]> review.fuel-infra Code Review - openstack-build/cinder-build.git/log
openstack-build/cinder-build.git
11 years agoMerge "Integrate oslo's periodic tasks."
Jenkins [Tue, 11 Jun 2013 18:17:52 +0000 (18:17 +0000)]
Merge "Integrate oslo's periodic tasks."

11 years agoMerge "Replace FLAGS with cfg.CONF in api"
Jenkins [Tue, 11 Jun 2013 07:08:58 +0000 (07:08 +0000)]
Merge "Replace FLAGS with cfg.CONF in api"

11 years agoIntegrate oslo's periodic tasks.
Svetlana Shturm [Mon, 10 Jun 2013 14:02:26 +0000 (15:02 +0100)]
Integrate oslo's periodic tasks.

Cinder use old style invocation of periodic tasks, which not based on oslo library.
It will be better to use main idea of periodic tasks for all services.

Blueprint: oslo-periodic-tasks

Change-Id: I6ac1ca28abefcc9d0bbfd41873c8f57d40b4a97a

11 years agoMerge "NFS drivers don't honor vm size with volume from an image"
Jenkins [Tue, 11 Jun 2013 05:04:29 +0000 (05:04 +0000)]
Merge "NFS drivers don't honor vm size with volume from an image"

11 years agoReplace FLAGS with cfg.CONF in api
Sergey Vilgelm [Mon, 10 Jun 2013 08:52:13 +0000 (12:52 +0400)]
Replace FLAGS with cfg.CONF in api

Replace all the FLAGS with cfg.CONF in cinder/api
Large commit was split into several parts

Change-Id: I2114d5fc45f5299c3b2011270034a3370e0ec388
Fixes: bug #1182037
11 years agoAdd the iscsi device check and exception processing.
Vincent Hou [Thu, 6 Jun 2013 08:46:52 +0000 (16:46 +0800)]
Add the iscsi device check and exception processing.

Before downloading the image and executing the command "qemu-img convert",
check if the iSCSI device is still available via the command "sudo dd
if=<device> of=/dev/null count=1". This command will raise an exception with
the message "Input/output error", if the back-end storage is disconnected to
the cinder-volume node, so we use it to test the availability of the storage
device. If it is unavailable, there is no need to download the image &
"qemu-img convert" and an exception DeviceUnavailable will be raised.

Fixed Bug1169290.

Change-Id: I133b4cc1bac493df073d42e240092cf2e6300454

11 years agoMerge "Imported Translations from Transifex"
OpenStack Jenkins [Sat, 8 Jun 2013 16:07:11 +0000 (16:07 +0000)]
Merge "Imported Translations from Transifex"

11 years agoMerge "Minor Logic bug in NFS Driver."
Jenkins [Sat, 8 Jun 2013 15:45:12 +0000 (15:45 +0000)]
Merge "Minor Logic bug in NFS Driver."

11 years agoMerge "Fix 'undefined symbol conn' error"
Jenkins [Sat, 8 Jun 2013 15:40:43 +0000 (15:40 +0000)]
Merge "Fix 'undefined symbol conn' error"

11 years agoMinor Logic bug in NFS Driver.
Morgan Fainberg [Sat, 8 Jun 2013 00:18:59 +0000 (17:18 -0700)]
Minor Logic bug in NFS Driver.

When checking oversubscription rate, we need to ensure that we
check the _allocated_ (apparant size) not the _available_ (actual
usage) on the NFS share.  This fix ensures we are now checking the
correct values.  This bug allowed (effectively) a one-time
volume allocation that would push the "apparant" size over the
oversubscription rate.

Change-Id: I189b689cb76ac2a243aa6e255ce0c80d86e247ce
Fixes: bug #1188877
11 years agoImported Translations from Transifex
OpenStack Jenkins [Fri, 7 Jun 2013 18:10:20 +0000 (18:10 +0000)]
Imported Translations from Transifex

Change-Id: I888eb000062b469a4b55274dfec2cfd28e9c329c

11 years agoFix 'undefined symbol conn' error
Dirk Mueller [Sat, 1 Jun 2013 12:48:03 +0000 (14:48 +0200)]
Fix 'undefined symbol conn' error

Use self.conn, not conn.

Fixes LP Bug #1188619

Change-Id: Idbd62016fda4cafb6df953ca57af94d56edd3918

11 years agoNFS drivers don't honor vm size with volume from an image
leseb [Thu, 23 May 2013 17:09:06 +0000 (19:09 +0200)]
NFS drivers don't honor vm size with volume from an image

The bug has been encountered with the NFS generic driver
(cinder.volume.nfs.NfsDriver) and the NetApp NFS driver
(cinder.volume.drivers.netapp.nfs.NetAppDirectCmodeNfsDriver). I believe
that drivers based on distributed filesystem such as GlusterFS, nexenta
and scality are also impacted however I didn't test it those backends.
However since most of them already inherit from the RemoteFsDriver
class, this should be fine.

