoenix wrote: | Thanks, you guys!
I don't know why new naming scheme is developed. But I think there must be some reason. There is a Chinese interesting sentence: " if I don't know why, I feel he is very hard."
Although, I stick to use wlan0 - the old name to keep the stability, I guess if I use systemd-tools to replace udev, maybe it can help something.
I will try it later. |
The purpose was to create predictable names that don't conflict with the kernel's naming of devices, so that device nodes will always point to the same device. The problem is, it does that by assuming that device will always exist in the same physical location - bus slot, USB port, etc. The simple fact is, USB ports change, bus slots can be renumbered when adding devices or changing BIOS options, etc. So, in effect, the names are neither predictable nor persistent and they can bite you when you least expect it, in particularly painful ways.
Not much thought was actually put into the design of the new system, it's just what the people behind it wanted and they declared it to be better, even though it really isn't. And it has the worst effects on people with limited networking options (1 wired and/or 1 wireless device is pretty common on most consumer hardware), all in the name of making it "easier" for them, while the corporate users should know what they are doing and know the best time to check/change configuration is when adding/removing hardware, not when you do a reboot days/weeks/months later only to find that your remote server no longer has any networking.
It's simply arrogance and ignorance working hand in hand and it's caused far more problems than it has fixed. The easiest, best solution is to simply assign device names based on MAC and, if you get cheap hardware with identical MACs, maybe it's best to buy something $1 more expensive since I'm sure that won't be the only issue you'll be having. |
|
Back to top |
|
 |
|
|
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum
|
|