Can't connect to intranet/isolated access points
asiag0 opened this issue · comments
I know there's an old closed ticket about this, but I don't see any kind of resolution on that ticket.
My use case is to be able to connect to an access point that does not have internet access at all; not a captive portal. I have a raspberry pi set up as a webserver to be used as data storage and access from isolated locations (such as a cabin in the backwoods where there's not even cellular coverage). One feature of this device would be a webpage with links to epubs that would be accessible from the e-reader's beta-feature browser as a simple and convenient method of side-loaded said epubs.
The server works perfectly when attached to a internet-connected access point, of course. Maybe we could have a setting accessible from the Settings window or in the config file to allow for this kind of functionality?
I don't remember the details, but you can patch the firmware to change the url used to check for internet access.
My H2O also has some problems to connect to wifi. I got some logs as follow:
Jun 7 02:09:47 dhcpcd-dbus: init completed, waiting for events
Jun 7 02:09:48 dhcpcd-dbus: scan results on interface eth0
Jun 7 02:09:48 wpa_supplicant[1146]: eth0: Trying to associate with d8:c8:e9:1a:0d:48 (SSID='666666' freq=2452 MHz)
Jun 7 02:09:48 wpa_supplicant[1146]: Failed to add supported operating classes IE
Jun 7 02:09:48 wpa_supplicant[1146]: eth0: Associated with d8:c8:e9:1a:0d:48
Jun 7 02:09:49 wpa_supplicant[1146]: eth0: WPA: Key negotiation completed with d8:c8:e9:1a:0d:48 [PTK=CCMP GTK=TKIP]
Jun 7 02:09:49 wpa_supplicant[1146]: eth0: CTRL-EVENT-CONNECTED - Connection to d8:c8:e9:1a:0d:48 completed [id=0 id_str=]
Jun 7 02:09:49 dhcpcd[1160]: eth0: carrier acquired
Jun 7 02:09:49 dhcpcd[1160]: eth0: executing/libexec/dhcpcd-run-hooks' CARRIER Jun 7 02:09:49 dhcpcd[1160]: eth0: delaying IPv4 for 0.6 seconds Jun 7 02:09:49 dhcpcd-dbus: event on interface eth0 (CARRIER) Jun 7 02:09:49 dhcpcd[1160]: eth0: using ClientID 01:70:25:59:e1:7f:f4 Jun 7 02:09:49 dhcpcd[1160]: eth0: reading lease/var/db/dhcpcd-eth0-666666.lease'
Jun 7 02:09:49 dhcpcd[1160]: eth0: rebinding lease of 192.168.2.222
Jun 7 02:09:49 dhcpcd[1160]: eth0: sending REQUEST (xid 0x4af6b056), next in 3.4 seconds
Jun 7 02:09:49 dhcpcd[1160]: eth0: acknowledged 192.168.2.222 from 192.168.2.1
Jun 7 02:09:49 dhcpcd[1160]: eth0: leased 192.168.2.222 for 86400 seconds
Jun 7 02:09:49 dhcpcd[1160]: eth0: renew in 43200 seconds, rebind in 75600 seconds
Jun 7 02:09:49 dhcpcd[1160]: eth0: writing lease/var/db/dhcpcd-eth0-666666.lease' Jun 7 02:09:49 dhcpcd[1160]: eth0: adding IP address 192.168.2.222/24 Jun 7 02:09:49 dhcpcd[1160]: eth0: adding route to 192.168.2.0/24 Jun 7 02:09:49 dhcpcd[1160]: eth0: adding default route via 192.168.2.1 Jun 7 02:09:49 dhcpcd[1160]: eth0: executing/libexec/dhcpcd-run-hooks' REBOOT
Jun 7 02:09:49 dhcpcd-dbus: event on interface eth0 (REBOOT)
Jun 7 02:09:49 dhcpcd-dbus: status changed to connected
Jun 7 02:09:49 fickel: bool FickelService::Start(const QString&) "eth0"