Change-Id: I14575da69a2c99c7cbcece27b40a171153371ee3
Fixes: bug #1183459
11 years agoAdd missing tests for backup_* methods
Yuriy Zveryanskyy [Tue, 4 Jun 2013 10:14:12 +0000 (13:14 +0300)]
Add missing tests for backup_* methods

There was no tests in test_db_api for backup_*
methods.

Add tests to ensure that all works.

blueprint db-api-tests

Change-Id: Ie83de7f9cb3d2c4de5442513a60bfedac312f62e

11 years agoMerge "Unset all stubs before running other cleanups."
Jenkins [Fri, 7 Jun 2013 04:11:53 +0000 (04:11 +0000)]
Merge "Unset all stubs before running other cleanups."

11 years agoMerge "Add missing tests for iscsi_* methods"
Jenkins [Fri, 7 Jun 2013 02:37:41 +0000 (02:37 +0000)]
Merge "Add missing tests for iscsi_* methods"

11 years agoRemove E12 errors from tox.ini Flake ignores.
John Griffith [Sat, 1 Jun 2013 17:27:04 +0000 (11:27 -0600)]
Remove E12 errors from tox.ini Flake ignores.

This removes the E12 from ignores and fixes up the
existing pep8 errors that we were ignoring.

Change-Id: I5d60f1eed768fcae01a708fcf9ea324844c6376d

11 years agoMerge "Fix config registration in cinder volume drivers."
Jenkins [Thu, 6 Jun 2013 07:28:05 +0000 (07:28 +0000)]
Merge "Fix config registration in cinder volume drivers."

11 years agoMerge "Add call to vol driver when accepting a transfer"
Jenkins [Wed, 5 Jun 2013 22:47:36 +0000 (22:47 +0000)]
Merge "Add call to vol driver when accepting a transfer"

11 years agoUnset all stubs before running other cleanups.
Yuriy Taraday [Tue, 28 May 2013 12:44:51 +0000 (16:44 +0400)]
Unset all stubs before running other cleanups.

Cleanup callbacks are run in reverse order so all stubs unsetting
should be added at the end of the cleanup list.

Fixes bug 1185004

Change-Id: I5ad3ea2cdce663aaeae6b7cc2cb9c5dac9d12435

11 years agoMerge "Elevate acceptors context on accept reserve udpate"
Jenkins [Wed, 5 Jun 2013 08:12:09 +0000 (08:12 +0000)]
Merge "Elevate acceptors context on accept reserve udpate"

11 years agoMerge "Re-write of the cinder-manage man page."
Jenkins [Wed, 5 Jun 2013 07:24:32 +0000 (07:24 +0000)]
Merge "Re-write of the cinder-manage man page."

11 years agoMerge "Handle IPv6 specifid glance servers gracefully"
Jenkins [Wed, 5 Jun 2013 07:24:29 +0000 (07:24 +0000)]
Merge "Handle IPv6 specifid glance servers gracefully"

11 years agoMerge "Log iSCSI target output on error."
Jenkins [Wed, 5 Jun 2013 07:24:25 +0000 (07:24 +0000)]
Merge "Log iSCSI target output on error."

11 years agoMerge "Imported Translations from Transifex"
OpenStack Jenkins [Wed, 5 Jun 2013 07:24:21 +0000 (07:24 +0000)]
Merge "Imported Translations from Transifex"

11 years agoMerge "Replace FLAGS with cfg.CONF in db"
Jenkins [Wed, 5 Jun 2013 04:23:51 +0000 (04:23 +0000)]
Merge "Replace FLAGS with cfg.CONF in db"

11 years agoFix config registration in cinder volume drivers.
John Griffith [Thu, 30 May 2013 22:05:45 +0000 (16:05 -0600)]
Fix config registration in cinder volume drivers.

The config documentation relies on options being registered
on a modules import.  Our need to move the drivers to using
self.configuration for multi-backend support means that options
wouldn't be loaded until object initialization which breaks
documentation.

This patch puts a dummy CONF init/load back in the drivers. While putting
this change together I came across a number of drivers still using FLAGS,
and even worse a number of drivers using a mixture of FLAGS and CONF and
self.configuraiton.  So most of those are cleaned up here as well.

