/juju

Merge master #2520

Merged
merged 557 commits into from Jun 8, 2015

Conversation

Reviewers
No reviews
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
@voidspace

voidspace commented Jun 8, 2015

Merge master back into the devices-api-maas feature branch.

(Review request: http://reviews.vapour.ws/r/1886/)

jujubot and others added some commits May 21, 2015

@jujubot
Merge pull request #2395 from dimitern/lp-1456989-1.24 
Fixed lp:1456989 - cloud-init 0.6.3 (precise) workaround

Reintroducing a change which was dropped with the introduction of CentOS
support (#2066): cloud-init 0.6.3 needs special handling for packages
needing the cloud-tools repo (precise-updates/cloud-tools). Without this
fix, deploying to precise fails at cloud-init time with errors running
apt-get install. For more details see http://pad.lv/1456989.

Live tested on MAAS and EC2.

(Review request: http://reviews.vapour.ws/r/1756/)
3911b51
@ericsnowcurrently
Revert "Merge pull request #2389 from ericsnowcurrently/fix-1457011-b… 
…ashisms"

This reverts commit 3159890, reversing
changes made to cc30b73.
4be2bc8
@jujubot
Merge pull request #2396 from ericsnowcurrently/revert-1.24 
Revert "Merge pull request #2389 ..."

(fixes https://bugs.launchpad.net/juju-core/+bug/1457031)

This reverts commit 3159890, reversing
changes made to cc30b73.

(Review request: http://reviews.vapour.ws/r/1757/)
5bec335
@anastasiamac
Storage add hook tool.
7794249
@anastasiamac
Reverting to implementation where ctx.unit is used.
db81b5a
@wallyworld
Don't complain when destroying a machine that is about to be cleaned up
463a74e
@wallyworld
Remove debug
7033987
@mjs
state: exposed txn pruning functionality 
github.com/juju/txn.Runner now has a MaybePruneTransactions
method. This change exposes that on multiEnvRunner and as a method on
State.
411e7a5
@mjs
Bumped the juju/txn dependency to get new txn pruning functionality 
Runner.MaybePruneTransactions
c1715d6
@mjs
worker/txnpruner: new simple worker to prune completed txn data 
Periodically calls PruneTransactions to keep the txns collection in
check.
c17bc12
@mjs
cmd/jujud: run the txnpruner worker in the master machine agent
0f9fdb4
@axw
provider/ec2: tag instances and volumes 
We tag instances started with StartInstances
with the machine tag and the environment UUID.
Similarly, we tag volumes with volume tag and
environment UUID.

State server instances are additionally tagged
with a "JujuStateServer" tag. This is not used
for anything yet, but we can later use it to
improve the way we identify state servers.

In the console, instances and volumes will now
display the tag as the instance's "name" column.
The environment UUID can be used for identifying
instances and volumes within an environment, e.g.
for AllInstances, and also for billing purposes.
6238cd3
@jujubot
Merge pull request #2386 from mjs/1453785-txn-pruning-1.24 
Prune completed transactions

This set of changes integrates the new txn pruning functionality in juju/txn into Juju. A new txnpruner worker for the master machine agent is introduced to call MaybePruneTransactions periodically. Transactions will be pruned if the transaction count has doubled since the last prune.

Fixes LP #1453785.

(Review request: http://reviews.vapour.ws/r/1747/)
fbae2a7
@jujubot
Merge pull request #2405 from axw/provider-ec2-tag-resources 
provider/ec2: tag instances and volumes

We tag instances started with StartInstances
with the machine tag and the environment UUID.
Similarly, we tag volumes with volume tag and
environment UUID.

State server instances are additionally tagged
with a "JujuStateServer" tag. This is not used
for anything yet, but we can later use it to
improve the way we identify state servers.

In the console, instances and volumes will now
display the tag as the instance's "name" column.
The environment UUID can be used for identifying
instances and volumes within an environment, e.g.
for AllInstances, and also for billing purposes.

(Review request: http://reviews.vapour.ws/r/1767/)
2e5fdb7
@anastasiamac
Added creation of Unit for tests.
cf927ec
@anastasiamac
Fixed comment and permission check.
3261a58
@anastasiamac
Created test suite where storage possible.
6606095
@anastasiamac
Added test for non-storage-capable charms.
2166dba
@anastasiamac
Comment.
9e04536
@anastasiamac
Check input arguments.
b7f1040
Fix bug 1454697 
Agents which use a close worker need to close the API they open
if their start functions fail.  This patch fixes the machine
and unit agents and mocks out the api state upgrader to simulate
the failure and test that the close occurs.

