~andreserl/pyjuju/bootstrap-cobbler : changes

~andreserl/pyjuju/bootstrap-cobbler » Changes from revision 299

From Revision 299 to 280

Rev	Summary	Authors	Date
299	Adding Initial Cobbler Bootstrap Support Adding Initial Cobbler Bootstrap Support	Andres Rodriguez	12 years ago
298	renamed cobblers to orchestras renamed cobblers to orchestras	William Reade	12 years ago
297	unsaved file :/ unsaved file :/	William Reade	12 years ago
296	do-nothing implementations of the classes we'll need to boot... do-nothing implementations of the classes we'll need to bootstrap	William Reade	12 years ago
295	merge from parent merge from parent	William Reade	12 years ago
294	added get_zookeeper_machines to cobbler.MachineProvider added get_zookeeper_machines to cobbler.MachineProvider	William Reade	12 years ago
293	merge latest trunk merge latest trunk	William Reade	12 years ago
292	fixes per Gustavo's review fixes per Gustavo's review	William Reade	12 years ago
291	* deleted failing EC2 bootstrap test; replaced it with a mor... * deleted failing EC2 bootstrap test; replaced it with a more comprehensive and general on in common	William Reade	12 years ago
290	* gutted evil on-demand-subclassing of LaunchMachine: everyt... * gutted evil on-demand-subclassing of LaunchMachine: everything we need to do can be handled with a pair of optional callbacks to LaunchMachine.run * reinstated EC2-specific bootstrap operation * added EC2ProviderMachine tests I failed to remember earlier * rearranged launch/bootstrap tests somewhat, to better fit new division of responsibilities (WIP; IMO not a blocker)	William Reade	12 years ago
289	* realised "instance" is not the same as "machine" * realised "instance" is not the same as "machine" * added ec2.machine.machine_from_instance * ProviderMachine now has .private_dns_name and .launch_time attributes (.launch_time is a datetime.datetime object) * changed some tests which assumed .launch_time to be a timestamp * LaunchMachine.start_machines implementations should now return a list of Machines, not instance_ids; consequently simplified launch callbacks * MachineProvider.get_zookeeper_machines should now also return a lsit of Machines, not instances, and we therefore no longer need LaunchMachine.get_zookeeper_hosts (because those returned ProviderMachines should all have .private_dns_names, and therefore don't have to be specially handled by different providers) * in fact there should be no naked instances accessible from the public API any more	William Reade	12 years ago
288	vars chain test makes more sense now vars chain test makes more sense now	William Reade	12 years ago
287	fixed logging tests fixed logging tests	William Reade	12 years ago
286	fixed trivial points in gustavo's review; added some tests f... fixed trivial points in gustavo's review; added some tests for Launchmachine; broke tests which use logging	William Reade	12 years ago
285	realised 169.254.169.254 is kinda EC2-specific, extracted realised 169.254.169.254 is kinda EC2-specific, extracted get_instance_id_cmd to LaunchMachine	William Reade	12 years ago
284	* moved bootstrappy stuff from common.launch into .bootstrap * moved bootstrappy stuff from common.launch into .bootstrap * added attributes to ec2.MachineProvider that make the EC2Bootstrap operation redundant: shortly everything will be able to just use Bootstrap (dummy provider may need some work)	William Reade	12 years ago
283	renamed get_machine_options to get_launch_options, because I... renamed get_machine_options to get_launch_options, because IMO it reveals intention better	William Reade	12 years ago
282	docs; occasional name changes docs; occasional name changes	William Reade	12 years ago
281	fix necessary import fix necessary import	William Reade	12 years ago
280	fix unnecessary import fix unnecessary import	William Reade	12 years ago

Older »