Note there are two drivers that were not updated at all here:
  1. windows.py
  2. zadara.py

The zadara folks have indicated that they're in the process of updating and
releasing a new version of their driver so I left that as is.

The windows driver needs a bit of work to switch over.

Fixes bug: 1179159

Change-Id: I90165299bf080da17741d027e36e361540da0ff8

11 years agoMerge "Fixes the 3PAR drivers CPG validation"
Jenkins [Wed, 5 Jun 2013 02:12:41 +0000 (02:12 +0000)]
Merge "Fixes the 3PAR drivers CPG validation"

11 years agoElevate acceptors context on accept reserve udpate
John Griffith [Tue, 4 Jun 2013 18:59:20 +0000 (12:59 -0600)]
Elevate acceptors context on accept reserve udpate

When accepting a transfer we need to update the
quota data for both the donor and the acceptor's.
The acceptor is easy enough (as that's the context
we currently are using) however to update the donor's
quota data we need an elevated context.

Fixes bug: 1187538

Change-Id: I1fb9e7726830dce2c3de23d3446de5bbe9a1fd24

11 years agoAdd call to vol driver when accepting a transfer
John Griffith [Tue, 4 Jun 2013 19:02:37 +0000 (13:02 -0600)]
Add call to vol driver when accepting a transfer

Some back-ends have the notion of tenancy on volumes
and set ownership in the driver.  The update of the DB
alone when doing volume transfers isn't enough, we need
to be able to propogate the update info all the way
down to the driver level.

This patch introduces the volume api/manger/rpc calls
and implements a stub in the base driver that can
be over-ridden for those that need to take some action.

Change-Id: Ica6ae368084c44b32af5d08df14bb3745f3a37ee

11 years agoImported Translations from Transifex
OpenStack Jenkins [Tue, 4 Jun 2013 18:11:20 +0000 (18:11 +0000)]
Imported Translations from Transifex

Change-Id: Id15f3465b126110983c73647530835585c31914c

11 years agoMerge "Replace FLAGS with cfg.CONF in scheduler"
Jenkins [Tue, 4 Jun 2013 17:36:11 +0000 (17:36 +0000)]
Merge "Replace FLAGS with cfg.CONF in scheduler"

11 years agoMerge "Don't set signing_dir by default."
Jenkins [Tue, 4 Jun 2013 16:33:19 +0000 (16:33 +0000)]
Merge "Don't set signing_dir by default."

11 years agoImplement DB migration for volume transfer BP
Ollie Leahy [Wed, 15 May 2013 10:05:49 +0000 (10:05 +0000)]
Implement DB migration for volume transfer BP

This patch implements the backend for volume transfers as specified in the
Volume transfer blueprint.

Change-Id: Id18e57942c3908e2539f3b0845263d0fbdd198f5
implement bp: volume-transfer

11 years agoReplace FLAGS with cfg.CONF in db
Sergey Vilgelm [Mon, 3 Jun 2013 09:12:32 +0000 (13:12 +0400)]
Replace FLAGS with cfg.CONF in db

Replace all the FLAGS with cfg.CONF in cinder/db
Large commit "https://review.openstack.org/31172" was split into several parts
Use the common pattern: CONF = cfg.CONF

Change-Id: Ibac0a4b233ba82e13e3a9bfb6bd3fd418cdab29f
Fixes: bug #1182037
11 years agoAdd missing tests for iscsi_* methods
Yuriy Zveryanskyy [Tue, 4 Jun 2013 12:56:54 +0000 (15:56 +0300)]
Add missing tests for iscsi_* methods

There was no tests in test_db_api for iscsi_*
methods.

Bug 1187367 found.

blueprint db-api-tests

Change-Id: Ia554e6030307a0c4f0510ab74fe0881ee160f743

11 years agoLog iSCSI target output on error.
Avishay Traeger [Tue, 4 Jun 2013 08:46:20 +0000 (11:46 +0300)]
Log iSCSI target output on error.

Add the command output on error to aid in debugging.

Change-Id: Ib505eaeba7f1ca0e54e99dec8ebbbf71df3e5a51
Fixes: bug 1186991
11 years agoRe-write of the cinder-manage man page.
Jay S. Bryant [Mon, 3 Jun 2013 15:01:58 +0000 (10:01 -0500)]
Re-write of the cinder-manage man page.

The cinder-manage man page didn't appear to have been re-written
since splitting from Nova.  There were numerous options that
no longer exist documented and references to Nova instead of
Cinder.

This version has been re-written based on the current code in
cinder-manage.

