Thread pool support in oslo.db was already deprecated[1], and it is not
actually used by Glance.
[1] 74c6bf266e0b86ceb4726cad9268f134d423bae8
Change-Id: I79e934f34f33953c821b6529a34e44dcf2a9688c
This change contains minor version bump to Images API to
indicate latest changes to the API.
Also has config file sync for for M3 release.
These two are combined for saving some gating time due to
busy infra at Train milestone 3 deadline.
Change-Id: I13133c32734751d43cc3afed9b68d015cebd5b6b
The documentation references the section DEFUALT
instead of DEFAULT. As this is most likely a
spelling error which will cause confusion, it is
better if we fix it.
This patch corrects the config option information
and includes an updated set of sample config files
generated from the current code including the fix.
Change-Id: If3c5e334aa1fa2ff5a28f52c00330d42cb9bcf9d
In preparation for the last Ocata beta milestone, let's update our
configuration files. This will also allow for other folks who are
updating config files to create much smaller diffs.
Change-Id: Id644879bb9e93a291c7935eaeacf664bbf46efc2
There's a bug in glance_store ( bug 1619487 ) that results in a
reordering of the store related configurations everytime the tool to
autogenerate configs is used.
This commit generates the Newton config in the deterministic way by
using the suggested solution described on the bug.
For more information and background discussions see change
Iea2c538b37182191445cc5c1f8b2d9dceed06343
Related-Change: I3146b8597a5d89da49b84d6653edacc3067c2c71
Change-Id: Idbea98865ccfc14563ab4357764383e1e7021542
During the release many config options are changed or updated. The
sample config files provided in tree (under etc/*) need to be updated
using the autogeneration tool as these files are being used in different
places. Some examples are devstack gates, downstream packagers shipping
with the same (or using defaults from these files), etc.
Hence, before we cut a release we need to refresh the config files
shipped with tree to match the changes done in the source code during
the release cycle.
This commit is intended to refresh the entire set of config files for
Newton. They include:
* glance api and registry related configs
* scrubber, cache and manage utilities' configs
* any other configs coming from downstream like oslo libraries,
glance_store, etc.
* belonging to affiliated code base like glare.
Change-Id: Iea2c538b37182191445cc5c1f8b2d9dceed06343
Co-Authored-By: Nikhil Komawar <nik.komawar@gmail.com>
Co-Authored-By: Hemanth Makkapati <hemanth.makkapati@rackspace.com>
Change IDs I123be151fdda8037e62ba21600e63c45fdc90c8d and
Ie332236d6e11e0cb63114bc1c68c3a8691f1f1ae fix adding some missing config
options to be included as a part of the sample configs. We need to
refresh the configs to include those to be part of the latest sample
set. This commit refreshes some of the config for that purpose.
This change has been inspired while writing release notes here
If0b8b731c2a1a1c41fc612d506c350073d832efa . It was observed that the
bugs mentioned missing config opts in the sample files and have been
marked fixed. However, in order to make the release notes consistent
with that claim and to mention, as a part of the release notes, that
those bugs have been fixed, refreshing this has become important.
Change-Id: I534a0dc2d9f82ab70517cde12bfd40d4f5ae69e2
Related-Bug: 1568723
Related-Bug: 1568894
The option was used to enable experimental Artifacts API
that was moved to glance-glare service. This config option
does not do anything and is removed to avoid confusion.
Removed apropriate sections from example config files as well.
Change-Id: Ie84f3f65ec88fd30197758ac319ef5b2f4c313e2
Closes-bug: 1556050
This should be the last time reordering happening by the
configgenerator. In future we should see only actual changes but to get
there we need to do this massive change once more.
Closes-Bug: 1553330
Change-Id: Icca0a94a40ba640ca75e01b5cd0d061dacd334ff
EXPERIMENTAL Glance v3 API has been removed in favor of standalone API
(EXPERIMENTAL Artifacts API of v0.1).
This patch introduces a new process entry point to run on a different
port (9494 by default), with its own configuration file and a paste
config.
A controller stub for old /v3 api remains in the glnace.api package for
the compatibility with existing paste configuration which may reference
it. This stub returns a 301 redirects to glare endpoint if it is present
or 410 errors otherwise.
To reuse the existing version_negotiation middleware some refactoring
has been made.
Implements blueprint: move-v3-to-glare
Change-Id: I5b7bd4cdcc5f2a40fc4a5f74bcc422fd700c4fb0
Adding taskflow_executor_opts into the opts so they will be included.
Closes-bug: #1496012
Depends-On: I52ebf810f4699826baa2bdf91d28e24d902cf950
Change-Id: I9c0988a70f691482258f5f3ba9a5cf5601a81ddf
Prior to this change, images were being scrubbed serially while
the image locations, if multiple, were being scrubbed parallely.
