* Register notification hooks first in init to allow sending .init() failure notifications
* Error from run_script is not consistently passed to deferred functions
---------
Co-authored-by: Frederik Ring <frederik.ring@gmail.com>
* Bump github.com/docker/cli
Bumps [github.com/docker/cli](https://github.com/docker/cli) from 28.5.2+incompatible to 29.0.2+incompatible.
- [Commits](https://github.com/docker/cli/compare/v28.5.2...v29.0.2)
---
updated-dependencies:
- dependency-name: github.com/docker/cli
dependency-version: 29.0.2+incompatible
dependency-type: direct:production
update-type: version-update:semver-major
...
Signed-off-by: dependabot[bot] <support@github.com>
* github.com/docker/docker has been superseded by moby packages
* Fix usage of new filter API
* Base test env off Docker 29
---------
Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Frederik Ring <frederik.ring@gmail.com>
* feat: Add Google Drive storage backend with custom endpoint support
- Implement Google Drive storage backend using service account authentication
- Add support for custom Google Drive API endpoints for testing
- Include comprehensive test setup with mock services
- Add configuration options for Google Drive credentials and folder ID
- Update documentation with Google Drive configuration examples
* remove debug messages
* Improve tests
* Add mounting tip on documentation
* Replace deprecated lib usage
* Fix identation
* Fix googledrive tests
* Fix googledrive support for _FILE credentials
* Remove googledrive test unecessary echos
* Current test setup is too restrictive, allow running a series of executables in different contexts instead
* Split existing tests that test against swarm
* Defining test sandbox in compose file allows swapping with multi node swarm setup
* Test cases can request a multi node swarm cluster
* Docker healthchecks can be used for awaiting ready state
* Compose profiles can be used
GPG is known to have usability issues and is generally cumbersome to
use. age [0] is a modern alternative to GPG that is designed by a
cryptographer that has worked and continues to work on Golang's crypto
packages for years.
Allowing age to be used to encrypt backups dramatically simplifies the
backup process.
[0]: https://age-encryption.org/
* Add "none" compression type
* Add "none" compression to docs
* Use passThroughWriteCloser for "none" compression
* Add test for none compression
---------
Co-authored-by: Frederik Ring <frederik.ring@gmail.com>
* Extend confd test case to test for env var propagation
* Env vars set in conf.d files are expected to propagate
* Lock needs to be acquired when instantiating script
* Hoist control for exiting script a level up
* Do not accidentally nil out errors
* Log when running schedule
* Remove duplicate log line
* Warn on cron schedule that will never run
* Do not await containers when there was an error on scaling
* Add test case for usage with socket proxy
* Add documentation on required permissions for docker-socket-proxy
* Add full list of used Docker APIs to doc
* CONTAINER_START and CONTAINER_STOP is not needed
* Query for labeled services as well
* Try scaling down services
* Scale services back up
* Use progress tool from Docker CLI
* In test, label both services
* Clean up error and log messages
* Document scale-up/down approach in docs
* Downgrade Docker CLI to match client
* Document services stats
* Do not rely on PreviousSpec for storing desired replica count
* Log warnings from Docker when updating services
* Check whether container and service labels collide
* Document script behavior on label collision
* Add additional check if all containers have been removed
* Scale services concurrently
* Move docker interaction code into own file
* Factor out code for service updating
* Time out after five minutes of not reaching desired container count
* Inline handling of in-swarm container level restart
* Timer is more suitable for timeout race
* Timeout when scaling down services should be configurable
* Choose better filename
* Reflect changes in naming
* Rename and deprecate BACKUP_STOP_CONTAINER_LABEL
* Improve logging
* Further simplify logging
* Try running tests in Docker
* Spawn new container for each test
* Store test artifacts outside of mount
* When requested, build up to date image in test script
* sudo is unneccessary in containerized test env
* Skip azure test
* Backdate fixture file in JSON database
* Pin versions for azure tools
* Mount temp volume for /var/lib/docker to prevent dangling ones created by VOLUME instruction
* Fail backdating tests with message
* Add some documentation on test setup
* Cache images
* Run compose stacks with shortened default timeout
* Replace envconfig with env
* Adjust config options and processing
* Added _FILE variant for all password vars.
* Try pathenvconfig
* Revert everything so far
* Use our fork of envconfig with custom lookup
* Use our fork of envconfig with custom lookup
* Test compose timeout option
* Remove secret resolving and specific _FILE config
* Fix timing issue in swarm tests
* Revert "Test compose timeout option"
This reverts commit ab50b21748, reversing
changes made to 0282514b2b.
Revert "Test compose timeout option"
This reverts commit 0282514b2b.
* Use offen/envconfig v1.5.0
* Add info about _FILE in README
* Value > File. Panic on file error. Panic on duplicate presence.
* Test panic on duplicate vars and panic on file error.
* Skip backends while pruning
* Add pruning test step and silence download log for better readability
* Add test cases for pruning in all backends
Also add -q or --quiet-pull to all tests.
* Add test case for skipping backends while pruning
* Adjusted test logging, generate new test spec file
* Gitignore for temp test file
