Jesse Dearing jessedearing

Risky Business

Presenter

Jesse Dearing

Abstract

As we operate production systems they inevitably fail. As they fail and fail and fail again we as operators start to identify the way they fail and we know the causes of the failures. What if we could see the future and predict failures before they happen. We're

	{
	"basics": {
	"name": "Jesse Dearing",
	"label": "Staff SRE & Platform Engineer \| Kubernetes · AWS · Terraform · Go · Rust",
	"image": "https://jesse.dev/images/avatar-2019-05-31.png",
	"email": "me@jesse.dev",
	"website": "https://jesse.dev",
	"summary": "I've spent over a decade at the intersection of building software and keeping it running. I care deeply about both sides of that equation.\n\nMy focus is platform engineering and site reliability: designing the infrastructure, tooling, and systems that let product engineers ship fast, securely, and with full visibility into how their code is serving customers. I've worked across the full spectrum from bare-metal in data centers to multi-region cloud deployments. Software-defined infrastructure is what makes it possible to manage, automate, and observe systems at scale, and building that foundation is where I do my best work.\n\nI'm most effective in Senior and Staff IC roles. Where shaping architecture, unblocking teams, solving problem

	#!/usr/bin/env python3

	import base64
	import json
	from kubernetes import config, client, dynamic
	from cryptography import x509
	from cryptography.hazmat.backends import default_backend

	config.load_kube_config()
	dc = dynamic.DynamicClient(client=client.ApiClient())

	#!/usr/bin/env python3

	import base64
	import argparse
	from datetime import datetime, timedelta
	from cryptography import x509
	from cryptography.hazmat.backends import default_backend
	from kubernetes import client, config

	config.load_kube_config()

	package main

	import (
	"fmt"
	"os"
	"os/exec"
	"strconv"
	"strings"
	"time"
	)

	resource "aws_db_instance" "main" {
	allocated_storage = 10
	storage_type = "gp2"
	engine = "mysql"
	engine_version = "5.7"
	instance_class = "db.t2.micro"
	name = "main"
	username = "root"
	password = "${data.external.something_in_one_password.result.password}"
	parameter_group_name = "default.mysql5.7"

	data "external" "something_in_one_password" {
	program = ["${path.cwd}/bin/one_password.sh"]
	query = {
	vault = "Personal"
	uuid = "abc123"
	}
	}

	#!/usr/bin/env bash

	set -e

	# Eval so we can set the JSON values as shell variables
	eval "$(jq -r '@sh "VAULT=\(.vault) UUID=\(.uuid)"')"

	# Get the token or password from the first field of the second section (this is just how I store them, but you could use jq's select
	password=`op get item --vault=$VAULT $UUID \| jq -r '.details.sections[1].fields[0].v'`

	const express = require('express')
	const Webtask = require('webtask-tools')
	const bodyParser = require('body-parser')
	const MessagingResponse = require("twilio").twiml.MessagingResponse
	const request = require('request-promise-native')

	var app = express();

	app.use(bodyParser.json());