Conflicts:
	cmd/jujud/unit.go
23e5718
Address review comments. 
Used a channel instead of a boolean for the patched reportClosedAPI
function.  Also changed the report* functions to take an io.Closer
rather than an interface{}.

Conflicts:
	cmd/jujud/unit_test.go
95b2c0f
Address review comments 
Added closure of the channel.
ad1089b
@wallyworld
Change logic so that a machine with a dying principal unit can still … 
…become dying
b0f0f3f
@jujubot
Merge pull request #2410 from cherylj/file_handle_124 
Forward port of fix for bug 1454697 (1.24)



(Review request: http://reviews.vapour.ws/r/1771/)
49d506b
@anastasiamac
Made comment readable.
0f4272d
@anastasiamac
Added check that mock is called.
3ad6afe
@wallyworld
Allow machine destroy if all units dying, not just 1
24a940d
@wallyworld
Check unit names match when validating machine destroy
c3adacf
@wallyworld
Small fix
ec94fd9
@anastasiamac
Changed to add storage on succeful hook.
f039f17
@wallyworld
Fix command doc
599f01f
@jujubot
Merge pull request #2404 from wallyworld/service-amchine-destroy 
Don't complain when destroying a machine that is about to be cleaned up

Fixes: https://bugs.launchpad.net/juju-core/+bug/1455158

Adds extra doc to service and unit destroy commands to explain that when the last unit is destroyed, the machine can also be cleaned up.

Add code to machine destroy to check that if there is a unit still on that machine, the service itself may be dying and thus the cleanup worker will eventually harvest the machine. So the machine destroy becomes a no-op and there's no error returned. Previously an error was returned which complained that the machine had assigned units, which it sort of did, but not alive for much longer and the user would get an error and see the machine destroyed anyway.

(Review request: http://reviews.vapour.ws/r/1766/)
bcacd5e
Resumer API facade implemented - server & client
0bda751
@anastasiamac
Fixing full-stack tests
6159189
Resumer worker now uses the API instead of State
1e97c09
@wallyworld
Fix lease txn structure; do not ignore errors; worker exists on error
5db5f30
Refactored tests not to use JujuConnSuite
9d638d0
Added missing files from last commit
1cf3a14
Improved amachine agent tests for resumer
c7add3c
Improved resumer worker tests - not using JujuConnSuite
c133f94
@anastasiamac
Added check for storage instance count 0.
e3c0aa4
@anastasiamac
Added check for storage instance count 0.
0083d55
@anastasiamac
Added check for storage instance count 0.
a2f3c29
@anastasiamac
Added constraints accumulation and renamed vars.
71722cd
@anastasiamac
Added writeChanges check :D
59cbb99
@anastasiamac
Corrected test for accumulation.
91a5051
@anastasiamac
Added test for behaviour during env blocks.
5a6d9b1
@anastasiamac
Fixed tests
bca0e54
@sinzui
Increment juju to 1.24-beta6
4f9ff0b
@wallyworld
Use a generic error to pass back to callers if lease calls fail
c87fd18
@waigani
Merge branch 'master' of https://github.com/juju/juju into upgrade-ho… 
…sted-env-master
39876f7
@anastasiamac
Cleaned up based on review. Added bug numbers.
df29667
@axw
provider/openstack: tag resources 
We tag instances and volumes with the environment UUID.

State server instances are additionally tagged with a
"JujuStateServer" tag. This is not used for anything yet,
but we can later use it to improve the way we identify
state servers.
1e17947
@waigani
Merge branch 'master' of https://github.com/juju/juju into storage-pr… 
…ovisioner-bug
81980dc
@axw
jujuc: send stdin to server 
Send stdin to the uniter's RPC server when invoking
jujuc commands. This enables the user to pipe data
to jujuc commands.

Fixes https://bugs.launchpad.net/juju-core/+bug/1454678
f35a8c6
@jujubot
Merge pull request #2407 from axw/provider-openstack-tag-resources 
provider/openstack: tag resources

We tag instances and volumes with the environment UUID.

State server instances are additionally tagged with a
"JujuStateServer" tag. This is not used for anything yet,
but we can later use it to improve the way we identify
state servers.