(fixes bug 1174502)

Change-Id: Ia03ca05bbd87444b2213bf719da0ffff96b67699

11 years agoReplace FLAGS with cfg.CONF in scheduler
Sergey Vilgelm [Mon, 3 Jun 2013 13:36:42 +0000 (17:36 +0400)]
Replace FLAGS with cfg.CONF in scheduler

Replace all the FLAGS with cfg.CONF in cinder/scheduler
Large commit "https://review.openstack.org/31172" was split into several parts
Use the common pattern: CONF = cfg.CONF

Change-Id: If293d36a9e1552f7af497edf947e82458488bf9c
Fixes: bug #1182037
11 years agoMerge "Fixed method db.api.reservation_expire"
Jenkins [Sun, 2 Jun 2013 17:10:34 +0000 (17:10 +0000)]
Merge "Fixed method db.api.reservation_expire"

11 years agoMerge "rbd: simplify configuration and use librbd and librados"
Jenkins [Sun, 2 Jun 2013 16:10:31 +0000 (16:10 +0000)]
Merge "rbd: simplify configuration and use librbd and librados"

11 years agoMerge "python3: Introduce py33 to tox.ini"
Jenkins [Sun, 2 Jun 2013 16:00:33 +0000 (16:00 +0000)]
Merge "python3: Introduce py33 to tox.ini"

11 years agoMerge "Fix AttributeError typo"
Jenkins [Sun, 2 Jun 2013 16:00:31 +0000 (16:00 +0000)]
Merge "Fix AttributeError typo"

11 years agoMerge "Add error reporting to generate_sample.sh on import failure."
Jenkins [Sun, 2 Jun 2013 16:00:28 +0000 (16:00 +0000)]
Merge "Add error reporting to generate_sample.sh on import failure."

11 years agoMerge "Update 3PAR driver session management"
Jenkins [Sun, 2 Jun 2013 10:43:58 +0000 (10:43 +0000)]
Merge "Update 3PAR driver session management"

11 years agoMerge "HDS Cinder Driver. Rev #1"
Jenkins [Sun, 2 Jun 2013 08:56:28 +0000 (08:56 +0000)]
Merge "HDS Cinder Driver. Rev #1"

11 years agopython3: Introduce py33 to tox.ini
Chuck Short [Sat, 1 Jun 2013 23:42:04 +0000 (18:42 -0500)]
python3: Introduce py33 to tox.ini

Introduce py33 to tox.ini to make testing
easier with python3.

Change-Id: I351fa475deffd75201e20e3629af6ed018b9ac7c
Signed-off-by: Chuck Short <chuck.short@canonical.com>
11 years agoFix AttributeError typo
Dirk Mueller [Sat, 1 Jun 2013 12:52:45 +0000 (14:52 +0200)]
Fix AttributeError typo

Fixes Bug LP #1186566

Change-Id: I8211d365c4d3532cd6ffc3b3041661ff6b9e637f

11 years agoFix path for pylint Gate
Dirk Mueller [Sat, 1 Jun 2013 14:20:45 +0000 (16:20 +0200)]
Fix path for pylint Gate

Change-Id: Ibd36a2a0a7ebe572a4dd660a3e35378ae02b8187

11 years agoFixed method db.api.reservation_expire
Sergey Skripnick [Wed, 29 May 2013 09:46:14 +0000 (12:46 +0300)]
Fixed method db.api.reservation_expire

There was missing relation in Reservation model.

Change-Id: I6efe558278297aea6d2348b347152c244b3323d6
Fixes: bug 1185325
11 years agoHandle IPv6 specifid glance servers gracefully
Dirk Mueller [Sat, 1 Jun 2013 12:25:37 +0000 (14:25 +0200)]
Handle IPv6 specifid glance servers gracefully

IPv6 netlocs can for valid reasons contain ':', so
splitting by ':' is dangerous. Instead of splitting
by host and port in order to only reassemble it into
a netloc again later, simply always pass netloc, which
avoids the bug.

Add extra textcases to verify the new behavior.
Fix Flake8 warnings.

Fixes LP Bug #1182830

Change-Id: I130528ae946e8888d35c25e468b4ea6ac29db0cf

11 years agoHDS Cinder Driver. Rev #1
Lakhinder Walia [Tue, 7 May 2013 23:05:46 +0000 (16:05 -0700)]
HDS Cinder Driver. Rev #1

blueprint hds-hus-iscsi-cinder-driver
This is the first rev of Hitachi Data Systems Cinder iSCSI driver.
This driver works with HUS (df850) array.
This driver contains all the base-line features specified for Havana release.
Amended into this submission are changes from code-reviews.

