Nuvla-job-engine

Changed

- Deployment actions - Put deployment in final state before raising the
exception

Changed

- Fix in Deployment start component regression
- Deployment action do not cat exception to force full stack trace

Added

- Added support for start stop state deployment of subtype application
- Docker cli connector

Changed

- When an error occur during execution of a job, the final state is set to '
FAILED'

Added

- add script to restore zookeeper jobs from elasticsearch

Changed

- Deployment - credential-id renamed parent
- Deployment parameter - field deployment/href renamed parent
- Executor is now mono-threaded. Use multiple executors to run multiple jobs in
parallel.

Added

- Allow authentication with server via request headers
- Nuvlabox delete job
- Deployment state job

Changed

- Reduce the sleep time to 30 seconds after an error when contacting the Nuvla
server
- Deployment stop - delete credential on stop
- Update start deployment options to support restart policy conditions, cpu, ram
- Deployment code reorganization
- Release script fix
- Move reusable parts in util directory for actions
- Connector docker, stop container, if service not found should not return an
error
- Deployment enhance exception management and always leave deployment in a final
state
- Deployment start, stop, deployment resource changed

Changed

- first release of job container