This document describes the policy for updating the snapd package in a stable supported distro, including LTS.
snapd is the tool to interact with Ubuntu Core Snappy. This package is also used in the generation of the OS snap package and snappy Ubuntu Core images. One of the goals of this project is to keep your system always up-to-date with the latest security fixes and with the newest developed features. It was designed in a way that makes it easily extensible, so every new release provides bug fixes and new features with a low risk of regressions. The team is working with a continuous delivery process which results in a new version ready to be released every week. The snapd package that was delivered at the time of the Ubuntu 16.04 release is also not feature-complete, and the nature of this project makes it important to be able to continue to deliver new features on top of a stable Ubuntu release. Therefore, in addition to critical bug fixes, new features and small improvements are allowed in an update as long as the conditions outlined below are met.
This is the mandatory QA process that the proposed packages have to pass. The following requirements must be met:
each change must be reviewed and approved by at least two members of the ubuntu-core github team before landing into the master branch.
- each change must be fully tested at the unit level.
- all the unit tests must pass in all the supported architectures. They are executed for one arch before the change is merged into master, and for all the architectures during the build of the deb package that will go into proposed.
- each bug fix that affects the user interface must have one QA review. The QA engineer will verify that the bug is fixed in a system with the proposed package installed by executing an automated or manual test.
- all the bugs reported in launchpad that will be fixed in this release must have a link to the pull request that fixes them. These bugs must be marked as "Committed" once that pull request is merged into master.
- all the new user-facing features will be tested in a real system.
- most of these tests are automated and executed as part of the autopkgtest suite of the deb and its reverse-dependencies in a classic ubuntu system, and as part of the automated user-acceptance suite in a snappy Ubuntu Core system.
- the tests that can't be automated are documented and manually executed when there are changes in the code that can affect the feature.
- when a new version is ready to be proposed, the QA team will perform extensive exploratory testing on the areas that will be changed by the release.
The resulting package, with all the changes in place, must undergo and pass the following additional QA procedures:
- upgrade test from previous version of the package. This test must be performed with:
- apt install/upgrade.
- test interaction with classic apt install and update of debs to make sure that snapd doesn't interfere with the classic system:
- install and update a deb with apt.
- test interaction with gnome software center:
- install and update a snap.
- install and update a deb.
The above tests can be performed by any QA engineer.
This is a package new in Ubuntu 16.04 LTS. Once we have another stable Ubuntu version released this should be added to the above process:
- upgrade test from previous distribution to the current one. If the current distribution is an LTS one, the upgrade path from the previous LTS distro must also be exercised.
Requesting the SRU
The SRU should be requested as usual (StableReleaseUpdates) with the description of the bug containing links to automatic testing results (travis unit test, jenkins autopkg tests, and jenkins integrations tests) so that any one can verify the testing occurred and its results. Additionally, the SRU bug should be verbose in documenting any manual testing that occurs an example of a good SRU bug can be found in http://launchpad.net/bugs/1588052. The SRU should be done with a single process bug for this stable release exception, instead of individual bug reports for individual bug fixes. However, individual bugs may be referenced in the from the changelog but each of those bugs will need to independently verified and commented on for the SRU to be considered complete.