Docimpact: Bug #1180648

Change-Id: Ia27d076443b10da2c653456f9292dd192362b853

11 years agoMerge "Imported Translations from Transifex"
OpenStack Jenkins [Fri, 31 May 2013 22:16:42 +0000 (22:16 +0000)]
Merge "Imported Translations from Transifex"

11 years agoMerge "Migrate to Oslo DB code"
Jenkins [Fri, 31 May 2013 21:22:29 +0000 (21:22 +0000)]
Merge "Migrate to Oslo DB code"

11 years agoImported Translations from Transifex
OpenStack Jenkins [Fri, 31 May 2013 18:11:06 +0000 (18:11 +0000)]
Imported Translations from Transifex

Change-Id: Ic2b5d57ad16da67e9b152853338a2ef488eaeee7

11 years agoMerge "Prevent force delete if the volume is attached"
Jenkins [Fri, 31 May 2013 13:48:27 +0000 (13:48 +0000)]
Merge "Prevent force delete if the volume is attached"

11 years agoMerge "Updating HACKING to disallow the use of locals()"
Jenkins [Fri, 31 May 2013 13:45:54 +0000 (13:45 +0000)]
Merge "Updating HACKING to disallow the use of locals()"

11 years agoMerge "Rename requires files to standard names."
Jenkins [Fri, 31 May 2013 11:01:45 +0000 (11:01 +0000)]
Merge "Rename requires files to standard names."

11 years agoAdd error reporting to generate_sample.sh on import failure.
Anton Frolov [Fri, 31 May 2013 10:27:49 +0000 (14:27 +0400)]
Add error reporting to generate_sample.sh on import failure.

Currently tools/conf/generate_sample.sh collect options from project
modules by importing each module and looking at it's global variables
for instances of classes derived from cfg.Opt (or lists of such
instances). Sometimes this import can fail due to various reasons,
cfg.DuplicateOptError for example. And if import fails no options
from imported module written to sample config file. Error reporting
will at least warn user that some of options didn't get into sample
config.

Change-Id: I831464b1c1575c106d866a936a50ec74b2d44e59
Fixes: bug #1182611
11 years agoMerge "Fix incorrect authorization rule in quota contrib api."
Jenkins [Fri, 31 May 2013 08:32:42 +0000 (08:32 +0000)]
Merge "Fix incorrect authorization rule in quota contrib api."

11 years agoUpdating HACKING to disallow the use of locals()
Mike Perez [Fri, 31 May 2013 08:11:13 +0000 (01:11 -0700)]
Updating HACKING to disallow the use of locals()

Change-Id: I7bf2720bdb0456274dc81a73d91296dff0e3fced

11 years agoPrevent force delete if the volume is attached
Mike Perez [Fri, 31 May 2013 07:11:24 +0000 (00:11 -0700)]
Prevent force delete if the volume is attached

Force deletes were eventually failing on the volume manager layer due to
being in an attached state. This will check that up front to inform the
user that they need to detach first.

Fixes: bug #1164929
Change-Id: I24ade24fd750dc647331ef25b835f45f29c10fd7

11 years agoMerge "InvalidUUID can not be raised"
Jenkins [Fri, 31 May 2013 05:02:02 +0000 (05:02 +0000)]
Merge "InvalidUUID can not be raised"

11 years agoInvalidUUID can not be raised
Sergey Skripnick [Wed, 29 May 2013 10:44:38 +0000 (13:44 +0300)]
InvalidUUID can not be raised

There is incomplete format in exception's message.

Fixes: bug 1184870
Change-Id: I27f231c8c5cb41b1ec5fdb34fba6a8dff6a556c7

11 years agoMerge "Add testrepository to git ignores."
Jenkins [Fri, 31 May 2013 04:59:47 +0000 (04:59 +0000)]
Merge "Add testrepository to git ignores."

11 years agoFix incorrect authorization rule in quota contrib api.
Nikolay Sobolevskiy [Thu, 30 May 2013 19:21:09 +0000 (23:21 +0400)]
Fix incorrect authorization rule in quota contrib api.

Incorrect policy and authorization extension name in
 contrib quota api. Legacy from nova.

Change-Id: Ida8262ad63d6f881dd50a385a9913c2ed57b441c
Fixes: bug #1185974
11 years agoRename requires files to standard names.
Zhenguo Niu [Wed, 29 May 2013 08:50:03 +0000 (16:50 +0800)]
Rename requires files to standard names.

Rename tools/pip-requires to requirements.txt and tools/test-requires
to test-requirements.txt. These are standard files, and tools in the
general world are growing intelligence about them.

