Re: routing changes from Sept 7th and possible ppp breakage (RELENG_13)
- In reply to: mike tancsa : "Re: routing changes from Sept 7th and possible ppp breakage (RELENG_13)"
- Go to: [ bottom of page ] [ top of archives ] [ this month ]
Date: Fri, 10 Sep 2021 16:01:47 UTC
On 9/10/2021 11:52 AM, mike tancsa wrote: > On 9/10/2021 11:34 AM, Helge Oldach wrote: >> Zhenlei Huang wrote on Fri, 10 Sep 2021 10:34:20 +0200 (CEST): >>>> On Sep 10, 2021, at 5:38 AM, mike tancsa <mike@sentex.net> wrote: >>>> default 10.255.255.1 UGS 6 1500 igb0 >>>> default 10.1.0.1 UGS 17 1500 tun10 >>>> >>>> Where as now, it doesnt seem to replace it, it seems to just add >>>> another default route. >>> At first glance it seems that ppp does not check existing default >>> route before replacing it with a new one. >> My thinking as well. Similar thing (creating kind-of an overlay network) >> with openvpn results in a single default route - apparently the original >> one is properly removed before adding openvpn's own default route. >> > One other thing I noticed is that net.route.multipath is RO on i386 but > RW on AMD64. Is that intentional for some reason ? > Actually, its neither. On my i386 image I was using an old kernel def that didnt have options ROUTE_MPATH yet the sysctl still shows up with a value of 1 # sysctl net.route.multipath net.route.multipath: 1 # sysctl -w net.route.multipath=0 sysctl: oid 'net.route.multipath' is read only # ---Mike