tshak

const std = @import("std");

pub fn main() !void {
    var list = std.ArrayList(u21).init(std.heap.page_allocator);
    try list.append(1);

    list.deinit();

    // Seg fault :(
    std.debug.print("{}", .{list.items[0]});

tshak

apiVersion: serving.knative.dev/v1
kind: Configuration
metadata:
  name: testdummy
  namespace: apps
spec:
  template:
    metadata:
      name: testdummy
    spec:

tshak

# Merges a kube config into your existing kube config
# Based on https://github.com/kubernetes/kubernetes/issues/46381#issuecomment-461404505
kmerge() {
  if [[ -z "$1" ]]; then
    echo "usage 'kmerge newKubeConfig'"
    exit 1
  fi

  cp ~/.kube/config ~/.kube/config.old
  KUBECONFIG=~/.kube/config:$1 kubectl config view --flatten > ~/.kube/mergedkub && mv ~/.kube/mergedkub ~/.kube/config

tshak

Name:                      voyager-repro-5c65d99685-f84xr
Namespace:                 default
Node:                      ip-172-20-119-215.us-east-2.compute.internal/172.20.119.215
Start Time:                Tue, 02 Oct 2018 15:07:51 +0200
Labels:                    origin=voyager
                           origin-api-group=voyager.appscode.com
                           origin-name=repro
                           pod-template-hash=1721855241
Annotations:               ingress.appscode.com/last-applied-annotation-keys=
                           kubernetes.io/limit-ranger=LimitRanger plugin set: cpu request for container haproxy

tshak

apiVersion: v1
kind: Service
metadata:
  labels:
    app: repro
  name: repro
  namespace: default
spec:
  ports:
  - port: 8080

tshak

Introduction

The following is a very rough set of benchmarks to get an initial sense of the performance difference between up with the nodejs shim and with using native Go support in Lambda. While not rigorous, the summary conclusion is that other than warmup there is not a sigdifference between the two to warrent more rigorous testing.

Methodology

This test utilized up 0.5.4 as well as a custom version that only utilizes native Go support in Lambda. The testing was done with the simple Golang example app found in the documentation in eu-west-3. An EC2 instance in the same region ran various tests using Vegeta using stock AWS Ubuntu 16.04 image with the ulimit increated to 65000. After a clean up deploy, a 3 second warmup at 10RPS

tshak

Keybase proof

I hereby claim:

• I am tshak on github.
• I am tshak (https://keybase.io/tshak) on keybase.
• I have a public key ASBd6ErPjb0hcj3Wa1f-Kf6BFXzZV6x81zkzDmqDrA9tXAo

To claim this, I am signing this object: