Add Peak EWMA load balancer (contrib)

[frittentheke]

What this PR does / why we need it:
This LB algorithm uses the peak exponentially-weighted moving average based on round-trip time (RTT) and outstanding requests to prefer targets resulting in low request latency and to respond to changes quickly. Peak EWMA is also well-suited for cross-data-center routing: it naturally prefers upstream hosts in the closest data center, but seamlessly fails over to other data centers during slowdowns (and fails back when performance recovers). In scenarios where all upstream hosts have similar request latency, Peak EWMA behaves equivalently to equal-weighted least request load balancing (using P2C selection).

Adding this contrib extension allows for evaluation of the potential for Istio service-meshes via EnvoyFilters.

Which issue this PR fixes

• Scope for dynamic probe / weight based loadbalancing algorithm as an alternative to RR and low RIF istio#54782 by @someshkoli (maybe)

Special notes for your reviewer:
This is simply adding the great work of @rroblak to the Envoy build that is Istio-Proxy.
See his feature issue and PRs

• Peak EWMA load balancing envoyproxy/envoy#20907
• contrib: implement Peak EWMA load balancing policy envoyproxy/envoy#40653

or maybe https://linkerd.io/2016/03/16/beyond-round-robin-load-balancing-for-latency/ for some background on why this algo is a great addition.

There also other Issues/PRs for other proxies discussing or adding it:

HAProxy

• New Balancing algorithm (Peak) EWMA haproxy/haproxy#1570
• Server weight modulation based on smoothed average measurement haproxy/haproxy#1977

NGINX

• https://github.com/kubernetes/ingress-nginx/blob/main/rootfs/etc/nginx/lua/balancer/ewma.lua

Linkerd

• https://linkerd.io/2.19/features/load-balancing/

Alibaba Cloud

• https://www.alibabacloud.com/help/en/asm/user-guide/load-balancing-based-on-workload-latency-using-exponentially-weighted-moving-average-ewma
• https://www.alibabacloud.com/blog/601610

[istio-testing]

Hi @frittentheke. Thanks for your PR.

I'm waiting for a istio member to verify that this patch is reasonable to test. If it is, they should reply with /ok-to-test on its own line. Until that is done, I will not automatically test new commits in this PR, but the usual testing commands by org members will still work. Regular contributors should join the org to skip this step.

Once the patch is verified, the new status will be reflected by the ok-to-test label.

I understand the commands that are listed here.

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository.