bcavagnolo

from elasticsearch import Elasticsearch, Urllib3HttpConnection
from pprint import pprint

class CompressedHttpConnection(Urllib3HttpConnection):

    def __init__(self, *args, **kwargs):
        super(CompressedHttpConnection, self).__init__(*args, **kwargs)
        self.headers = {'Accept-Encoding': 'gzip,deflate'}

    def perform_request(self, *args, **kwargs):

bcavagnolo

#!/usr/bin/env bash
set -e
set -x

[ "${DYNAMODB_HOME}" = "" ] && DYNAMODB_HOME=/opt/dynamodb
DYNAMODB_FILE=dynamodb_local_latest.tar.gz
DYNAMODB_URL=http://dynamodb-local.s3-website-us-west-2.amazonaws.com/${DYNAMODB_FILE}
DYNAMODB_JAR=DynamoDBLocal.jar
DYNAMODB_LOG=/var/log/dynamodb.log

bcavagnolo

import unittest
import os

FOO_IS_ON=os.environ.get('FOO')


def seriously_skip_this_if(condition, reason):
    '''skip the test class (including the setUpClass method

    nose (and friends) will run your setUpClass and tearDownClass methods even

bcavagnolo

/*
 * Derived from debugfs.c in e2fsprogs and inspired by:
 * http://unix.stackexchange.com/questions/50177/birth-is-empty-on-ext4/50184
 *
 * Copyright (C) 1993 Theodore Ts'o.  This file may be redistributed
 * under the terms of the GNU Public License.
 *
 * Modifications by Robert Sanders <gt8134b@prism.gatech.edu>
 *
 * To build:

bcavagnolo

Asset Store Coding Challenge

The doves are loose and lots of other satellites are showing up to the party. Antennas to communicate with the ever expanding fleet are popping up like wildflowers. To keep track of them all, we need an asset store. We've already worked out the initial requirements and made a few design decisions:

• The asset store is going to be exposed as a custom RESTful web API written in python.

• Every asset that we store has an "asset name". There are some rules about asset names:

  • They must be globally unique across all assets

• They can only contain alphanumeric ascii characters, underscores, and dashes

bcavagnolo

# Most of the kinesis examples out there do not seem to elucidate the
# opportunities to parallelize processing of kinesis streams, nor the
# interactions of the service limits. You may want to start your
# journey by familiarizing yourself with the concepts (e.g., what is a
# shard?) and the service limits:
# http://docs.aws.amazon.com/streams/latest/dev/service-sizes-and-limits.html

# The idea is to spawn a process per shard in order to maximize
# parallelization while respecting the service limits. If you run
# multiple instances of this script (or equivalent) you will exhaust

bcavagnolo

add_ons:
  cni:
    disable: false
    options:
      calico:
        felix_input_mtu: 1440
        log_level: info
        mode: overlay
        workload_mtu: 1500
    provider: calico