| View previous topic :: View next topic |
| Author |
Message |
stardotstar
l33t
Joined: 10 Feb 2006
Posts: 887
Location: 2074/SYD/NSW/AU
|
 Posted: Wed May 03, 2006 3:15 am Post subject: strange - I'm stumped with a dhcp wifi issue? after pointers |
 |
|
I am working on three wireless IPs: 1 at home and 2 at work. the one at home and one of the two at work are no brainers - I connect to them with my simple scripts - which basically copy the appropriate /etc/wpa_supplicant.conf into place and launch a restart on /etc/init.d/net.eth1...
so I have skink_up (sets me up at home - never any hassels)
casper_up (associates me with one of the hidden ssids with no key management and a simple wep key at work)
sentinel_up (associates and tries to get dhcp but always times out... (again hidden ssid but no key management and simple wep key)
skink and casper are D-Link, sentinel is linksys
The two at work are differentiated by being on different private subnets - casper is 192.168.0.1 and sentinel is 192.168.100.1
Now: here are the casper and sentinel wpa_supplicant.configs:
| Code: |
geko dhcpc # cat /etc/wpa_supplicant.conf_casper
ctrl_interface=/var/run/wpa_supplicant
ctrl_interface_group=0
eapol_version=2
ap_scan=2
driver_param="hwcrypto=0"
update_config=1
network={
ssid="Casper"
key_mgmt=NONE
wep_key0="*****"
}
geko dhcpc # cat /etc/wpa_supplicant.conf_sentinel
ctrl_interface=/var/run/wpa_supplicant
ctrl_interface_group=0
eapol_version=2
ap_scan=2
driver_param="hwcrypto=0"
update_config=1
network={
ssid="sentinel"
key_mgmt=NONE
wep_key0=########
}
|
the scripts are basic:
| Code: |
geko dhcpc # cat /usr/bin/casper_up
#!/bin/sh
cp -f /etc/wpa_supplicant.conf_casper /etc/wpa_supplicant.conf
/etc/init.d/net.eth1 restart
export http_proxy="nswnrbm1.ca.canon.com.au:80"
geko dhcpc # cat /usr/bin/sentinel_up
#!/bin/sh
cp -f /etc/wpa_supplicant.conf_sentinel /etc/wpa_supplicant.conf
/etc/init.d/net.eth1 restart
|
This is what happens:
| Code: | geko dhcpc # casper_up
* Starting eth1
* Stopping wpa_supplicant on eth1 ... [ ok ] * Starting wpa_supplicant on eth1 ... [ ok ] * eth1 connected to "Casper" at 00:13:46:4C:43:0C
* Bringing up eth1
* dhcp
* Running dhcpcd ... [ ok ] * eth1 received address 192.168.0.102
geko dhcpc # tail /var/log/messages
May 3 12:56:42 geko bridge-eth1: down
May 3 12:56:44 geko bridge-eth1: enabling the bridge
May 3 12:56:44 geko bridge-eth1: up
May 3 12:57:46 geko bridge-eth1: disabling the bridge
May 3 12:57:46 geko dhcpcd[22265]: timed out waiting for a valid DHCP server response
May 3 12:57:46 geko bridge-eth1: down
May 3 13:00:01 geko cron[22281]: (root) CMD (test -x /usr/sbin/run-crons && /usr/sbin/run-crons )
May 3 13:00:01 geko cron[22283]: (root) CMD (rm -f /var/spool/cron/lastrun/cron.hourly)
May 3 13:04:23 geko bridge-eth1: enabling the bridge
May 3 13:04:23 geko bridge-eth1: up
|
so all is good with Casper...
With sentinel though things just time out:
| Code: |
geko dhcpc # sentinel_up
* Stopping eth1
* Bringing down eth1
* Stopping dhcpcd on eth1 ... [ ok ] * Shutting down eth1 ... [ ok ] * Stopping wpa_supplicant on eth1 ... [ ok ] * Starting eth1
* Starting wpa_supplicant on eth1 ... [ ok ] * eth1 connected to "sentinel" at 00:14:BF:E3:EC:84
* Bringing up eth1
* dhcp
* Running dhcpcd ... [ !! ]geko dhcpc # tail /var/log/messages
May 3 13:04:23 geko bridge-eth1: enabling the bridge
May 3 13:04:23 geko bridge-eth1: up
May 3 13:05:13 geko bridge-eth1: disabling the bridge
May 3 13:05:13 geko dhcpcd[22740]: terminating on signal 15
May 3 13:05:13 geko bridge-eth1: down
May 3 13:05:14 geko bridge-eth1: enabling the bridge
May 3 13:05:14 geko bridge-eth1: up
May 3 13:06:16 geko bridge-eth1: disabling the bridge
May 3 13:06:16 geko dhcpcd[23557]: timed out waiting for a valid DHCP server response
May 3 13:06:16 geko bridge-eth1: down
|
iwconfig shows sentinel is found and available and although further away a colleague of mine with windows has been able to confirm that the dhcpd on sentinel is serving addresses and at a more distant range to boot...
