Age | Commit message (Collapse) | Author | |
---|---|---|---|
2017-07-26 | CI: Cleanup and enhance to specify platform branch | Micah Anderson | |
Fix indentation; setup some more clear variables. Add a third variable to ensure the proper platform branch is checked out. This is necessary because otherwise environment deploys get stuck because the platform directory is in a detached state and then the environment deploys will not proceed because the branch check fails. This will fix #8843. | |||
2017-07-16 | Update tests/platform-ci/README.md how to run platform tests | Varac | |
2017-07-16 | Use unqualified path to bundler | Varac | |
So users can run it locally when they have `bundler` installed as debian package. | |||
2017-07-04 | Fix non-existent CI variable CI_COMMIT_REF (#8844) | Micah Anderson | |
2017-07-04 | Additional fix for CI cache failure | Micah Anderson | |
2017-07-04 | Fix upgrade_test failure. | Micah Anderson | |
See https://0xacab.org/leap/platform/-/jobs/14029 for an example. | |||
2017-07-04 | leap_cli - make fog gem optional | elijah | |
2017-07-01 | [CI] Fix node name for local tests | Varac | |
2017-06-30 | [CI] Run bundle install after checking out different platform branch | Varac | |
2017-06-30 | [CI] Add upgrade tests to ci-build.sh | Varac | |
2017-06-30 | [CI] Fail when CI_JOB_NAME is not recognized | Varac | |
2017-06-30 | [CI] Move more commands into functions | Varac | |
2017-06-30 | [CI] Use CI_JOB_NAME to determine what action to take | Varac | |
2017-06-30 | [CI] Check for mandatory env variables on platform builds | Varac | |
2017-06-30 | [CI] Pin leap_cli so fog-aws is installed | Varac | |
2017-06-28 | platform test - pin ruby version to 2.1.10 when using rbenv, since that is ↵ | elijah | |
the latests that will work. | |||
2017-06-23 | Use default apt component for CI | Varac | |
Resolves: #8828 | |||
2017-06-17 | Renewed commercial cert for platform CI | Varac | |
2017-06-16 | [CI] Use older commit for puppet-catalog-test | Varac | |
After `puppet-catalog-test` has been recently updated it failed in our CI with: File[/etc/apt/sources.list] has notify relationship to invalid resource Exec[apt_updated] See #8814 for more details. Resolves: #8814 https://github.com/invadersmustdie/puppet-catalog-test/commit/ac386793c2c456d2071dd0adda716224128f0bb3 | |||
2017-06-16 | [CI] Use master branch of leap_cli | Varac | |
We moved from develop to master some time ago so we should use master for CI testing as well. | |||
2017-05-23 | [vagrant] Use private networking IP from eth1 if present | varac | |
2017-05-23 | [vagrant] Move $OPTS to vagrant config | varac | |
2017-05-23 | Lint configure-leap.sh | varac | |
2017-05-23 | [vagrant] Use private networking for direct acces | varac | |
Without private networking, the box cannot get directly accessed, only via port forwardings. https://www.vagrantup.com/docs/networking/private_network.html - Resolves: #7769 | |||
2017-05-10 | Remove pixelated submodule from the example provider | Tulio Casagrande | |
2017-04-27 | change environment names to match ci-build.sh | Micah Anderson | |
2017-04-27 | Improve ci-build.sh (Closes #8771) | varac | |
* Change environment names for clarity: . Use staging for deploying to latest . Use production environments to deploy to demo: production/vpn production/mail * Install leap_cli if not present and define default values * Remove old nodes from cached runs * Remove no longer used SEEDS variable * Debugging improvements: . Hide secrets when calling ci-build.sh with xtrace enabled . Use unbuffer to we can add debug output locally . Add debugging to build_from_scratch() | |||
2017-04-25 | Add a production environment for demovpn, demomail | Micah Anderson | |
Pull duplicated bits into a function | |||
2017-04-25 | Fix the pipefail by putting ts inside of ci-build.sh | Micah Anderson | |
2017-04-20 | switch to using CI_ENVIRONMENT_NAME and defaulting to the basic deployment | Micah Anderson | |
2017-04-20 | Enhance ci-build.sh for latest CI builds. | Micah Anderson | |
. Reorganize script to allow for multiple builds . Add latest build, pulling from the ibex provider . Run the build as the cirunner unprivileged user . Set pipefail because job is run within a pipe . Change name of 'build' stage to 'deploy' . Setup an environment for the latest CI deployment | |||
2017-04-20 | Ensure leap command is setup properly for CI | Micah Anderson | |
Add a `leap help` command at the end of the CI setup.sh to ensure that the command is setup properly before continuing | |||
2017-03-29 | Run leap info after deploy | varac | |
2017-03-16 | Use http://deb.leap.se/platform jessie snapshots for platform CI | varac | |
2017-03-15 | [8144] Remove Haproxy tests | varac | |
2017-02-09 | tests: check process by either process scan or service name. closes #8753 | elijah | |
2017-01-31 | Platform CI: Dont run bundle install in parallel | varac | |
Closes: #8684 | |||
2017-01-10 | New unbound runs with debug flag by default | varac | |
2016-12-23 | bugfix: remove deprecated parameter from soledad test script | drebs | |
2016-12-23 | Change regex for soledad process check | varac | |
2016-12-20 | [Vagrant] Remove versioning of provider config | varac | |
2016-12-08 | Use webapp/nickserver:master on leap_platform:master (#8678) | varac | |
2016-11-24 | Merge branch 'bug/avoid-hanging-soledad-sync-test-scripts' into 'master' | Varac | |
Bug/avoid hanging soledad sync test scripts See merge request !53 | |||
2016-11-22 | use lock to avoid running multiple soledad tests (#8590) | drebs | |
If a soledad sync test script is already running, there's no need to run another one. This avoids having multiple test script hanging and eating up resources. We have seen this situation under development circumstances, when the soledad server has been modified in a way that the client hangs and never finishes. | |||
2016-11-22 | add a timeout for the soledad sync test script (#8590) | drebs | |
If for any reason a sync takes too long, the script will timeout and log an errro. | |||
2016-11-09 | Don't do strict checking for nickserver | varac | |
Sometimes nickserver is listed with `ruby /usr/bin/nickserver start` in the process table, sometimes with `/usr/bin/ruby /usr/bin/nickserver start`. We should do proper checking with `systemctl status nickserver` to make sure the service is up though (https://leap.se/code/issues/8579). Meanwhile it's ok to not do strict checking. | |||
2016-11-08 | New nickserver is using fully qualified ruby path now | varac | |
2016-10-25 | Change CI build webapp source branch to master. | Micah Anderson | |
The develop branch was removed, and current master is the same as develop was before. | |||
2016-10-18 | Fix Are_daemons_running test for nickserver | varac | |
2016-10-18 | Use random vm name when running local | varac | |
when using gitlab-runner locally, CI_BUILD_ID is always 1 which will conflict with running/terminating AWS instances in subsequent runs therefore we pick a random number in this case |