From the blog.

Managing Digital Racket
The more I tune out, the less I miss it. But that has presented me with some complex choices for a nuanced approach to curb

NMC DOiT Vol.2 Scenario 7 Day 4-4 – RIPng Unequal-Cost Load-Balancing

227 Words. Plan about 1 minute(s) to read this.

I wasn’t going to blog about this scenario anymore, but this one issue is so freaky, I just have to. The scenario asked for RIPng to load-balance IPv6 traffic over the ethernet and serial links in a 2:1 ratio. So, for every packet routed over the serial, 2 would go over the ethernet link. My brain pondered this, and after maybe 3 seconds of contemplation, I didn’t even try to figure it out. I didn’t know of any feature that would allow me to do anything other than equal-cost load-balancing, so I skipped right to the answer key. And, in truth, there is no such thing other than equal-cost load-balancing. This is one of those “think outside the box” solutions.

The answer is to create a tunnel that goes over the ethernet link, run RIPng over it, and tweak the metric with “metric-offset” so that the tunnel comes up as an equal-cost path along with the ethernet and serial. So, RIP will think it has 3 paths – the tunnel, the ethernet, and the serial. 1 packet will go over the ethernet encapsulated as a tunnel packet. 1 will go over ethernet natively. 1 will go over serial. Thus, a 2:1 ratio in favor of the ethernet.

If I ever saw someone trying to put this into production on my network at work, I’d kill them outright, and wouldn’t feel bad about having done so.

Ethan Banks writes & podcasts about IT, new media, and personal tech.
about | subscribe | @ecbanks