| Code: |
eth1 IEEE 802.11g ESSID:"sentinel"
Mode:Managed Frequency:2.432 GHz Access Point: 00:14:BF:E3:EC:84
Bit Rate=54 Mb/s Tx-Power=20 dBm
Retry limit:7 RTS thr:off Fragment thr:off
Encryption key:9A4A-B4F2-F5 Security mode:open
Power Management:off
Link Quality=61/100 Signal level=-64 dBm Noise level=-84 dBm
Rx invalid nwid:0 Rx invalid crypt:11 Rx invalid frag:0
Tx excessive retries:0 Invalid misc:725 Missed beacon:8
|
So.... hmmm; I have tried playing around with the cache and info files in the /var/lib/dhcpc:
| Code: |
geko dhcpc # ls
dhcpcd-eth0.cache dhcpcd-eth0.info.old dhcpcd-eth1.info
dhcpcd-eth0.info dhcpcd-eth1.cache dhcpcd-eth1.info.old
|
ie moving them so that they are not holding earlier information.
I have also tried clearing resolv.conf to no avail.
Perhaps my timeout is too low but it is set at 30 seconds which should be heaps of time!
:
| Code: |
geko conf.d # cat net
modules_eth0=( "ifconfig" )
modules_eth1=( "wpa_supplicant" )
wpa_supplicant_eth1="-Dwext"
wpa_timeout_eth1=60
modules=( "!iwconfig" )
fallback_eth0=( "192.168.0.200 netmask 255.255.255.0" )
fallback_route_eth0=( "default via 192.168.0.1" )
wpa_timeout_eth1=60
config_Casper=( "dhcp" )
#dhcpcd_Casper="-t 5"
config_skink=( "dhcp" )
#dhcpcd_skink="-t 5"
config_sentinel=( "dhcp" )
#config_sentinel=( "192.168.100.207 netmask 255.255.255.0" )
#routes_sentinel=( "default gw 192.168.100.1" )
config_eth0=( "dhcp" )
config_eth1=( "dhcp" )
dhcpcd_eth0="-t 20"
dhcpcd_eth1="-t 30"
|
What else do I need to investigate - I jhave associated with this point before because sometimes I need to administer it but I can't work out what circumstances cause trouble or not!
This is the latest ipw2200bg driver and firmware and the recommended iee80211 modules...
| Code: |
geko conf.d # lsmod
Module Size Used by
vmnet 28708 17
vmmon 187020 5
ieee80211_crypt_wep 4224 1
ipw2200 173384 0
ieee80211 44840 1 ipw2200
ieee80211_crypt 4544 2 ieee80211_crypt_wep,ieee80211
phidgetservo 5248 0
cytherm 5312 0
s2io 61584 0
dummy 2116 0
ibm_acpi 24192 0
i2c_dev 7300 0
ipr 61024 0
firmware_class 8000 2 ipw2200,ipr
sata_sx4 11716 0
sata_sis 5572 0
pcspkr 2436 0
|
TIA
WIll
_________________
]8P |
|
| Back to top |
|
 |
niltonvolpato
n00b
Joined: 28 May 2003
Posts: 68
Location: Campinas, Brazil
|
| Posted: Sun May 07, 2006 6:49 pm Post subject: |
|
|
Very similar problem here: no dhcp lease in another access point. But i'm not using wpa supplicant.
-- Nilton |
|
| Back to top |
|
|
stardotstar
l33t
Joined: 10 Feb 2006
Posts: 887
Location: 2074/SYD/NSW/AU
|
| Posted: Sun May 07, 2006 9:28 pm Post subject: |
|
|
Interesting to hear of a similar experience - I am drawing a blank here.
Further to this issue I have just been off line at work and therefore away from the problem assesspoint for four days.
This time I booted up and immediately associated with the DHCP server that in the post above was always timing out. ie i did not attempt to associate with the other accesspoint that always works. Nor had I been recently associated with any other.