Change-Id: Ifb2735151dc8c7f89ec01d237226669f7051ffa8
Fixes: bug #1179008
11 years agoMerge "Fix typos"
Jenkins [Thu, 30 May 2013 20:26:58 +0000 (20:26 +0000)]
Merge "Fix typos"

11 years agorbd: simplify configuration and use librbd and librados
Josh Durgin [Wed, 22 May 2013 00:49:02 +0000 (17:49 -0700)]
rbd: simplify configuration and use librbd and librados

Add an rbd_ceph_conf options to mirror glance configuration, and use
the existing rbd_user option to choose how to connect to the cluster
instead of relying on an environment variable.  Use these settings
when running command line programs and when connecting via librados.

Use absolute imports so that importing the python librbd bindings
via 'import rbd' does not try to import cinder.drivers.rbd again.

Create some convenience wrappers to simplify librbd and librados
error handling and cleanup. Using these everywhere also simplifies
testing. Mock out all the librados and librbd calls in the tests
so these libraries don't need to be installed.

Remove the local_path() method since it's never used. It was
left over from nova-volume.

There are only three things still relying on the command line:
- importing an image
- exporting to an image
- getting monitor addresses

Importing and exporting on the command line include zero-detection
that would be little benefit to replicate here. librados and librbd
don't have a simple interface to obtain the monitor addresses, so
leave that to a command line tool as well.

Fixes: bug 1083540
Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
Change-Id: I32d059c5e460c2dd8423119b3dbe4a9921f5e907

11 years agoUpdate 3PAR driver session management
Walter A. Boring IV [Thu, 30 May 2013 19:05:26 +0000 (12:05 -0700)]
Update 3PAR driver session management

The 3PAR REST API server has a limited number of active sessions.
This change to the driver removes the ability of keeping the
REST API session alive for the life of the driver.  Now on every
driver entry point, which is synchronized, we login to the 3PAR
and when the work is complete, we log out of the 3PAR.  This has
the side affect of more overhead to the 3PAR but helps eliminate
failed commands due to maximum sessions reached to the 3PAR.

This patch also changes the locking to be the same lock for all
driver entry points and adds locking around the volume stats.
This patch also migrates all of the client code access to the common
class, which simplifies the drivers.

Change-Id: Ibcec4cf1781262521ccbdf99c4ba4167634a09c4

11 years agoFix typos
Eric Harney [Tue, 28 May 2013 14:41:26 +0000 (10:41 -0400)]
Fix typos

Fix servcie -> service typo in test code

Fix typo: "metedata" -> "metadata" typo in
Glance metadata database migration code error message.

Change-Id: I21f4539b7dedeaed6b2e6c3b34803d1b996ce79c

11 years agoAdd testrepository to git ignores.
John Griffith [Thu, 30 May 2013 17:13:49 +0000 (11:13 -0600)]
Add testrepository to git ignores.

Now that we're updating our test runner we need
to add the git ignore rule for testrepository.

Change-Id: I0cf33598ef947750953624eccb5d34d78df462ed

11 years agoFix incorrect copyright
John Griffith [Thu, 30 May 2013 17:05:28 +0000 (11:05 -0600)]
Fix incorrect copyright

Copy paste when drivers were moved out of san.py carried
over the copyright that was included in that file.

Remove that and update.

Change-Id: I38315009a41b9a868c674f63ebf449a78642b2a4

11 years agoMerge "Add missing tests for cinder.db.api.quota_"
Jenkins [Thu, 30 May 2013 16:59:35 +0000 (16:59 +0000)]
Merge "Add missing tests for cinder.db.api.quota_"

11 years agoMerge "Imported Translations from Transifex"
OpenStack Jenkins [Thu, 30 May 2013 16:59:31 +0000 (16:59 +0000)]
Merge "Imported Translations from Transifex"

11 years agoMerge "Fix incorrect status for volume clone from image"
Jenkins [Thu, 30 May 2013 16:53:21 +0000 (16:53 +0000)]
Merge "Fix incorrect status for volume clone from image"

11 years agoMerge "Fix 'Inheritance-based rule deprecated' log warning"
Jenkins [Thu, 30 May 2013 16:52:38 +0000 (16:52 +0000)]
Merge "Fix 'Inheritance-based rule deprecated' log warning"

11 years agoMerge "Return 404 from delete of extra spec if not found"
Jenkins [Thu, 30 May 2013 15:57:21 +0000 (15:57 +0000)]
Merge "Return 404 from delete of extra spec if not found"

11 years agoAdd missing tests for cinder.db.api.quota_
Sergey Skripnick [Wed, 29 May 2013 15:02:56 +0000 (18:02 +0300)]
Add missing tests for cinder.db.api.quota_

blueprint db-api-tests

Change-Id: Ib47c1620ce124919d9f3365a0abaa53b4b3e5cd4

11 years agoMerge "Add missing tests for cinder.db.api"
Jenkins [Thu, 30 May 2013 15:11:42 +0000 (15:11 +0000)]
Merge "Add missing tests for cinder.db.api"

11 years agoReturn 404 from delete of extra spec if not found
Cian O'Driscoll [Thu, 30 May 2013 13:00:54 +0000 (13:00 +0000)]
Return 404 from delete of extra spec if not found

API service now returns a 404 if the volume type that the
user is trying to delete doesn't exist.