Depends on go-goose/goose#9

(Review request: http://reviews.vapour.ws/r/1768/)
f1100da
@jujubot
Merge pull request #2400 from anastasiamac/storage-add-hook-tool 
Storage add hook tool.



(Review request: http://reviews.vapour.ws/r/1761/)
64d6c3f
@natefinch
move syslog config to /var/lib/juju/
ce898a6
@natefinch
upgrade steps
536d006

jujubot and others added some commits Jun 3, 2015

@jujubot
Merge pull request #2476 from natefinch/upgrade-fix 
make the move-syslog-config upgrade more lenient

…about not being able to delete old config files.  We don't actually care overly much if the old ones are left around, so now we just log if we can't delete them, rather than screwing the upgrade because of it.
f479fac
@davecheney
Fix LP 1460893 
Fix various errors spotted with Gordon Klaus' `ineffassign`.

Most are harmless, but some are symptoms of dead code, or unchecked conditions.
a1c2723
@davecheney
fix test
8ec6a53
@axw
Add resource tags to resource provisioning info 
We now pass tags to set on resources (initially
instances, volumes and filesystems) through their
provisioning parameters, via the API.

We introduce a new environs/tags package that holds
the names we use for common tags, as well as helpers
for creating tag sets.
5cb1112
@jujubot
Merge pull request #2379 from perrito666/uas_status_set_service 
added satus-set and status-get for service.

Still needs more tests (Writing them but I would appreciate reviews in the mean time) and a fake leader check.

(Review request: http://reviews.vapour.ws/r/1739/)
01fb9f6
@jujubot
Merge pull request #2479 from axw/resource-tags-units-storagename-etc 
Add resource tags to resource provisioning info

We now pass tags to set on resources (initially
instances, volumes and filesystems) through their
provisioning parameters, via the API. Tags make
their way into StartInstanceParams.InstanceConfig,
and into VolumeParams and FilesystemParams.

We introduce a new environs/tags package that holds
the names we use for common tags, as well as helpers
for creating tag sets.

The ec2 and openstack providers are updated to use
the tags when creating instances and volumes. Later
we should introduce a worker to keep tags up-to-date,
but for now we only set tags at instance creation
time.

(Review request: http://reviews.vapour.ws/r/1845/)
a7dfe0e
@anastasiamac
Added check to api level, corrected doc comment.
4e13ac2
@perrito666
Added better error management.
cd6ec66
@wwitzel3
removing manual validation logging from state
cd398c9
@wwitzel3
removing manual validation logging from state
f7f6e64
@jujubot
Merge pull request #2486 from wwitzel3/juju-1.24 
removing manual validation logging from state

Remove some logging that was left behind in state.

FIxes LP: https://bugs.launchpad.net/juju-core/+bug/1453853

(Review request: http://reviews.vapour.ws/r/1852/)
1e8a11b
@natefinch
fix bug in namespace usage
b4a9c55
@natefinch
add unit test
bf5a8ad
@jujubot
Merge pull request #2488 from wwitzel3/master 
removing manual validation logging from state

Patching from juju-1.24 fix.
Previous review: http://reviews.vapour.ws/r/1852/

(Review request: http://reviews.vapour.ws/r/1854/)
0c7b39f
@jujubot
Merge pull request #2489 from natefinch/fix-1461354-tryagain 
fix bug in namespace usage (lp:1461354)

fixes-1461354

https://bugs.launchpad.net/juju-core/+bug/1461354
debug-log EOF: invalid tag and panic seen in state server logs

This was a result of me changing the namespace stored in the config to no longer store the leading "-". Instead, it just stores the namespace itself, and where we use it, we have to format it correctly. This seems like a much more sane way to do it, so the value you set is the value that exists, and we don't dictate formatting to everyone using the namespace.... however, it meant I forgot to check for an empty namespace in this one case.

(Review request: http://reviews.vapour.ws/r/1855/)
7e97a2b
@anastasiamac
Only count matters.
f3a53ea
@jujubot
Merge pull request #2484 from perrito666/fix_1.24_1431372 
fix 1431372 restore fails

Added better error management.
This makes juju restore plugin give more useful output when non vital tasks fail.

