Using a memory database causes tests/test_build to intermittently
fail, because using the same pysqlite3 connection object from multiple
threads - as was done so that the threads shared the same memory database
- is not, in the end, thread safe. One thread will stomp on the transaction
state of other threads, resulting in errors from starting a new transaction
when another is already in progress, or trying to commit a transaction
that is not in progress.
To avoid a significant speed penalty, the session-scope fixture sets up
a database in the pytest temporary directory, which will typically be on
tmpfs. Time to complete all tests:
memory backend: 38 seconds
file on tmpfs: 40 seconds
file on nvme ssd with btrfs: 137 seconds
MBSSQLAlchemy, which attempted to make the memory backend work, is removed.
Session hooks are installed on the Session class rather than on the
scoped_session instance - this works better when we're changing from
one database to another at test setup time.
This also includes `from __future__ import absolute_import`
in every file so that the imports are consistent in Python 2 and 3.
The Python 2 tests fail without this.
This puts backend specific code in either the builder or scheduler
subpackage. This puts API specific code in the new web subpackage.
Lastly, any code shared between the API and backend is placed in the
common subpackage.
This merges the configuration from conf/config.py to
module_build_service/config.py. This also greatly simplifies the logic
in `init_config`. Additionally, `init_config` is no longer aware of
Flask. This will allow us to eventually break up the configuration
between the API and the backend.
It turns out we still have some backend code relies on the flask app,
so revert the change to init_config, and make minor change to create
config object for frontend and backend with the same function.
1. init_web_config: create Config object for frontend, load
configuration from `web_config.py`.
2. init_backend_config: create Config for backend, load configuration
from `backend_config.py`.
And two new classes inherit from `Config` in config.py:
1. WebConfig: representing the orchestrator frontend web configuration
2. BackendConfig: representing the orchestrator backend workers
configuration
Before calling init_{web,backend}_config, check sys.argv, if
"fedmsg-hub*", "celery" or "build_module_locally" is present, it's
running as backend.
Please note that this patch does not change the use of database session
in MBS. So, in the frontend, the database session is still managed by
Flask-SQLAlchemy, that is the db.session. And the backend, running event
handlers, has its own database session created from SQLAclehmy session
API directly.
This patch aims to reduce the number of scoped_session created when call
original function make_db_session. For technical detailed information,
please refer to SQLAlchemy documentation Contextual/Thread-local
Sessions.
As a result, a global scoped_session is accessible from the
code running inside backend, both the event handlers and functions
called from handlers. The library code shared by frontend and backend,
like resolvers, has no change.
Similarly, db.session is only used to recreate database for every test.
Signed-off-by: Chenxiong Qi <cqi@redhat.com>
This also removes the outdated comments around authorship of each
file. If there is still interest in this information, one can just
look at the git history.
The original motivation for this refactor is to reuse make_module and
drop TestMMDResolver._make_mmd. Some tests require a modulemd created
and some tests also require those modulemd to be stored into database as
a module build. The problem is db_session has to be passed to
make_module even if no need to store into database.
Major changes in this patch:
* Argument db_session is optional.
* Arguments requires_list and build_requires_list are replaced by a
single argument dependencies which is a list of group of requires and
buildrequires
* A new make_module_in_db is created for creating and storing the new
modulemd into database conveniently.
* Tests are updated with the new make_module and make_module_in_db.
Signed-off-by: Chenxiong Qi <cqi@redhat.com>
This patch separates the use of database session in different MBS components
and do not mix them together.
In general, MBS components could be separated as the REST API (implemented
based on Flask) and non-REST API including the backend build workflow
(implemented as a fedmsg consumer on top of fedmsg-hub and running
independently) and library shared by them. As a result, there are two kind of
database session used in MBS, one is created and managed by Flask-SQLAlchemy,
and another one is created from SQLAclhemy Session API directly. The goal of
this patch is to make ensure session object is used properly in the right
place.
All the changes follow these rules:
* REST API related code uses the session object db.session created and
managed by Flask-SQLAlchemy.
* Non-REST API related code uses the session object created with SQLAlchemy
Session API. Function make_db_session does that.
* Shared code does not created a new session object as much as possible.
Instead, it accepts an argument db_session.
The first two rules are applicable to tests as well.
Major changes:
* Switch tests back to run with a file-based SQLite database.
* make_session is renamed to make_db_session and SQLAlchemy connection pool
options are applied for PostgreSQL backend.
* Frontend Flask related code uses db.session
* Shared code by REST API and backend build workflow accepts SQLAlchemy session
object as an argument. For example, resolver class is constructed with a
database session, and some functions accepts an argument for database session.
* Build workflow related code use session object returned from make_db_session
and ensure db.session is not used.
* Only tests for views use db.session, and other tests use db_session fixture
to access database.
* All argument name session, that is for database access, are renamed to
db_session.
* Functions model_tests_init_data, reuse_component_init_data and
reuse_shared_userspace_init_data, which creates fixture data for
tests, are converted into pytest fixtures from original function
called inside setup_method or a test method. The reason of this
conversion is to use fixture ``db_session`` rather than create a
new one. That would also benefit the whole test suite to reduce the
number of SQLAlchemy session objects.
Signed-off-by: Chenxiong Qi <cqi@redhat.com>
Most of the issues are caused by the use of SQLAlchemy database session. Some
inline comments describe the issues in detail.
Signed-off-by: Chenxiong Qi <cqi@redhat.com>
This commit fixes issue with following situation:
Module `foo` has been built with `buildrequires: platform: [f28]`
and `requires: platform: []`. It can therefore be used as a buildrequirement
on any platform stream. But if you want to build module `app` which
buildrequires `foo` on `platform:f30`, the MBS won't pull-in the `foo`
module build, because MBS currently limits the modules by the platform
they have been built for.
