John lgastako

All You Zombies

2217 Time Zone V (EST) 7 Nov. 1970--NTC-- "Pop's Place": I was polishing a brandy snifter when the Unmarried Mother came in. I noted the time---10:17 P. M. zone five, or eastern time, November 7th, 1970. Temporal agents always notice time and date; we must.

The Unmarried Mother was a man twenty--five years old, no taller than I am, childish features and a touchy temper. I didn't like his looks---I never had---but he was a lad I was here to recruit, he was my boy. I gave him my best barkeep's smile.

Maybe I'm too critical. He wasn't swish; his nickname came from what he always said when some nosy type asked him his line: "I'm an unmarried mother." If he felt less than murderous he would add: "at four cents a word. I write confession stories."

If he felt nasty, he would wait for somebody to make something of it. He had a lethal style of infighting, like a female cop---reason I wanted him. Not the only one.

Don't use MongoDB

I've kept quiet for awhile for various political reasons, but I now feel a kind of social responsibility to deter people from banking their business on MongoDB.

Our team did serious load on MongoDB on a large (10s of millions of users, high profile company) userbase, expecting, from early good experiences, that the long-term scalability benefits touted by 10gen

Yesterday I upgraded our running elasticsearch cluster on a site which serves a few million search requests a day, with zero downtime. I've been asked to describe the process, hence this blogpost.

To make it more complicated, the cluster was running elasticsearch version 0.17.8 (released 6 Oct 2011) and I upgraded it to the latest 0.19.10. There have been 21 releases between those two versions, with a lot of functional changes, so I needed to be ready to roll back if necessary.

Our setup:

elasticsearch

We run elasticsearch on two biggish boxes: 16 cores plus 32GB of RAM. All indices have 1 replica, so all data is stored on both boxes (about 45GB of data). The primary data for our main indices is also stored in our database. We have a few other indices whose data is stored only in elasticsearch, but are updated once daily only. Finally, we store our sessions in elasticsearch, but active sessions are cached in memcached.

	# 5 minutes Lisp in Python
	# Pepijn de Vos <http://pepijndevos.nl>
	#
	# Inspired by 30 minutes Lisp in Ruby
	# http://gist.github.com/562017
	#
	# This Lisp does not read or parse anything at all.
	# A generator and a Decorator are abused to run sexps.
	#
	# Usage:

	VOICE="Vicki"
	BALANCE="100"

	function voice_prompt () {
	echo " SPEAKING: $1"
	say -v "$VOICE" "$1"
	}

	function get_user_keypress () {
	KEY=""

	<!DOCTYPE html>
	<html>
	<head>
	<meta charset="utf-8" />
	<title>Canvas Hexagonal Map</title>
	<style type="text/css">
	canvas {
	border:0;
	display:block;
	margin:0 auto;

	package tests

	import (
	"testing"
	)

	type S struct {
	t *testing.T
	}

	;;5-minute sketch of tagged literals in JSON via the little known second argument to the browser's JSON.parse.
	;;It's no EDN, but it's still a hell of a lot nicer than the usual JSON-deserialization state of affairs.


	(ns shim.json
	(:require [goog.date.DateTime :as DateTime]))

	(defn parse-tagged-string
	"Parse a tagged string within JSON.
	Assumes string begins with a throwaway first character and there is a single space between the tag and its data."

	;;framework.clj
	;;...
	(definterface ES
	(putIfAbsent [k v])
	(rep [k old-val new-val])
	(get [k]))

	(defn ^ES jes []
	(let [m (atom {})]
	(reify ES

	-- You need to install IPPrint and HsColour with cabal to use this config.
	-- If you find all the loading messages anoying you may want to set an alias:
	-- alias ghci="ghci -v0"
	-- Or if you are using stack ghci:
	-- alias ghci="stack --silent ghci --ghci-options -v0"

	:set -XNoMonomorphismRestriction
	:set prompt "\ESC[34mλ\ESC[m "

	-- remove the following comments if ghci >= v7.8.1 for multiline support: