| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| ... | |
| | | |_|_|_|_|_|/ / / / / / /
| |/| | | | | | | | | | | |
| | | | | | | | | | | | | |
| | | | | | | | | | | | | |
| | | | | | | | | | | | | |
| | | | | | | | | | | | | |
| | | | | | | | | | | | | |
| | | | | | | | | | | | | |
| | | | | | | | | | | | | | |
XenAPI bandwidth polling would fail or be unreliable if there
was any other virtual machines on the same node (such as the nova
compute node). This uses the nova_uuid which is much more
reliable to be there. Fixes LP Bug #918383
Change-Id: I1eb9e6289a8171d6fc93a49b327df474b1e2e503
|
| |\ \ \ \ \ \ \ \ \ \ \ \ \ \
| | |_|_|_|_|_|_|_|_|_|/ / /
| |/| | | | | | | | | | | /
| |_|_|_|_|_|_|_|_|_|_|_|/
|/| | | | | | | | | | | | |
|
| | | |_|_|_|_|_|_|_|_|/ /
| |/| | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | | |
The following commit
commit b63a3e8cbd8fe57d0be119368bf47964543270c2
Author: Josh Kearney <josh@jk0.org>
Date: Fri Jul 13 15:57:54 2012 -0500
Attach ISO as separate disk if given proper instruction.
This gives the option to attach the ISO as a separate drive, like we do in
XenServer, and boot from that instead of replacing the root drive.
Fixes bug 914484.
Change-Id: I0e0560f5fcbc361b60ff1284e7369997dd770fd9
Results in this XML config
<devices>
<disk type="file" device="cdrom">
<driver type="qcow2" cache="none"/>
<source file="instance-00000001/disk"/>
<target bus="virtio" dev="vda"/>
</disk>
<disk type="file" device="cdrom">
<driver type="qcow2" cache="none"/>
<source file="instance-00000001/disk"/>
<target bus="ide" dev="vda"/>
</disk>
...
</devices>
which is broken in many ways
- The same backing file is used in read/write mode on 2 disks
which will cause data corruption on write.
- The 'virtio' bus type does not allow device="cdrom" disks
- The 'dev' attribute is given the same value on 2 disks
To compound the problems, the original commit completely lacks
any tests which would have caught the brokeness and/or identified
what was actually intended.
Change-Id: If1ec286f98fc91f9f411800bbb0adbf2797d2114
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
|
| |\ \ \ \ \ \ \ \ \ \ \ \
| |_|_|_|_|_|_|_|_|_|_|/
|/| | | | | | | | | | | |
|
| | | |_|_|_|/ / / / / /
| |/| | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | | |
Add a generic scheduler retry for build failures. Failed
build requests get casted back to scheduler for retry until
success or the maximum number of attempts is reached. The
number of attempts to make is configurable or can be
simply set to 1 to disable retries altogether.
Partially implements blueprint: scheduler-resource-race
DocImpact:
Adds a new capability to filter scheduler to enable retries of
scheduling requests.
1) New flag: scheduler_max_attempts (int) - Number of attempts to make
to schedule an instance before giving up and settting the instance to
error.
2) New RetryFilter. Avoids re-scheduling to the same host multiple
times. (nova.scheduler.filters.RetryFilter)
Change-Id: I1127caeed4418c75372a42ca7fafacb4f061ffe3
|
| |\ \ \ \ \ \ \ \ \ \ \
| |_|/ / / / / / / / /
|/| | | | | | | | | | |
|
| | | | | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | |
| | | | | | | | | | | |
This fixes bug 1026307
In the periodic task that generates .exists notifications, if there is
an exception generating the notification for an instance, it attempts
to log the exception and continue. Alas, there is an error in the error
handler that clobbers the original error, preventing it from being logged.
Turns out a variable in for loop was 'instance_ref' in one part, but 'instance' in
except clause. Fixed.
Change-Id: Ic887d9078725ec9b676c4f49cf58b567759b6210
|
| |/ / / / / / / / / /
| | | | | | | | | |
| | | | | | | | | |
| | | | | | | | | |
| | | | | | | | | |
| | | | | | | | | |
| | | | | | | | | |
| | | | | | | | | |
| | | | | | | | | |
| | | | | | | | | |
| | | | | | | | | |
| | | | | | | | | |
| | | | | | | | | |
| | | | | | | | | |
| | | | | | | | | |
| | | | | | | | | |
| | | | | | | | | | |
Fixes bug 1026709
The code in quota_reserve acquires SQL locks in a different order than
the code in reservation_commit/reservation_rollback. This can result in
an SQL deadlock under heavy load. Due to an (unrelated) bug in
SQLAlchemy, this can result in this exception:
ResourceClosedError: This result object does not return rows. It has
been closed automatically.
This patch reorganizes the code to always fetch (and thusly lock) the
quota_usages table before the reservations table.
