The new cloud package set is a JSON file which has proper mapping of
source packages that Britney has to the binary packages that
cloud-test-framework expects.
Add optional reporting of cloud tests. Cloud reporting includes the
package name, version, series and whether the test result was
pass/fail/error.
Cloud reporting requested by partners to be able to determine the amount
of packages tested.
Additional verdict info is being provided by the CloudPolicy to improve
clarity on the state of cloud tests. When tests pass this will be marked
in the verdict. Also when a package is skipped because it is not in the
cloud package set this will be made clear in the verdict.
Cloud policy now uses the ADT_PPAS field as the source but will only
take private PPAS due to requiring a fingerprint. If ADT_PPAS is empty
then the 'proposed' archive is used.
Adds logic to try and retrieve the install source of a package from the
logs. This is useful if multiple PPAs are defined since the policy won't
explicitly know when contains the package under test.
Changes made
- Multiple hardcoded fields moved to config
- Series is now retrieved from options
- Pocket is now called source and retrieved from config
- Adds source type config which can be either archive or ppa
- Returns REJECTED_PERMANENTLY policy verdict when test failures or
errors occur. Adds verdict info the the excuse.
The cloud policy is currently used to test whether the proposed
migration will break networking for an image in the Azure cloud. Cloud
testing will likely increase in scope in the future.
This reverts commit 7ff150ced70d0db565378683909e22dfac5383e1.
Sadly we can't use the X-Auth-Token approach due to implementational details (the token is valid only for an hour). So we need to switch back to using swiftclient.
This way there's less secrets that need to be shared and less new code to introduce. We also modified the test tooling to be able to check for authentication tokens in the queries.
On some distros (Ubuntu), arch:all packages are built along with one of
the architectures. We shouldn't be listing 'all' as its own arch in this
case. Instead we filter out the binaries except for on the
'all_buildarch'.
The apt version comparison sorts 'blacklisted' greater than most version
numbers, which means that we accidentally apply force hints for version
'blacklisted' to all uploads. Since this is the only case of a hacked
version number, let's special case it so that 'blacklisted' hints only
match packages with 'blacklisted' version.
This is a test carried over from Ubuntu which ensures that a package
which builds no binaries on an arch doesn't have tests requested.
It was disabled. Enable it.
Most DKMS packages do not declare Testsuite: autopkgtest-pkg-dkms, but
we can detect this anyway, and this way we can enforce that the module
is buildable.
It works like this. We wait until all tests have finished running. and
then grab their results. If there are any regressions, we mail each bug
with a link to pending-sru.html. There's a state file which records the
mails we've sent out, so that we don't mail the same bug multiple times.
We want to treat linux-$flavor and linux-meta-$flavor as one set in
britney which goes in together or not at all. We never want to promote
linux-$flavor without the accompanying linux-meta-$flavor.
Add a new LinuxPolicy which runs after most of the other policies, which
invalidates linux-foo if linux-meta-foo is invalid.
For packages with lots of reverse dependencies, new versions of those reverse
dependencies may keep on showing up in testing. If migration is blocked until
the results for these new version, migration may take extremely long. If there
are results for the current trigger but for the previous version of the reverse
dependency, use those until the fresh resuts are available.
Similar for the reference runs.
Currently, britney only schedules reference runs when they don't exist. It does
strip out runs against older versions of the autopkgtest, but the current version
may exist for a while and the reference run can be old. So, add an option to
ignore old results.