For the general case, this may not achieve much performance gain as
the number of images is likely to be more than the number of image
locations per image. Hence, this change attempts to parallelize
image scrubbing while leaving image locations to be scrubbed
serially.
Also, though eventlet greenpool was being used, greening the world
wasn't done. As is, it's unlikely to achieve the intended gains. So,
this change also monkey patches essential python modules.
Finally, this change also makes the pool size configurable. This
offers the flexibility to choose between serial or parallel
scrubbing. Also, parallel scrubbing can be regulated to a desired
level by altering the pool size.
DocImpact
Implements: blueprint scrub-images-in-parallel
Change-Id: I5f18a70cd427e2c1e19a6bddeff317a46396eecc
From Havana version glance was enhanced to be able to clean
pending-delete images up based on their location status db recorded [0].
So from the release, we don't really need two separated clean up logics
in scrubber internal any more - one logic to deal with booking file and
db, other one to handle db duplicately which I'm going to remove here.
[0]
https://wiki.openstack.org/wiki/ReleaseNotes/Havana#Multiple_Image_Locations
Change-Id: I6a1b42f4557d1822f049b2c88ce9f2c00d5c6483
Signed-off-by: Zhi Yan Liu <zhiyanl@cn.ibm.com>
Replace processutils and lockutils modules of oslo-incubator with
oslo.concurrency lib.
Change-Id: Ic1af8753a70f1aada22efe8132e48cbc16e14f3f
Signed-off-by: Zhi Yan Liu <zhiyanl@cn.ibm.com>
1. Sync glance.openstack.common.policy up to latest
version from oslo-inc.
2. Clean useless modules which depended by policy
module and pinned gettextutils module there. For
latter one, we are going to use glance.i18n instead.
* jsonutils
* strutils
docImpact
Closes-bug: #1288178
Closes-bug: #1387973
Partial-bug: #1381870
Change-Id: I84511ab1ee600e618985448dfbfbdc26cb130370
Signed-off-by: Zhi Yan Liu <zhiyanl@cn.ibm.com>
Adding a status field to image's each location property, each location
status can be 'active', 'pending_delete' and 'deleted'.
Under location's status information Scrubber service can make cleanup
based on DB records also but not a dedicated queue-file for each image.
This is third part of this change which covered scrubber DB queue
enablement and double queue supporting.
Partially-Implements BP: image-location-status
Change-Id: I8ee9e9254c8371bbbce8b28d643e3d474b2361bc
Signed-off-by: Zhi Yan Liu <zhiyanl@cn.ibm.com>
It's not immediately obvious where log output will go if the 'log_file' and
'use_syslog' directives are either missing or commented out. This patch updates
the documentation in the config files to mention the stdout default. This
should help remove any confusion when configuring logging for glance.
Change-Id: I36f2a9d30bdee1949e60f413b7f181631c564ed5
Closes-Bug: #1248292
The provided conf file for glance-scrubber was missing the auth settings
stanza required to make it work 'out of the box' in a Keystone environment
such as Devstack.
Also add the "Security Options" line just below for consistency with the
other configs.
Change-Id: I1182fd3a7a79c0f6aeb601d65c31e462fb3610bb
Files written to scrubber_datadir can contain credentials in plain text.
Update schedule_delayed_delete_from_backend to make use of
metadata_encryption_key to encrypt the uri before writing it out.
Fixes LP Bug #1112586
Change-Id: I03ff36f6b57f58a5e5de28bf8d48e7ce0216e5b3
The default for the verbose config option is False and the sample config
file should reflect that.
Also, comment out the defaults of the config options so that it is clear
that the sample file is showing the default in the code rather than
overriding it.
Change-Id: I3913ea54465658d93dc56e014dfe5d911b0541d6
These generic factories allow us to dump the copied and pasted
app_factory and filter_factory methods in the codebase.
The main difference is the paste configuration changes from:
[app:apiv1app]
paste.app_factory = glance.api.v1:app_factory
...
[filter:cache]
paste.filter_factory = glance.api.middleware.cache:filter_factory
to this:
[app:apiv1app]
paste.app_factory = glance.common.wsgi:app_factory
glance.app_factory = glance.api.v1:API
...
[filter:cache]
paste.filter_factory = glance.common.wsgi:filter_factory
glance.filter_factory = glance.api.middleware.cache:CacheFilter
Apart from reducing code duplication, this will also allow us to have
the generic factories inject other data into the apps and filters.
Change-Id: I1d2be5630ab61d29b8948ff88d58e6e1b11c2e5f
* glance-api writes out to a scrubber "queue" dir on delete
* Scrubber determines images to deleted from "queue" dir not db
Change-Id: Ia5574fc75f1a9c763bdef0f5773c2c182932b68a