lp:~natefinch/juju-core/052-write-majority

Branch merges

Propose for merging

No branches dependent on this one.

Related source package recipes

No recipes using this branch. (?)

Create packaging recipe

Related snap packages

Related bugs

Related blueprints

2711. By Nate Finch on 2014-05-09

switch to write majority

2710. By John A Meinel on 2014-05-08

[r=jameinel] bzr: fix a test about paths

When bzr looks up the root of a branch, it expands symlinks to get the
'canonical' path. On Mac OS X, /tmp and /var are actually symlinks to
/private/tmp and /private/var which broke the tests.

So just EvalSymlinks on the path and it works again.

https://codereview.appspot.com/98040043/

2709. By John A Meinel on 2014-05-08

[r=jameinel] version/version.go: refactor to allow series on Mac

This pulls out a bunch of code for determining the series of the current host and makes it possible to implement platform specific calls.

It then exposes Mac platform as "darwin13" which matches the $OSTYPE environment variable. Most of the code is written so that it can be tested on all platforms, and only the last little connection is done as platform specific code.

I don't know if we plan on doing OS Version specific binaries for Mac/Win like we do for Ubuntu (trusty vs precise), but I figured I could start with the specific version, and we can decide if we want to do version agnostic later. ('darwin' vs 'darwin13')

I know we intend that the 'juju' client is supported on Mac, and this is a step towards getting the test suite to pass there. Right now the first test that wants 'tools' fails because the tools version is "unknown", even though all of the code builds successfully.

2708. By Andrew Wilkins on 2014-05-08

[r=axwalk] environs/manual: use old AddMAchines API

Manual provisioning was implicitly changed
to use the new AddMachinesV2 client API,
with no backwards compatibility failover.
I've changed it to explicitly use the old
API, since we do not require anything of
the new one.

Fixes lp:1317267

https://codereview.appspot.com/95150043/

2707. By Ian Booth on 2014-05-08

[r=jameinel] Merge 1.18 into trunk.

There were a bunch of conflicts from things that had to be backported but in a slightly different way. I verified that all the conflicts could be resolved in favor of current code.

There was one patch that wasn't in trunk at all, but according to the bug "didn't apply to trunk".
However, I think at least one of the assertions (bootstrap machine has its addresses set) does still apply, so I kept that bit.

2706. By Menno Finlay-Smits on 2014-05-08

[r=menno.smits],[bug=1302005] cmd/juju: deprecate --series, added --upload-series

Emit a deprecation warning to stderr when --series is used

To make testing easier an interface is now used to allow the calls
made by BootstrapCommand.Run to environs/bootstrap to be mocked out.

https://codereview.appspot.com/94060044/

R=thumper

2705. By Dimiter Naydenov on 2014-05-07

[r=dimitern],[bug=1316762] maas: Fix bug #1316762 - modprobe 8021q if needed

Conditionally modprobe 8021q driver used for VLAN
support, only if it's not already loaded. This is
needed to enable the cloud-installer to bootstrap
inside an lxc container, or deploy an lxc container
at all for that matter (it was broken).

Live tested on MAAS with precise and trusty.

https://codereview.appspot.com/93130043/

R=natefinch

2704. By Andrew Wilkins on 2014-05-07

[r=axwalk] Improve mgo testing

- Use testing.MgoSuite in the store package
- Wait for "waiting for connections" message in mongod
  output before starting tests.
- Fix testing.MgoInstance.DialInfo to configure
  the client based on whether TLS is being used.

https://codereview.appspot.com/97110043/

2703. By John A Meinel on 2014-05-07

[r=jameinel] refactor to remove apiRootForEntity

This was going to be used for giving different roots for differnt enitities, but we don't actually do that, and it made for some very awkward code.

https://codereview.appspot.com/92920043/

2702. By Andrew Wilkins on 2014-05-07

[r=axwalk] state: reduce default dial timeout to 30s

Previously we could have been dialling state connections
across the Internet, since the CLI would connect to Mongo
directly. Now, with the CLI going to API only, we can
safely reduce the timeout.

It is possible that the timeout will be exceeded if the
peers in an HA setup are not all ready. In this case the
dial operation will fail and the caller (state server agent)
will restart.

This change will reduce the time to failure in tests, and
allow us to capture some information about the failed test
without exceeding the Go test timeout.

https://codereview.appspot.com/97100045/