Fixed Bug 1090306.

Change-Id: I3c1d4fdac1c2492f532ab4dbfd5c7aaf0ffa6fa4

11 years agoFix incorrect status for volume clone from image
Edward Hope-Morley [Wed, 29 May 2013 22:35:32 +0000 (23:35 +0100)]
Fix incorrect status for volume clone from image

If, when using a Ceph as a backend for Cinder and Glance, you
create a volume from a Glance image, the request is actioned
in one of two ways; (1) if Glance does not supply sufficient
information to Cinder, a full copy is made of the Glance
image or (2) Ceph uses it's copy-on-write capability to do a
clone. If a full copy is done, the volume is marked as
Bootable=True but if a clone is done it is not.

This does not appear to break nova boot from volume at present
but the status displayed by the cinder client is incorrect.

Change-Id: I2959fc3dfaf0e073781583019797bcd44f7789d6
Fixes: bug 1185533
11 years agoMerge "Support for NFS shares with spaces in path."
Jenkins [Thu, 30 May 2013 11:58:53 +0000 (11:58 +0000)]
Merge "Support for NFS shares with spaces in path."

11 years agoMerge "Fixes 3PAR Host already exists error."
Jenkins [Thu, 30 May 2013 11:52:30 +0000 (11:52 +0000)]
Merge "Fixes 3PAR Host already exists error."

11 years agoMerge "rbd: send ceph monitor addresses with connection info"
Jenkins [Wed, 29 May 2013 18:17:08 +0000 (18:17 +0000)]
Merge "rbd: send ceph monitor addresses with connection info"

11 years agoMerge "Ensure that pbr>=0.5.10 is installed"
Jenkins [Wed, 29 May 2013 18:16:03 +0000 (18:16 +0000)]
Merge "Ensure that pbr>=0.5.10 is installed"

11 years agoImported Translations from Transifex
OpenStack Jenkins [Wed, 29 May 2013 18:11:27 +0000 (18:11 +0000)]
Imported Translations from Transifex

Change-Id: Ia0a6aa414a71f25baa1555dc004340e7b29c0f20

11 years agoMerge "test_glance.py: Stub out _get_member_model as well."
Jenkins [Wed, 29 May 2013 17:53:45 +0000 (17:53 +0000)]
Merge "test_glance.py: Stub out _get_member_model as well."

11 years agoSupport for NFS shares with spaces in path.
Anton Frolov [Tue, 28 May 2013 15:34:41 +0000 (19:34 +0400)]
Support for NFS shares with spaces in path.

Unescape share address string listed in nfs_shares_config so that
'x.y.z.w:/foo\040bar' would become 'x.y.z.w:/foo bar' to support
mounting shares with spaces in name.
Use stat command to get available and total space on device instead
of df since it's output is easier to parse.

Fixes: bug #1180984
Change-Id: I552aba91ca8db14130f854c739594a818186bbe2

11 years agoFixes 3PAR Host already exists error.
Jim Branen [Fri, 24 May 2013 16:30:10 +0000 (09:30 -0700)]
Fixes 3PAR Host already exists error.

Fixed “Host already exists” error by,
* Screen scraping error message and extracting the host name
used by the 3PAR backend.
* Cache the host name used by the 3PAR backend for subsequent
methods using host name.
* After a restart, and 3PAR host name it no longer available,
retrieve 3PAR host name using wwn/iqn, if a terminate_connection
fails.

Fixes bug 1182134

Change-Id: Ia08a311af168ae19dbe7c1405db9c606fd878e53

11 years agoEnsure that pbr>=0.5.10 is installed
Jay S. Bryant [Mon, 20 May 2013 20:02:11 +0000 (15:02 -0500)]
Ensure that pbr>=0.5.10 is installed

A bug in PBR levels less than 0.5.10 causes the
/usr/lib/python2.X/site-packages/cinder/db/sqlalchemy/migrate_repo/migrate.cfg
file to not be properly installed.  This causes 'cinder-manage db sync'
to fail due to the missing file.