The DHCP server Sentinel immediately gave me a lease and I am working fine through it now *(I even have net.eth0 stopped).
The only clue that may be associated with this is in the messages - despite the successful lease I see:
| Code: |
May 8 07:07:21 geko bridge-eth1: enabling the bridge
May 8 07:07:21 geko bridge-eth1: up
May 8 07:07:24 geko dhcpcd[11203]: DHCP_NAK server response received: requested address on bad subnet
|
I repeated the association and lo and behold more interesting clues:
| Code: | May 8 07:09:22 geko dhcpcd[10163]: terminating on signal 15
May 8 07:09:23 geko bridge-eth0: disabling the bridge
May 8 07:09:23 geko bridge-eth0: down
May 8 07:09:26 geko process `host' is using obsolete setsockopt SO_BSDCOMPAT
May 8 07:09:51 geko dhcpcd[11205]: terminating on signal 15
May 8 07:09:51 geko bridge-eth1: disabling the bridge
May 8 07:09:51 geko bridge-eth1: down
May 8 07:09:51 geko bridge-eth1: enabling the bridge
May 8 07:09:51 geko bridge-eth1: up
|
So despite the successful association with this DHCP server and subsequent fully operational session there appears to be some problems at one end or the other. Is the message pointing to an incompatible negotiation from the DHCP Server or my client daemon?
Can anyone shed any further light on this?
UPDATE:
I associate with Casper (the other AP that always gives me an address) and the association works perfectly straight away with no aberant messages:
| Code: |
May 8 07:27:38 geko bridge-eth1: disabling the bridge
May 8 07:27:38 geko dhcpcd[12518]: terminating on signal 15
May 8 07:27:38 geko bridge-eth1: down
May 8 07:27:40 geko bridge-eth1: enabling the bridge
May 8 07:27:40 geko bridge-eth1: up
May 8 07:27:45 geko dhcpcd[13553]: DHCP_NAK server response received
May 8 07:29:20 geko dhcpcd[13555]: terminating on signal 15
May 8 07:29:20 geko bridge-eth1: disabling the bridge
May 8 07:29:20 geko bridge-eth1: down
May 8 07:29:22 geko bridge-eth1: enabling the bridge
May 8 07:29:22 geko bridge-eth1: up
|
subsequent attempts to connect Sentinel now simply time out...
however association with casper is immediate and without errors or messages:
| Code: |
May 8 07:30:24 geko bridge-eth1: disabling the bridge
May 8 07:30:24 geko dhcpcd[14488]: timed out waiting for a valid DHCP server response
May 8 07:30:24 geko bridge-eth1: down
May 8 07:30:33 geko bridge-eth1: enabling the bridge
May 8 07:30:33 geko bridge-eth1: up
|
Wierd huh?
Perhaps there is a need to flush previous caching from the dhcpcd ?? Maybe a reboot will fix the problem each time? I don't have time right now to test that but if I do I will update this post.
_________________
]8P |
|
| Back to top |
|
|
stardotstar
l33t
Joined: 10 Feb 2006
Posts: 887
Location: 2074/SYD/NSW/AU
|
| Posted: Mon May 08, 2006 10:09 pm Post subject: |
|
|
Rebooting seems to resolve the issue and I think I have confirmed that the presence or not of eth0 wired lan and dhcp does not effect it.
Is there some way to flush the dhcp cache or reset the dhcpcd more thoroughly than what happens when init.d/net.ethx is restarted/stoped/started ??
Will
_________________
]8P |
|
| Back to top |
|
|
niltonvolpato
n00b
Joined: 28 May 2003
Posts: 68
Location: Campinas, Brazil
|
| Posted: Tue May 09, 2006 7:54 pm Post subject: |
|
|
It seems I solved the problem here. I was using a router for 802.11b (only), but it was configured for "basic transfer rates" as "1, 2 Mbps (default)", I changed it to "1, 2, 5.5, 11 Mps". Until now the problem has gone. But it's still very strange.
-- Nilton |
|
| Back to top |
|
|
stardotstar
l33t
Joined: 10 Feb 2006
Posts: 887
Location: 2074/SYD/NSW/AU
|
| Posted: Tue May 09, 2006 10:24 pm Post subject: |
|
|
My Routers are all B/G Mixed but I have switched the basic rate on the problematic one to All and will see what happens.
Thanks for the update.
Will
_________________
]8P |
|
| Back to top |
|
|
|
Networking & Security