Change-Id: Ia364496a996870d754094915ea0501ff19052037
|
| |\ \ \ \ \ \ \ \ \ \ |
|
| | | |_|_|_|_|_|_|_|/
| |/| | | | | | | |
| | | | | | | | | |
| | | | | | | | | |
| | | | | | | | | |
| | | | | | | | | |
| | | | | | | | | |
| | | | | | | | | |
| | | | | | | | | |
| | | | | | | | | |
| | | | | | | | | |
| | | | | | | | | | |
Avoid maintaining needless copies of this dict, by instantiating
lazily once and only once.
This approach will bring master into sync with the corresponding
patch proposed to stable/essex:
https://review.openstack.org/9446
Change-Id: I8a7cd5fc4fe0effd436e91a6c481df7b0d5a8b01
|
| |\ \ \ \ \ \ \ \ \ \
| |_|_|_|_|_|_|_|_|/
|/| | | | | | | | | |
|
| | | |_|_|_|_|_|/ /
| |/| | | | | | |
| | | | | | | | |
| | | | | | | | | |
Change-Id: I91ea24747eba9f0f0090f923c5728ba2b85eb141
|
| |\ \ \ \ \ \ \ \ \
| |_|/ / / / / / /
|/| | | | | | | | |
|
| | | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | | |
use the helper function instead of the libvirt call directly
to allow NovaExceptions instead of libvirt exceptions to be thrown.
Change-Id: Ieac171e1a2bce1f5542ae73f3d8f6b7a5ccfa8f7
|
| |\ \ \ \ \ \ \ \ \ |
|
| | | |/ / / / / / /
| |/| | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | | |
NotImplementedError is for exceptions, NotImplemented is an error
singleton
Change-Id: I2bc42783c5bb92a32d7c3b0f5efab32a2e2f8fb6
|
| |\ \ \ \ \ \ \ \ \
| |_|_|_|_|_|_|_|/
|/| | | | | | | | |
|
| | | |_|_|/ / / /
| |/| | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | | |
This is dependent on refactoring of libvirt live migration.
Enables live migration using xenapi:
* works when both hosts are part of the same aggregate
* assumes shared storage is configured as the default SR
Limitations in this version:
* only works for images that don't have an external
ramdisk/kernel
Change-Id: I74fc6aae9a615be7cefb1cf07755764df1af957a
|
| |\ \ \ \ \ \ \ \ |
|
| | | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | | |
This patch changes get_lock() to _get_lock() in the compute manager.
This method is not used via rpc, so this name change helps clarify that
it is (at least currently) an internal-only function. Marking it as
such allows the method signature to be changed without updating the RPC
API version and having to maintain backwards compatibility.
Related to blueprint no-db-messaging.
Change-Id: Ib608ba06215e0ab8be23cbd32c284420b5230dba
|
| | |_|/ / / / / /
|/| | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | | |
An exception object NotImplementedError() should be raised instead
of NotImplemented.
Change-Id: I0a3f0d99529b18d613f8fdb5fafa2d0307a66d86
|
| |\ \ \ \ \ \ \ \ |
|
| | | |_|_|/ / / /
| |/| | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | | |
GET /servers/xxx will display the key_name only if the keypairs extension is
being used.
Fixes bug 1024986
Change-Id: Ie37880f63b7595ae3dd2f3b23c6c3097dac2f57c
|
| |\ \ \ \ \ \ \ \
| |_|_|_|_|_|_|/
|/| | | | | | | |
|
| | | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | | |
This change adds currently missing absolute limits to the /limits response.
Fixes bug 1025294
Change-Id: I6a758dd683f70a0d006d5aa2840fcc2cb3243297
|
| |\ \ \ \ \ \ \ \ |
|
| | | |/ / / / / /
| |/| | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | | |
This is unused since edf3e39c remove the v1.0 API.
Change-Id: I1f7c6203ffbe0248c7ef9e95a231c05a0113a347
|
| |\ \ \ \ \ \ \ \
| |_|/ / / / / /
|/| | | | | | /
| | |_|_|_|_|/
| |/| | | | | |
|
| | | |_|/ / /
| |/| | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Fixes bug lp 988335
For fidelity with AWS, where the EC2 CreateImage API is only
supported for EBS-backed instances, we now support this API
for booted-from-volume nova servers.
We create a "place-holder" image in glance with the image data
being effectively empty, and the following properties set:
* the imaged instance's kernel and ramdisk IDs
* block device mapping containing the appropriate snapshot ID(s)
so that we can boot from this image without providing additional
context (such as via the nova boot --block-device-mapping option)
Change-Id: I0b3d18d7922f2ad1bc687fa88e2f5d4cf5aa068b
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Adding this log will help keep track if the
intended compute driver has loaded or it has been
misconfigured.
fixes bug 1021417
Change-Id: I6fd9e3b45aad36056882e8cef12b90bb8d7aea7b
|
| |\ \ \ \ \ \ |
|
| | | |/ / / /
| |/| | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
xenapi/connection.py was renamed xenapi/driver.py, it appears, but
vm_vdi_cleaner was not updated.
Also removed unused optparse import.
