1. 31 May, 2017 3 commits
  2. 23 May, 2017 1 commit
  3. 12 May, 2017 2 commits
  4. 10 May, 2017 2 commits
    • micah's avatar
      [bug] fix docker image generation · 681dd27c
      micah authored
      The image location needs to be changed to the new location and
      the docker.io package needs to be available in the image to
      build it
    • micah's avatar
      [feat] Build/push docker (see leap/platform#8755). · 95f378a4
      micah authored and drebs's avatar drebs committed
      When the testing/docker directory is changed, the `build_docker_image`
      CI job is triggered to build the docker image and push it to the
      soledad container registry.
      This is part of an effort to move all the docker images into their own
      project's container registry, and to build/provide them
      automatically (see leap/platform#8755).
      This is accomplished by the following:
      added the necessary overlay DOCKER_DRIVER variable to .gitlab-ci.yml
      added a build_docker_image job with docker:dind service enabled
      enabled container registry in soledad project
      added LEAP_CODE_O_MATIC_PRIVATE_TOKEN to project
      moved docker image from scripts:docker/soledad to soledad:testing/docker
      docker push gitlab-buildpackage:soledad to soledad container registry
      removed gitlab-buildpackage:soledad image and scripts/docker/soledad
      changed image in .gitlab-ci.yml to use new location
      Note: until https://gitlab.com/gitlab-org/gitlab-ce/issues/19813 is
      resolved, we have to use this method.
  5. 09 May, 2017 1 commit
  6. 08 May, 2017 1 commit
  7. 04 May, 2017 7 commits
  8. 02 May, 2017 2 commits
  9. 01 May, 2017 12 commits
  10. 27 Apr, 2017 3 commits
  11. 25 Apr, 2017 4 commits
  12. 23 Apr, 2017 1 commit
  13. 19 Apr, 2017 1 commit
    • drebs's avatar
      [test] tune parameters for asynchronous sqlcipher access · a052c150
      drebs authored
      When hundreds of concurrent document creations are spawned during tests,
      test runs in slower machines may fail with "db is locked" error.
      Increasing the timeout and number of retries in the sqlcipher adbapi
      level may help avoiding that.