~themue/juju-core/024-cleanup-watcher : changes

~themue/juju-core/024-cleanup-watcher » Changes from revision 1262

From Revision 1262 to 1243

Rev	Summary	Authors	Date
1262	Move assignment policy to global env config Move assignment policy to global env config Previoysly, each provider defined it's own AssignmentPolicy, which was hard coded to AssignNew. Now, it's been made an env setting with default AssignNew. With containers, it will be possible to add a unit to an existing machine (if not dirty) so this work is a step in that direction. R=jameinel, fwereade CC= https://codereview.appspot.com/9824043	Ian Booth	10 years ago
1261	Add dirty flag to machine state Add dirty flag to machine state When choosing an unused machine to assign a unit to, a check was being made to see if there were any principal units for the machine. If machines had units assigned and then unassigned, the machine would appear clean since number of principal unit goes to 0 but the machine should still be considered dirty. The machine state has a "clean" flag (boolean). Machines are created with the clean flag set to true. It is set false when a machine has a unit assigned, and not reset when units are unassigned. The assign to unused machine logic is updated to look at the clean flag. R=fwereade, thumper CC= https://codereview.appspot.com/9820043	Ian Booth	10 years ago
1260	First part of the provisioner refactoring. First part of the provisioner refactoring. This branch breaks up the current provisioner and defines some interfaces that we'll be using for containers. The broker interface is what we have starting, stopping instances, and listing instances and related machines. An environment broker is written that defers the actual calls to the environ. The provisioner now creates a provisioning task for the environment provider using a machine watcher and the environment broker. The common provisioning methods are now in provisioner_task.go. I refactored some of the methods as we were duplicating a lot of calls, like get all instances, and then getting instances for individual ids. Also, looking up specific machines, and also getting all machines. Now we just do the "all" gets, and keep a map around for the duration of the checks. R=wallyworld, mue, fwereade, rog CC= https://codereview.appspot.com/9937046	Tim Penhey	10 years ago
1259	state: added CleanupWatcher state: added CleanupWatcher The CleanupWatcher signals the demand for the running of state.Cleanup(). It is the first in a row of CLs to add an according worker. R=fwereade, dimitern CC= https://codereview.appspot.com/10078043	Frank Mueller	10 years ago
1258	revert r1257 due to a failing test: revert r1257 due to a failing test: https://bugs.launchpad.net/juju-core/+bug/1188549 R=gz CC= https://codereview.appspot.com/10046047	Dimiter Naydenov	10 years ago
1257	Update add machine command to create containers Update add machine command to create containers The add machine command accepts additional command line arguments to specify a container to create. The container can be on a new machine or an existing instance. Container constraints are also supported eg create a container with 2G memory. The state is updated with the new machine document but the machine watcher ignores the machine since there's no code to provision containers yet. examples: juju add-machine /lxc juju add-machine --constraints mem=4G /lxc juju add-machine 1/lxc R=mue, fwereade, thumper CC= https://codereview.appspot.com/9886045	Ian Booth	10 years ago
1256	Have the verbose flag set the log level to INFO. Have the verbose flag set the log level to INFO. When the loggo branch landed yesterday, there was a change in behaviour with the --verbose flag, where it did log to stderr, but it was at the WARNING level, not INFO. R=fwereade, dimitern CC= https://codereview.appspot.com/9698047	Tim Penhey	10 years ago
1255	state/api: Split client Machiner into subpackage state/api: Split client Machiner into subpackage Following up on https://codereview.appspot.com/10044043/ this does the same for the client-side: splits machiner facade into its own pacakge. R=fwereade, rog CC= https://codereview.appspot.com/10026044	Dimiter Naydenov	10 years ago
1254	state/apiserver: Split Machiner into subpackage state/apiserver: Split Machiner into subpackage Minor refactoring to put server-side Machiner facade into its own subpackage as suggested previously. As a side-effect errors and interfaces used had to be moved into another "common" subpackage to resolve circular imports. R=fwereade, jameinel, rog CC= https://codereview.appspot.com/10044043	Dimiter Naydenov	10 years ago
1253	state/open.go: Remove RetryDelay state/open.go: Remove RetryDelay The latest version of mgo always delays 500ms between every retry. This would mean that we delay both that 500ms and the extra time that we had internally. Instead, this removes our internal delay, and the test that asserted we did not delay when it was a TLS failure rather than a connection failure. mgo does not appear to expose any way to control the retry delay, but it does seem close enough for what we wanted to do that we can live with it. R=gz, dimitern CC= https://codereview.appspot.com/9984046	John A Meinel	10 years ago
1252	state/apiserver: Split tests into multiple suites state/apiserver: Split tests into multiple suites Trivial change to split huge "suite" test suite into multiple sub-suites, reusing common code where possible. R=mue, rog CC= https://codereview.appspot.com/10003044	Dimiter Naydenov	10 years ago
1251	state/api: Client-side Machiner API facade state/api: Client-side Machiner API facade This adds client-side API Machiner facade and a lightweight proxy MachinerMachine, which implements the methods on state.Machine, as needed by the machiner worker, but using the facade internally for the actual calls. Also introduced client-side tests. Fixed and re-enabled a test I broke earlier in r1247. R=mue, rog, fwereade CC= https://codereview.appspot.com/9686047	Dimiter Naydenov	10 years ago
