TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Container Networking with Vxlan, BGP and WireGuard

74 pointsby tobbybalmost 6 years ago

6 comments

KaiserProalmost 6 years ago
One does not simply go from a flat network to overlays. Overlays are slow, difficult, cause really odd failures and are often hilariously immature. They are the experimental graph database of the network world.<p>Just have a segregated network, and let the VPC&#x2F;dhcp do all the hard stuff.<p>Have your hosts on the default VLAN(or Interface if your cloudy), with its own subnet (Subnets should only exist in one VLAN.) Then if you are in cloud land, have a second network adaptor on a different subnet. If you are running real steel, then you can use a bonded network adaptor with multiple VLANs on the same interface. (The need for a VLAN in a VPC isn&#x27;t that critical because there are other tools to impose network segregation.)<p>Then use macvtap, or macvlan(or which ever thing that gives each container a macaddress) to give each container its own IP. This means that your container is visible on that entire subnet, either inside the host or without.<p>There is no need to faff with routing, it comes for free with your VPC&#x2F;network or similar. Each container automatically has a hostname, IP, route. It will also be fast. As a bonus it call cane be created at the start using cloudformation or TF.<p>You can have multiple adaptors on a host, so you can separate different classes of container.<p>Look, the more networking that you can offload to the actual network the better.<p>If you are ever re-creating DHCP&#x2F;routing&#x2F;DNS in your project, you need to take a step back and think hard about how you got there.<p>70% of the networking modes in k8s are batshit insane. a large amount are basically attempts at vendor lock in, or worse someone&#x27;s experiment thats got out of hand. I know networking has always been really poor in docker land, but there are ways to beat the stupid out of it.<p>The golden rule is this:<p>Always. Avoid. Network. Overlays.
评论 #20603454 未加载
评论 #20604522 未加载
评论 #20603840 未加载
exabrialalmost 6 years ago
Site is having issues atm... but I&#x27;ll throw something out there I&#x27;d really like to see.<p>We encrypt 100% of our machine-to-machine traffic at the TCP level. There&#x27;s a lot of shuffling of certs around to get some webapp to talk to postgres, then have that webapp serve https to haproxy, etc.<p>I&#x27;d be awesome if there was a way your cloud servers could just talk to each other using wiregaurd by default. We looked at setting it up, but it&#x27;d need to be automated somehow for anything above a handful of systems :&#x2F;
评论 #20601447 未加载
评论 #20602048 未加载
评论 #20600951 未加载
评论 #20601656 未加载
j0057almost 6 years ago
In my mind, a &quot;layer 2 subnet&quot; really doesn&#x27;t mean anything. Subnets are things that happen in IP, that is, layer 3, and layer 2 is the physical connection, ie. Ethernet or WLAN, which don&#x27;t have the concept of subnets.<p>Edit: also the OSI layer model was specified in the eighties, and isn&#x27;t all that accurate in 2019 to describe how our networks actually work.
评论 #20602034 未加载
chaz6almost 6 years ago
Can we have a version using IPv6 instead of legacy IPv4? It would make things a lot simpler (no need for any fancy routing or nat).
评论 #20600908 未加载
corndogealmost 6 years ago
This article uses Quagga - they really should be using FRRouting, which was forked from Quagga in 2017 by the core Quagga developers and has 4 times as many commits (16000[0] vs 4000[1]), far more features, bugfixes, etc. Quagga has been dead for over a year.<p>[0] <a href="https:&#x2F;&#x2F;github.com&#x2F;FRRouting&#x2F;frr" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;FRRouting&#x2F;frr</a><p>[1] <a href="http:&#x2F;&#x2F;gogs.quagga.net&#x2F;Quagga" rel="nofollow">http:&#x2F;&#x2F;gogs.quagga.net&#x2F;Quagga</a>
评论 #20601974 未加载
评论 #20601085 未加载
评论 #20601376 未加载
alexandre_malmost 6 years ago
&quot;Vxlan uses multicast which is often not supported on most cloud networks. So its best used on your own networks.&quot;<p>Not entirely correct.<p>Linux has had unicast vxlan for quite some time.<p>Flannel is doing unicast and works pretty much anywhere.<p>See &quot;Unicast with dynamic L3 entries&quot; section: <a href="https:&#x2F;&#x2F;vincent.bernat.ch&#x2F;en&#x2F;blog&#x2F;2017-vxlan-linux" rel="nofollow">https:&#x2F;&#x2F;vincent.bernat.ch&#x2F;en&#x2F;blog&#x2F;2017-vxlan-linux</a>
评论 #20604506 未加载