This change sets the required PBR level to >=0.5.10,<0.6 to avoid the problem.
In the case that user already has a 0.5.X level of PBR installed that is less
than 0.5.10 it should be updated rather than encountering issues as our
test system did.

Change-Id: I532158277a4acbb2f8cb75d03c91097e6d7e0312
fixes: bug 1182203

11 years agoMerge "Remove execute permissions from test files."
Jenkins [Wed, 29 May 2013 14:44:36 +0000 (14:44 +0000)]
Merge "Remove execute permissions from test files."

11 years agoAdd missing tests for cinder.db.api
Sergey Skripnick [Mon, 27 May 2013 14:17:01 +0000 (17:17 +0300)]
Add missing tests for cinder.db.api

Methods covered:

 reservation_*
 service_*
 volume_*

blueprint db-api-tests

Change-Id: Ieb396b3de40eaa2fca9d285f75c0a3d631745e74

11 years agoRemove execute permissions from test files.
Avishay Traeger [Fri, 24 May 2013 08:19:51 +0000 (11:19 +0300)]
Remove execute permissions from test files.

Remove execute permissions from test_volumes.py and
test_storwize_svc.py, which caused the tests defined in them
to be skipped.

Fixing the permissions caused test_create_volume_with_volume_type to
execute, which fails. The cause of the failure is that
(1) LoggingVolumeDriver doesn't implement get_volume_stats(), and
(2) The fake RPC doesn't pass the statistics to the scheduler for
some reason.

Because this test doesn't test the scheduler, just change to use the
simple scheduler and work around the issue.

Change-Id: Ie01331d484c5c48c78b626eff1ad793dbbe7182c
Fixes: bug 1183416
11 years agoMigrate to Oslo DB code
Roman Podolyaka [Thu, 23 May 2013 14:55:10 +0000 (17:55 +0300)]
Migrate to Oslo DB code

 - engine and session handling code is reused
 - integrity errors caused by duplicating entries
   are handled properly
 - the base class for models is mostly reused
   (the only exception is SoftDeleteMixin, which
    requires changes to existing DB schema; this
    will be implemented in BP db-unique-keys)
 - use common exceptions handling code

Blueprint: db-session-cleanup

Change-Id: I7f5eb53daddc57066af9c73caa633b37a2afd9c2

11 years agoCatch and report errors from copy image to volume.
John Griffith [Mon, 27 May 2013 20:04:56 +0000 (20:04 +0000)]
Catch and report errors from copy image to volume.

The copy image-to-volume errors weren't being handled
properly and the result was the lvcreate being retried
even though the lvcreate itself succeeded.

The result of this was misleading errors stating that
the volume couldn't be created because it already existed
(which it did, becuase the create itself was succesful).

Fixes bug: 1183283

Change-Id: I23f05fe64475c3efe285e05a258c4625b801375c

11 years agoMerge "Update cloned volumes QoS settings."
Jenkins [Wed, 29 May 2013 02:46:47 +0000 (02:46 +0000)]
Merge "Update cloned volumes QoS settings."

11 years agotest_glance.py: Stub out _get_member_model as well.
Adam Gandelman [Tue, 28 May 2013 21:34:30 +0000 (14:34 -0700)]
test_glance.py: Stub out _get_member_model as well.

glanceclient's v2 client needs to also have its _get_member_model()
stubbed out as well, to avoid attempts to reach 'fake_host' during
client instantiation.

Update: Only stub when _get_member_model() when required.

Change-Id: I943cbb017ccdbe4f6c994bc83cbefe8693f4672b
Fixes: bug #1185178.
11 years agorbd: send ceph monitor addresses with connection info
Josh Durgin [Sat, 25 May 2013 01:18:30 +0000 (18:18 -0700)]
rbd: send ceph monitor addresses with connection info

Previously we relied on a ceph configuration file on the compute host
for this information. Sending the info directly from cinder makes more
complex setups with multiple ceph clusters talking to the same compute
hosts possible.

Refresh the monitor addresses for each initialize_connection() call,
since monitors may be added or removed while cinder-volume is
running.

Fixes: bug 1077817
Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
Change-Id: I34a1fa16ce1f4524ba25832faf3129303e755100

11 years agoDon't set signing_dir by default.
Dan Prince [Tue, 28 May 2013 16:20:01 +0000 (12:20 -0400)]
Don't set signing_dir by default.

Allow the auth_token middleware to setup its own signing_dir
if required.

Fixes LP Bug #1185098.

Change-Id: Id8766efb34d37ca382b865167ec2465d0a1ce8dc