flupke

    def _get_next_or_previous_by_FIELD(self, field, is_next, **kwargs):
        if not self.pk:
            raise ValueError("get_next/get_previous cannot be used on unsaved objects.")
        op = is_next and 'gt' or 'lt'
        order = not is_next and '-' or ''
        param = smart_str(getattr(self, field.attname))
        q = Q(**{'%s__%s' % (field.name, op): param})
        q = q|Q(**{field.name: param, 'pk__%s' % op: self.pk})
        qs = self.__class__._default_manager.using(self._state.db).filter(**kwargs).filter(q).order_by('%s%s' % (order, field.name), '%spk' % order)
        try:

flupke

default = 'light gray'
strings = 'h217'
keywords = add_setting('h222', 'bold')
callables = 'h120'
breakpoints = 'h203'
comments = 'h116'

background = 'h236'
focused_background = 'h238'
current_background = 'h240'

flupke

#!py

import pprint
import json
import binascii
import functools
import traceback
import os
import os.path as op
import sqlite3

flupke

In the end, salt-cloud is just used to retrieve instances details from EC2 API.
You still need to setup a provider in salt for the region where your instances
are launched.

Here is what the reactor does when a minion comes up:

* wait for the new instances to ping the salt master when it receives
a notification from Amazon SNS
* use salt-cloud CloudClient to retrieve instance instance details from EC2 API
* stores the instance-id / minion-name association in a sqlite database