Questions[iproute2](server)

Linux Kernel before 3.6 used route caching to do IPv4 multipath routing, which meant routing between two separate lines/ISPs was quite easy. From 3.6 the algorithm changed to being per-packet, meaning that some route table/rule/iptables marker tricks were required to achieve the two lines/ISPs.

However, if you had two lines with the same ISP who could route a single IP down both lines on a per-packet basis in a balanced/failover fashion, then from 3.6 you could easily achieve line bonding (at the IP level) because of the per-packet routing in both directions.

From 4.4, the kernel changed again to flow-based load balancing based on a hash over the source and destination addresses.

I am currently running Kernel 4.4.36, and am using multipath routing over PPPoE connections. My downstream traffic from the ISP is routed across the two separate lines on a per-packet basis (one IP routed down both lines). This gives me a download speed faster than the speed of one individual line. Nearly the speed of both lines added together. It works really well, Skype video, VoIP (UDP), YouTube etc. all work great.

Because of having such a good downstream experience I want to try it upstream but my upstream traffic is routed according to the newer flow-based algorithm across both ppp devices (which have the same IP address). This means that I cannot achieve an upload speed that is faster than the speed of a single line.

Is there a way to configure the current Kernel to use the per-packet algorithm? Or some other method to achieve per-packet multipath routing? Would I need to revert to an older Kernel (which I don't want to do for various other reasons)?

My ISP does not support multi-link ppp.

In case it is relevant, I am currently running Arch Linux ARMv7 on a Raspberry Pi 3.

If I move an interface temporarily into a netns with

ip link set eth10 netns myns

then it no longer is visible in the root, only within the namespace myns.

How do I move it back, something like (these obviously don't exist):

ip link unset eth10

or perhaps

ip link set eth10 netns root

or similar?

I've been working with custom route tables on Linux, and I'm a bit confused by some of the documentation and behavior of the "ip route" command. It seems that the only valid values should be 0-255 plus the names defined in /etc/iproute2/rt_tables:

255 local
254 main
253 default
0   unspec

This would leave 1-252 for custom tables. Attempting to use an undefined table name gives an error:

$ ip route show table kermit
Error: argument "kermit" is wrong: table id value is invalid

However, it seems that I can use numbers far higher than 255 without error:

$ ip route show table 1000
[no output]
$ ip route add 10.10.10.0/24 dev eth0 table 1000
[no output]
$ ip route show table 1000
10.10.10.0/24 dev eth0  scope link

At some point, things get even weirder. Right at maxint (2^31), it "overflows" into the local table (255):

$ ip route show table 2147483647
[no output]
$ ip route show table 2147483648
[exact output of table 255 (local)]

Can anyone explain what is happening? Are there actually maxint custom routing tables that can be used?

Suppose you've configured IPv6 proxy NDP on one of your Linux systems like so:

ip -6 neighbor add proxy 2001:db8:1234::5 dev eth1

How do you verify that the configuration took? ip -6 neighbor show doesn't appear to show proxy entries and ip -6 neighbor show proxy isn't a supported command.

We have 2 different ISP connections. Our previous "IT guy" setup our firewall like so:

When /etc/rc.local was executed on startup, it did a bunch of ip rule add and ip route add commands in order to route certain internal hosts to use certain ISP connections.

Then at the end of /etc/rc.local, he executed our iptables firewall rules that were generated by Firewall Builder. These iptables rules have both Policy and NAT rules setup in them.

What I don't understand, is why did he use iproute2 to specify rules and routes but also specify NAT rules for iptables? Why didn't he just do it all in one or the other instead of using them both? Could he have got rid of the iproute2 rules and routes and just put all those same rules into the iptables NAT settings?