4

I am not the strongest Linux admin. I just thought I would throw that out there. I have a Linux wireless access point with HOSTAPD and DNSMASQ. I have been working on this for quite some time and I have resolved some issues but the primary problem still remains. I did have a problem where HOSTAPD would not function correctly after a reboot or service restart. I was able to correct that with service dependencies. There are so many google responses for HOSTAPD fails I cannot find a problem that matches mine. I have tried to implement a few such as not letting the NIC go to sleep and so forth but it has not corrected the primary issue.

I will have my devices connected to the WAP and it will run for 6 to 10 hours then the clients will start to drop one by one. They do not all drop at the same time but when one drops then the others will start to drop with in 10 minutes. I have set DNSMASQ to give leases for 2 minutes. This was to make sure the DHCP service was not the issue. I was able to get the exact time of a failure but syslog did not have any juicy details. It just has working then not working in an infinite loop.

Oct 17 17:17:01 raspberrypi CRON[1084]: (root) CMD (   cd / && run-parts --report /etc/cron.hourly)
Oct 17 17:17:18 raspberrypi dnsmasq-dhcp[598]: DHCPREQUEST(wlan0) 192.168.3.101 00:bb:3a:35:74:e1
Oct 17 17:17:18 raspberrypi rsyslogd-2007: action 'action 17' suspended, next retry is Sat Oct 17 17:18:48 2015 [try http://www.rsyslog.com/e/2007 ]
Oct 17 17:17:18 raspberrypi dnsmasq-dhcp[598]: DHCPACK(wlan0) 192.168.3.101 00:bb:3a:35:74:e1 android-8edf05d3f461104e
Oct 17 17:17:18 raspberrypi dnsmasq-dhcp[598]: not giving name raspberrypi to the DHCP lease of 192.168.4.114 because the name exists in /etc/hosts with address 127.0.1.1
Oct 17 17:18:14 raspberrypi dnsmasq-dhcp[598]: DHCPREQUEST(wlan0) 192.168.3.101 00:bb:3a:35:74:e1
Oct 17 17:18:14 raspberrypi dnsmasq-dhcp[598]: DHCPACK(wlan0) 192.168.3.101 00:bb:3a:35:74:e1 android-8edf05d3f461104e
Oct 17 17:18:14 raspberrypi dnsmasq-dhcp[598]: not giving name raspberrypi to the DHCP lease of 192.168.4.114 because the name exists in /etc/hosts with address 127.0.1.1
Oct 17 17:19:21 raspberrypi dnsmasq-dhcp[598]: DHCPREQUEST(eth1) 192.168.4.114 b8:27:eb:05:9b:c8
Oct 17 17:19:21 raspberrypi dnsmasq-dhcp[598]: DHCPACK(eth1) 192.168.4.114 b8:27:eb:05:9b:c8 raspberrypi
Oct 17 17:19:21 raspberrypi rsyslogd-2007: action 'action 17' suspended, next retry is Sat Oct 17 17:20:51 2015 [try http://www.rsyslog.com/e/2007 ]
Oct 17 17:19:21 raspberrypi dnsmasq-dhcp[598]: not giving name raspberrypi to the DHCP lease of 192.168.4.114 because the name exists in /etc/hosts with address 127.0.1.1
Oct 17 17:19:44 raspberrypi hostapd: wlan0: STA 00:bb:3a:35:74:e1 IEEE 802.11: deauthenticated due to local deauth request
Oct 17 17:19:44 raspberrypi hostapd: wlan0: STA 00:bb:3a:35:74:e1 IEEE 802.11: disassociated
Oct 17 17:19:44 raspberrypi hostapd: wlan0: STA 00:bb:3a:35:74:e1 IEEE 802.11: associated
Oct 17 17:19:48 raspberrypi hostapd: wlan0: STA 00:bb:3a:35:74:e1 IEEE 802.11: deauthenticated due to local deauth request
Oct 17 17:19:48 raspberrypi hostapd: wlan0: STA 00:bb:3a:35:74:e1 IEEE 802.11: disassociated
Oct 17 17:19:48 raspberrypi hostapd: wlan0: STA 00:bb:3a:35:74:e1 IEEE 802.11: associated

That is the excerpt from the syslog at the time of failure. Is it possible that perhaps some other service is failing and causing HOSTAPD to stop working? Most of the time the WAP will still be visible even though it is not allowing clients. Some time it is not visible. Some times it is not visible after reboot, but it does appear as "Other Network". Windows makes me type in the SSID and the password even though it already has it. When it fails both HOSTAPD and DNSMASQ say they are active (Running). Does anyone have any ideas?

Edit 1 - I added syslog level 0 and rebooted. My devices were connected for roughly 7 hours and then dropped. I have the logs that I have edited for easier reading.

IEEE 802.11: associated
WPA: event 1 notification
WPA: start authentication
IEEE 802.1X: unauthorizing port
WPA: sending 1/4 msg of 4-Way Handshake
WPA: EAPOL-Key timeout
WPA: sending 1/4 msg of 4-Way Handshake
WPA: EAPOL-Key timeout
WPA: sending 1/4 msg of 4-Way Handshake
WPA: EAPOL-Key timeout
WPA: sending 1/4 msg of 4-Way Handshake
WPA: EAPOL-Key timeout
WPA: PTKSTART: Retry limit 4 reached
IEEE 802.1X: unauthorizing port
IEEE 802.11: deauthenticated due to local deauth request
IEEE 802.11: disassociated

Clearly this indicates that it thinks it did not receive the response, but this does not happen on the cisco wap, why would this work for 7 hours? does this indicate something else?

Goff
  • 261
  • 1
  • 2
  • 7

1 Answers1

0

I had a similar problem and it was due to lack of signal strength, are you using an external antenna, give it a try it worked for me.

sparks
  • 166
  • 4