Benchmark results of Kubernetes network plugins over 10Gb/s network

Re:Benchmark results of Kubernetes network plugins over 10Gb... by londons_explore on Friday November 30, 2018 @10:13AM (#18569623)

All these 'plugins' are simply scripts and tools which configure the linux kernel iptables and network forwarding rules right?

None of them actually handle packets, connections, routing etc directly I assume?

Given that, it seems disingenuous to benchmark them... You're simply benchmarking the Linux Kernel vs the Linux Kernel.

[ Parent | on HN ]

Re:Benchmark results of Kubernetes network plugins over 10Gb... by whalesalad on Friday November 30, 2018 @10:22AM (#18569691)

At the end of the day you’re benchmarking a strategy. It doesn’t matter that Linux is the underlying thing executing that strategy. All of these tools have branding and marketing around them... It’s important to see which ones live up to claims.

[ Parent | on HN ]

Re:Benchmark results of Kubernetes network plugins over 10Gb... by detaro on Friday November 30, 2018 @10:53AM (#18569941)

They do not all work the same way, and using them clearly produces in different results, so why is it disingenuous to compare what performance you get with each of them, even if all they did was "configure the kernel" (which afaik isn't the case)?

[ Parent | on HN ]

Re:Benchmark results of Kubernetes network plugins over 10Gb... by gauravphoenix on Friday November 30, 2018 @10:50AM (#18569910)

No. This is not all true. Many tools use encapsulation of packets and do not modify netfilter or routing tables.

[ Parent | on HN ]

Re:Benchmark results of Kubernetes network plugins over 10Gb... by dharmab on Friday November 30, 2018 @11:21AM (#18570161)

No. For example, Cilium uses eBPF for L4 routing/filtering and Envoy for some L7 filtering.

[ Parent | on HN ]