Jun 7 02:10:21 nickel: ( 459.681 @ 0x76f5d0 / packetdump.warning) "https://storeapi.kobo.com/v1/products/featured/81844d90-8f90-40b0-98e4-f0cb6c3534e7?page_index=0&page_size=2700&sort=Temperature" => "Operation canceled"
Jun 7 02:10:21 nickel: ( 459.684 @ 0x76f5d0 / packetdump.warning) HTTP Status Code: 0
Jun 7 02:10:21 nickel: ( 459.687 @ 0x76f5d0 / packetdump.warning) Error: 5
Jun 7 02:10:21 nickel: ( 459.688 @ 0x76f5d0 / packetdump.warning) Source: "https://storeapi.kobo.com/v1/products/featured/81844d90-8f90-40b0-98e4-f0cb6c3534e7?page_index=0&page_size=2700&sort=Temperature"
Jun 7 02:10:49 nickel: ( 487.781 @ 0x4c7898) static bool EntitlementManager::subscriptionExpired(const Entitlement&, const UsageData&) Check isSubscription first ""
Jun 7 02:10:50 nickel: ( 489.668 @ 0x76f5d0 / packetdump.warning) "https://storeapi.kobo.com/v1/products/featured/81844d90-8f90-40b0-98e4-f0cb6c3534e7?page_index=0&page_size=9&sort=Temperature" => "Socket operation timed out"
Jun 7 02:10:51 nickel: ( 489.672 @ 0x76f5d0 / packetdump.warning) HTTP Status Code: 0
Jun 7 02:10:51 nickel: ( 489.673 @ 0x76f5d0 / packetdump.warning) Error: 4
Jun 7 02:10:51 nickel: ( 489.674 @ 0x76f5d0 / packetdump.warning) Source: "https://storeapi.kobo.com/v1/products/featured/81844d90-8f90-40b0-98e4-f0cb6c3534e7?page_index=0&page_size=9&sort=Temperature"
Jun 7 02:10:51 nickel: ( 489.675 @ 0x76f5d0 / packetdump.warning) ""
Jun 7 02:11:16 fickel: bool FickelService::Stop() "eth0"
Jun 7 02:11:16 dhcpcd[1160]: received signal TERM from PID 1211, stopping
Jun 7 02:11:16 dhcpcd[1160]: eth0: removing interface
Jun 7 02:11:16 dhcpcd[1160]: eth0: executing `/libexec/dhcpcd-run-hooks' STOPPED
Jun 7 02:11:16 dhcpcd[1160]: exited
Jun 7 02:11:16 dhcpcd-dbus: lost connection to dhcpcd
Jun 7 02:11:16 dhcpcd-dbus: status changed to down
Jun 7 02:11:16 dhcpcd-dbus: status changed to down
Jun 7 02:11:16 wpa_supplicant[1146]: ioctl[SIOCSIWESSID]: Resource temporarily unavailable
Jun 7 02:11:16 wpa_supplicant[1146]: eth0: CTRL-EVENT-DISCONNECTED bssid=d8:c8:e9:1a:0d:48 reason=3 locally_generated=1
Jun 7 02:11:16 wpa_supplicant[1146]: ioctl[SIOCSIWESSID]: Resource temporarily unavailable
Jun 7 02:11:16 wpa_supplicant[1146]: eth0: CTRL-EVENT-TERMINATING
My H2O also has some problems to connect to wifi. I got some logs as follow:
Is your wifi network isolated? It's unclear from your dhcpcd logs.
The process is: when the kobo reader connects to an access point, it attempts to access a file on the internet to verify connectivity, and if it can't (as in the case of non-internet access points), it de-auths from the AP. Very clear that this is by design, and has a negative effect on (probably) an extremely small number of users (myself included), who wish to connect to intranet wifi that has no internet connectivity.
In my example, I wish to be able to hit the road and go to the cabin, for extended periods of time, and retain the flexibility of accessing my own collection of epubs without having to load them all onto the kobo. I don't want to bring a computer/laptop with me, hence the tiny battery-powered device.
I have found a workaround to the phone-home function, but it still would be nicer to not have to go to these extra efforts in order to get it to work, hence the request for an additional firmware issue.