[OpenWrt-Devel] Loss of connectivity on ADSL reset

Tim Coote tim+ietf.org at coote.org
Thu Jan 12 05:24:45 EST 2017


Hullo

I’ve tried this request to openwrt-users and got no response, so I’m trying the developer community in the hope that you’ve got a deeper understanding of how openwrt should work.

I keep losing my openwrt router’s IPv4 configuration when my ISP bounces the link to my home.

I’m using commit 1a7b132013 (Jan 9 2016)

My ISP link is over ADSL. I have a dsl modem attached to the phone line and a WRT1200 attached to that, which connects with pppoe. I’m trying to run homenet, so I’d like to have the WAN interface controlled by homenet (hence, I’m avoiding interfaces called ‘wan’ or ‘lan’ to avoid them being treated as ’special'). Based on advice that I cannot immediately put my finger on, I’m using these stanzas to set up the WAN links:

config interface 'e0'
	option ifname 'eth0'
	option proto 'pppoe'
	option username ‘<login username>'
	option password ‘<login password>'

config interface 'e0ext'
	option ifname 'pppoe-e0'
	option proto 'hnet'
	option mode 'external'
	option _orig_ifname 'pppoe-e0’ # I think that this line is redundant.


The idea is that e0ext is ‘stacked’ on pppoe-e0 (the interface name from e0).

The configuration works most of the time. Until the link bounces. What seems to happen is that pppoe is torn down when the link is reset, and then rebuilt correctly. However, the interface e0ext does not seem to receive any notification of the link coming back up (I’m assuming that’s how things are supposed to work).  Hnet is designed to keep the ipv6 links up in the event of loss of internet, so I’m assuming that this doesn’t require notification of link restoration.

I’m guessing that the events on the links/interfaces should be propagated through ubus (?)

If I use ubus to take down e0, e0ext is disabled, but on sending up to e0, e0ext stays down. An explicit up to e0ext creates this error (from ubus call network.interface.e0ext status):

root at OpenWrt:~# ubus call network.interface.e0ext status
{
	"up": false,
	"pending": false,
	"available": false,
	"autostart": true,
	"dynamic": false,
	"proto": "hnet",
	"data": {
		
	},
	"errors": [
		{
			"subsystem": "interface",
			"code": "NO_DEVICE"
		}
	]
}



Am I missing something from the configuration, have I stumbled upon a bug, or do I have to live with rebooting the router ~once per week?  (and am I posting to the right group?)

cheers
Tim
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.infradead.org/pipermail/openwrt-devel/attachments/20170112/32b5eda3/attachment.htm>
-------------- next part --------------
_______________________________________________
openwrt-devel mailing list
openwrt-devel at lists.openwrt.org
https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel


More information about the openwrt-devel mailing list