This commit adds new config option to allow including modules built
against any platform stream.
Currently, we are using just `conf.arches` and `conf.base_module_arches`
to define the list of arches for which the RPMs in a submitted module are
built. This is not enough, because RCM needs to generate modules based
on the base modules which should use different arches.
This commit changes the MBS to take the list of arches from the buildrequired
module build. It checks the buildrequires for "privileged" module or base
module and if it finds such module, it queries the Koji to find out the list
of arches to set for the module.
The "privileged" module is a module which can override base module arches
or disttag. Previously, these modules have been defined by
`allowed_disttag_marking_module_names` config option. In this commit,
this has been renamed to `allowed_privileged_module_names`.
The list of arches are stored per module build in new table represented
by ModuleArch class and are m:n mapped to ModuleBuild.
This also moves the methods load_mmd and load_mmd_file to
module_build_service.utils.general.
This also removes some MSE unit tests with a mix of positive and
negative streams since this is not supported in libmodulemd v2. The
user will be presented with a syntax error if they try to submit
such a modulemd file.
GenericResolver.extract_modulemd is not removed, but deprecated. Call of it
will result in a deprecation message printed. Any new code should call
load_mmd.
Signed-off-by: Chenxiong Qi <cqi@redhat.com>
This patch proposes another code style to create module builds and associated
component builds. The major purpose to make this refactor is to make it easier
to follow up the lines of code and understand the test data and the
relationship between module builds and component builds.
Signed-off-by: Chenxiong Qi <cqi@redhat.com>
Before this commit, the compatible base modules for Module Stream Expansion
have been found without any limitation, just based on the stream version.
It was therefore possible that `platform:lp29.0.0` was found as compatible
module for `platform:f29.1.0` although those platform streams are not
compatible at all.
In this commit, the module can be treated as compatible only if it has
the same virtual stream as the input module. The idea behind this
is that both `platform:f29.0.0` and `platform:f29.1.0` should include
the `virtual_streams: [f29]` in their XMD section which tells MBS
that they are actually compatible. The `lp29` stream will not
have the same virtual stream (most likely it won't have any virtual
stream at all).
The `virtual_streams` is already used for this use-case in `MMDResolver`,
but it was not used to limit the inputs to `MMDResolver` which is what
this commit is doing.
This commit also fixes the issue in `get_last_builds_in_stream_version_lte`
which was simply broken if multiple stream_versions of single base module
existed and their builds had different version. In this case, only
builds with single (randomly chosen) version were returned.
- Keep scratch module builds in the 'done' state.
- Make koji tagging for scratch modules unique so the same
commit can be resubmitted.
- Use alternate prefix for scratch module build components so they can
be identified later.
- Prevent scratch build components from being reused.
- Assorted code and comment cleanup.
Signed-off-by: Merlin Mathesius <mmathesi@redhat.com>
This commit introduces new to_text_type helper method and calls it
for return value of all mmd.dumps() calls. That way, we always
end up with proper unicode string represntation on both python
major versions.
This commit also adds unicode character to description of all
the yaml files we use in the tests so we can be sure MBS can
handle unicode characters properly.
This might be temporary fix, depending on the result of discussion
at https://github.com/fedora-modularity/libmodulemd/issues/184.
* Fix failure test after rebasing on master branch.
* Fix some grammar issues.
* Only check stream collision modules on new created module build.
* Logging message properly.
Signed-off-by: Chenxiong Qi <cqi@redhat.com>
This resolve the stream collision by adding specific RPMs to
module-build-macros SRPM as Conflicts.
For more information about module stream collision, please refer to
docstring in utils/ursine.py
Signed-off-by: Chenxiong Qi <cqi@redhat.com>
ModuleBuild.json now includes xmd/mbs/buildrequires in the output JSON
data. As a result, these APIs will show buildrequires by default.
* /module-builds/
* /module-builds/?verbose=True
* /module-builds/$build_id
Buildrequires is accessible like this:
br_module_names = response_json['buildrequires'].keys()
br_module_dict = response_json['buildrequires'][module name]
Fixes: FACTORY-2201
Signed-off-by: Chenxiong Qi <cqi@redhat.com>
Imagine we have "platform:f29.0.0" and "platform:f29.1.0" base modules.
We also have "DBI" module we want to build agaisnt "platform:f29.1.0".
This "DBI" module depends on "perl" module which is only build against
"platform:f29.0.0".
Currently, DBI build would fail to resolve the dependencies, because
it wouldn't find "perl" module, because it is built against different
platform stream.
This PR changes the MSE code to include buildrequired module builds built
against all the compatible platform streams.
It does so by introducing following changes:
- MSE code uses new get_base_module_mmds() method to find out all the
compatible platform modules. This needed new methods in DBResolver
and MBSResolver.
- For each buildrequired module defined by name:stream, the MSE code then
finds particular NSVC built against each compatible platform module.
Side effect of these code changes is that every module now must buildrequire
some base module.
TestUtilsModuleStreamExpansion._make_module is moved to
tests/__init__.py so that other tests could reuse this to create a
module build.
Signed-off-by: Chenxiong Qi <cqi@redhat.com>
Future use cases will require the ability to find compatible module builds
to buildrequire based on the base module the module used to build. This
commit adds an association table that will contain module build IDs
and the base module they buildrequire.
Addresses FACTORY-3353
This reverts commit 8173040ea6. Since it didn't
provide any real benefit after talking with sgallagh, it seems like we'd better
off not introduce more risk to the next deployment.