(Review request: http://reviews.vapour.ws/r/1850/)
8dae5dc
@davecheney
fix test
71708f7
@anastasiamac
Check only count is provided in apiserver and copy other constraints … 
…from what unit has.
99d47af
@jujubot
Merge pull request #2482 from davecheney/100-fix-lp-1460893 
Fix LP 1460893

Fixes LP 1460893

Fix various errors spotted with Gordon Klaus' `ineffassign`.

Most are harmless, but some are symptoms of dead code, or unchecked conditions.

(Review request: http://reviews.vapour.ws/r/1848/)
392a9b4
@anastasiamac
Check count valid.
7d720ca
@anastasiamac
Corrected tests.
7ecae49
@jujubot
Merge pull request #2478 from anastasiamac/no-pool-size-bug1461342 
storage-add is provided only count.

 Units should only be able to request more of storage that has already been assigned, i.e. they should be able to specify count.


(Review request: http://reviews.vapour.ws/r/1844/)
e4d2f6f
@jujubot
Merge pull request #2467 from cherylj/vivid_containers_124v2 
Move cloud-init logic out of service/systemd

This PR addresses additional review comments for the original fix for:
https://bugs.launchpad.net/juju-core/+bug/1442308

The behavior is the same, but moves around some of the logic to match the changes made in master.  The commits for master were cherry picked to 1.24 and manually tested.

Review on master:  http://reviews.vapour.ws/r/1797/

(Review request: http://reviews.vapour.ws/r/1831/)
f5b84ae
@rogpeppe
cmd/juju/agent: stop watcher in waitForOtherStateServers
86d241c
@jujubot
Merge pull request #2492 from rogpeppe/029-upgradeWorkerContext-watch… 
…er-stop

cmd/juju/agent: stop watcher in waitForOtherStateServers

Without this change, UpgradeSuite fails more often than not (it panics). With it,
it passes consistently for me.


(Review request: http://reviews.vapour.ws/r/1859/)
8faff02
@rogpeppe
worker/envworkermanager: always stop worker on exit
919bf43
@jujubot
Merge pull request #2494 from bz2/lp_bug_link_fix 
Fix two lp bug links from github transition

Rog pointed out the move to github and auto-rename of all
the imports also accidentally broke a couple of bug links
in his comments. Switch them back to pointing at launchpad.

(Review request: http://reviews.vapour.ws/r/1861/)
0ca2d0f
@sinzui
Increment juju to 1.24-beta7
ddc63f4
@jujubot
Merge pull request #2495 from sinzui/inc-1.24-beta7 
Increment juju to 1.24-beta7



(Review request: http://reviews.vapour.ws/r/1862/)
53bad81
@wallyworld
Port a number of 1.24 fixes to master
92f79ce
@wallyworld
Add missing dep
6c9909c
@wallyworld
Fix test
0eac46c
@jujubot
Merge pull request #2501 from wallyworld/merge-1.24v2 
Merge 1.24v2

Fixes:
https://bugs.launchpad.net/bugs/1457728
https://bugs.launchpad.net/bugs/1461150
https://bugs.launchpad.net/bugs/1431372
https://bugs.launchpad.net/bugs/1459057
https://bugs.launchpad.net/bugs/1459060
https://bugs.launchpad.net/bugs/1459611
https://bugs.launchpad.net/bugs/1459616
https://bugs.launchpad.net/bugs/1459885
https://bugs.launchpad.net/bugs/1459912
https://bugs.launchpad.net/bugs/1460171
https://bugs.launchpad.net/bugs/1461111
https://bugs.launchpad.net/bugs/1461342
https://bugs.launchpad.net/bugs/1454678

As well as adding support for tags on machines and volumes.

(Review request: http://reviews.vapour.ws/r/1868/)
e7fcf10
@gabriel-samfira
Replace windows services package 
This PR branch replaces the current windows service package, which is highly dependent on executing powershell commandlets with a native implementation.

For the time being, the securestring package will be added to the windows service package. It has to be made more generally useful before we can split it out into its own project. It still lacks entropy, encrypting binary blobs and using encryption stores other then the user one.
1644855
@gabriel-samfira
respond to feedback
8a209ea
@gabriel-samfira
respond to feedback
993568f
@gabriel-samfira
respond to feedback
037bcde
@gabriel-samfira
respond to feedback
669d31d
@gabriel-samfira
respond to feedback
5af6ae1
@gabriel-samfira
respond to feedback
883a81b
@gabriel-samfira
respond to feedback
d7b02df
@jujubot
Merge pull request #2491 from rogpeppe/028-envworkermanager-kill-worker 
worker/envworkermanager: always stop worker on exit

This fixes an intermittent test failure in TestLoopExitKillsRunner
which happened because the underlying worker was not killed
when the loop terminated with an error.