Fixes bug 1025912
Change-Id: Ia1a93400dc24455fe91fdf4283eb32bcca4037cf
|
| | |_|/ / /
|/| | | |
| | | | |
| | | | | |
Change-Id: Ic0b04953f27467dfa479dc94582a4c2d3480f53d
|
| |\ \ \ \ \ |
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Currently the same metadata is returned no matter what metadata
version is requested. Metadata is currently implemented to version
2009-04-04 (apart from existing TODOs) as defined here:
http://docs.amazonwebservices.com/AWSEC2/latest/UserGuide/instancedata-data-categories.html
This change adds version checking conditionals throughout
get_ec2_metadata and only includes the metadata if the requested version
should contain it.
When there is a difference between Amazon API implementation and documentation
I've gone with replicating the implementation. So far differences include:
- undocumented hostname introduced in 1.0, deprecated in favour of
local-hostname
- kernel-id and ramdisk-id seem have to been implemented in 2007-12-15,
not 2008-02-01 as documented
get_ec2_metadata has also been reordered so that data is added oldest to
newest by API version. In a future change I may attempt to add
unimplemented API versions (2011-01-01, 2012-06-01).
Change-Id: Ibf6e6e3b9807f29dccd31006515df550bf9bf57b
|
| |\ \ \ \ \ \
| |_|/ / / /
|/| | | | | |
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
In addition to the following changes from openstack-common, this patch
includes some tweak to nova unit tests. timeutils.strtime() will raise
an exception if the year in a datetime object is before 1900.
Changes from openstack-common:
commit ce3071437d1871f77c4d8573cbe5f4ea8c817650
Author: Russell Bryant <rbryant@redhat.com>
Date: Mon Jul 16 10:30:25 2012 -0400
Use strtime() in to_primitive() for datetime objs.
This patch updates jsonutils.to_primitive() to use timeutils.strtime()
to convert a datimetime object to a string instead of just using str().
This ensures that we can easily convert the string back to a datetime
using timeutils.parse_strtime().
Required for the nova blueprint no-db-messaging.
commit 4c9d439ef24f5afdd74aa9153aa8fc772051e6cb
Author: Tim Daly Jr <timjr@yahoo-inc.com>
Date: Tue Jun 26 02:48:42 2012 +0000
Add 'filedecoder' method to the jsonutils wrapper module.
Fixes bug #1017765
After version 3.3.2, the anyjson library will throw a KeyError if
filedecoder isn't present. The filedecoder is just like the decoder
except it takes a file instead of a string, like json.load() instead
of json.loads().
Change-Id: Ib51f0da8641c035371e09047de9abe3cb83203e9
|
| | |_|/ / /
|/| | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
This seems to have been left over from a prior implementation.
Remove it for code hygene.
Change-Id: I6181713c3228632a50af055109f60728ab0f22e4
|
| |\ \ \ \ \ |
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
This refactoring of the libvirt live migration code is
required to enable live migration in the xenapi driver.
This change ensures libvirt specific checks are performed
only when the libvirt driver is enabled.
The complication is that some of these checks require
information to be passed between the source and destination
hosts. For example, when comparing CPU flags.
Change-Id: I7389f0b7f03313d7f04b907f481787dadf0716fd
|
| |\ \ \ \ \ \
| |_|/ / / /
|/| | | | | |
|
| | | |_|_|/
| |/| | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Related to LP 1021373.
Previously we returned a generic 500 Server Error on an over-quota
conditions, whereas this is arguably more appropriately reported
as a 413 status.
Change-Id: I5c1cdc9db54804c512d60e4179c1faa13516d6f9
|
| |\ \ \ \ \ |
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Partially implements blueprint general-host-aggregates: Xen decoupling
* keep rpcapi to compute-controller for Hypervisor-Pools
* Add compute_rpcapi.add_aggregate_host to libvirt
* Remove aggregate states
* Remove One aggregate per host restriction
* Keep Xen support for both hypervisor pools and aggregates
Change-Id: I6ac0f8b6fc31dff589363331e99e0755301f2172
|
| |\ \ \ \ \ \ |
|
| | | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
The strategy for removing the limit is to refactor migration so that
they work nearly identically to snapshots, meaning sequence-numbered
VHDs are rsynced over into a staging-area and then imported into the SR
using the `import_vhds` function.
Change-Id: Ibf5c82c52ae7d505ea9e54d64fcc8b8fdce4d05d
|
| |/ / / / / /
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
added quotas:show and quotas:update to policy.json
defaulting quotas:update to an admin-api default
Change-Id: Ib328912be859512b9c5391d9b7aa317ecf29aeab
|
| | |_|/ / /
|/| | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
This propagates the following patch from common into nova:
Author: Jason Kölker <jason@koelker.net>
Date: Tue Jul 10 11:12:50 2012 -0500
Fix Zmq so it doesn't rely on nova flags
* Embeded newlines in logging makes me cry
* String formatting uses parens not curly braces
* Fix LP1023039
* Hard skip Zmq tests so jenkins passes. Someone will have to fix.
Change-Id: I651aa10861cb26808fa46e74a7db545fb2fd23b8
|
| |\ \ \ \ \
| |_|/ / /
|/| | | | |
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Now it's 'nova.openstack.common.exception', thanks to a more modern
update.py in common.
Change-Id: I4b3caf9014f866da3a5bbf07ab194fc9910159b0
|