1250	Integrate loggo as a hierarchical logger. Integrate loggo as a hierarchical logger. The existing logging is very flat and has very little control over what is output. I wrote loggo as a Go port of a C++ library I wrote for nux/unity (my third or fourth logging library). Loggers are very lightweight objects that have a single pointer to the module. A module is defined by a dotted name, like "juju.uniter" or "juju.provider.ec2". There is no limit to the depth of nesting. Logging levels are inherited unless explicitly set. The root logger, identified by the module "" (empty string), defaults to WARNING. The existing log package has been updated to use the loggo methods so we don't have to change all the call sites immediately. However we should start to move to more detailed modules for logging as we can have better control over it. We should, as a soon to be done step, provide a way, probably using set-env, to configure logging for various machine agents or unit agents. This would allow a high level of control over what information gets logged. R=mue, jameinel CC= https://codereview.appspot.com/9738051	Tim Penhey	10 years ago
1249	Fix trunk so it builds and runs tests Fix trunk so it builds and runs tests r1247 introduced a regression to MachineSuite.TestServeAPI in cmd/jujud/machine_test.go It was not obvious to me in five minutes how to make the test still test what it should, so I have disabled the test by commenting it out, and left a TODO with the original branch author to either fix the test or delete it. R=wallyworld CC= https://codereview.appspot.com/10022043	Tim Penhey	10 years ago
1248	Make the API address easily accessible by charms Make the API address easily accessible by charms The API server addresses are now available to charms in an environment variable, included in the hooks context. This allows charms (e.g. Juju GUI) to easily retrieve the API address to use when connecting to the juju WebSocket server. Also added distro-info-data to the list of apt dependencies required for testing. Pre implementation discussion with Roger. R=fwereade, dimitern CC= https://codereview.appspot.com/9641044	Francesco Banconi	10 years ago
1247	state/api: Machiner API facade state/api: Machiner API facade Implemented Machiner facade providing API methods with bulk operations support, but only the subset of state calls needed by the machiner worker. Removed domain objects like Machine and Unit both server-side and client-side. Instead, the machiner facade at server-side provides all the API calls needed by the machiner. In a follow-up at the client-side there will be another Machiner facade and lightweight MachinerMachine wrapper preserving the state.Machine interface, while using the facade to do the actual calls. Thus the machiner facade effectively implements the state.State object, but for the machiner. Also removed the Pinger at both sides and some obsolete state/api/params types. Added internal server-side tests for the machiner. R=fwereade, mue CC= https://codereview.appspot.com/9896046	Dimiter Naydenov	10 years ago
1246	various: improve 'juju status' time for large envs various: improve 'juju status' time for large envs The net improvement is to change the time of 'juju status' in my extra-large environment from ~120s down to about 60s. The presence Watcher has a sync() method to grab all the currently alive beings. The first time it does sync, it hasn't seen any beings, so it was doing a one-by-one request for the being's sequence number and string key. This code now just does a batch request for all beings, and uses that to populate its map. There is a small concern that if there is a lot of churn, the total number of beings that have ever existed will get significantly larger than the set of currently active beings. However, fetching a batch of 750 takes 200ms, while fetching them one by one takes 50s. So we have to have 200x more dead items than currently living items before we break even. R=fwereade, rog, Danilo CC= https://codereview.appspot.com/9302046	John A Meinel	10 years ago
1245	Make simplestreams metadata loading more robust Make simplestreams metadata loading more robust When looking at each base URL to try and load the simplestreams metadata, unauthorised errors would be considered fatal and abort the search. This branch makes such errors similar to not found errors in that the next URL is attempted. R=jameinel, dimitern CC= https://codereview.appspot.com/9666050	Ian Booth	10 years ago
1244	Consolidate common error code Consolidate common error code I need to do some work which involves using an UnauthorisedError. There's one defined, but on the state package. There's also two different NotFound error implementation - one in state and one in environs. These errors, and possibly others, should be in a common, generic package so they can be used across the code base. I've created an errors packge and moved the structs and associated help methods there eg IsNotFoundError(), NotFoundf() etc. The changes are just mechanical and eliminate existing duplicate code. But now in the next branch, I can reuse UnauthorisedError and avoid defining another instance of this error. R=jameinel, dfc CC= https://codereview.appspot.com/9859045	Ian Booth	10 years ago
1243	Implement a add-machine command Implement a add-machine command add-machine will add a new machine record to the environment which results in a new empty machine being spun up. Optional constraints can be supplied, which are merged with the existing environ constraints and assigned to the machine. The machine is created with JobHostUnits running on it, so it is ready to accept new units. R=thumper, fwereade CC= https://codereview.appspot.com/9662048	Ian Booth	10 years ago

Older »