We also change the API into a slightly more general form - there's
no need for the EnvWorkerManager to know that the worker
it's dealing with is actually a worker.Runner, and this makes
the tests slightly simpler to write (and arguably the API easier
to understand)


(Review request: http://reviews.vapour.ws/r/1858/)
884972e
@gabriel-samfira
respond to feedback
ae4ba71
@jujubot
Merge pull request #2248 from gabriel-samfira/services 
Replace windows services package

This PR branch replaces the current windows service package, which is highly dependent on executing powershell commandlets, with a native implementation.

For the time being, the securestring package will be added to the windows service package. It has to be made more generally useful before we can split it out into its own project. It still lacks entropy, encrypting binary blobs and using encryption stores other then the user one. Fot the purpose of this package, the functionality it offers is enough.

This PR also removes the following dependencies:

bitbucket.org/kardianos/osext
bitbucket.org/kardianos/service

by implementing a small service handler using winsvc package.

Apologies for the long PR.

(Review request: http://reviews.vapour.ws/r/1609/)
bccc999
@waigani
Fix: 1441478 availability zone upgrade fails if containers are present 
Don't add availability zones to containers, don't error on missing
instances.
63e0610
@waigani
FIX 1403165 CMD: destroy server without jenv bootstrap cfg 
When destroying the state server from a jenv that does not have
bootstrap info, get that info from the API server instead and use it
to destroy the environment.
bbfddb7
@jujubot
Merge pull request #2507 from waigani/fix-1441478-avalzones-125 
Fix: 1441478 availability zone upgrade fails if containers are present

Don't add availability zones to containers, don't error on missing
instances.

(Review request: http://reviews.vapour.ws/r/1873/)
964819a
@jujubot
Merge pull request #2508 from waigani/fix-1403165-destroy-without-boo… 
…tstrap-125

FIX 1403165 CMD: destroy server without jenv bootstrap cfg

When destroying the state server from a jenv that does not have
bootstrap info, get that info from the API server instead and use it
to destroy the environment.

(Review request: http://reviews.vapour.ws/r/1874/)
293eafa
@mjs
worker/txnpruner: allow for the worker to run exactly on the interval 
This happens on Windows occasionally.

Fixes LP #1459291
b83d1ae
@jujubot
Merge pull request #2474 from waigani/deps-whitespace 
Check for white space in dependencies.tsv

If a revision id begins or ends with a space,
TestDependenciesTsvFormat will pass, but godeps will fail. Update the
test to check that the prefix and suffix are not spaces.

(Review request: http://reviews.vapour.ws/r/1840/)
dd9b590
@jujubot
Merge pull request #2516 from mjs/1459291-fix-testprunes 
worker/txnpruner: allow for the worker to run exactly on the interval

This happens on Windows occasionally.

Fixes LP #1459291

(Review request: http://reviews.vapour.ws/r/1882/)
699d183
Fixed lp:1462412: worker/resumer flaky test
c8fb832
Changes after review, some tests improvements
c5733d4
@jujubot
Merge pull request #2518 from dimitern/lp-1462412-worker-resumer-flak… 
…y-test

Fixed lp:1462412: worker/resumer flaky test

Tested with:

`time go test -c && for i in `seq 50`; do ./resumer.test & sleep 0.01;
done`

Before the fix I got 3-4 failures, now none are reproducible.

(Review request: http://reviews.vapour.ws/r/1884/)
28ead85
@voidspace

voidspace commented Jun 8, 2015

$$merge$$
@jujubot
Contributor

jujubot commented Jun 8, 2015

Status: merge request accepted. Url: http://juju-ci.vapour.ws:8080/job/github-merge-juju

jujubot added a commit that referenced this pull request Jun 8, 2015

@jujubot
Merge pull request #2520 from voidspace/devices-master-merge-1 
Merge master

Merge master back into the devices-api-maas feature branch.

(Review request: http://reviews.vapour.ws/r/1886/)
0608788

@jujubot jujubot merged commit 0608788 into juju:devices-api-maas Jun 8, 2015

@voidspace voidspace deleted the voidspace:devices-master-merge-1 branch Oct 6, 2016

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment