Closed marcionemec-daitan closed 5 months ago
rolled over from Sprint 24
As bcgov/nr-spar-data-sync#27 issue was solved, including a solution for this current ticket (Jenkins Cron file migration to Jenkins Polaris) this ticket will handle all documentation about the changes in confluence page.
Be warned: migration is only to bring minimal code execution into new environment. Vault and secrets management will be handled in bcgov/nr-spar-data#47 issue.
Documentation can be found in: https://apps.nrs.gov.bc.ca/int/confluence/display/FSADT2/Jenkins+configuration
Closing ticket.
Situation:
While trying to build/deploy the Vault issue solution (see bcgov/nr-spar-data#47 issue), Jenkins Cron server raised error trying to connect in github.com. This error is impeditive to deploy the pipeline of the project.
Asking to 1Team for a guidance, the team suggested us to use the Jenkins [Polaris server:](Polaris Jenkins: https://apps.nrs.gov.bc.ca/int/confluence/display/1TEAM/Polaris+Jenkins)
So, to use the Jenkins Polaris server, Data Sync Engine needs to be containerized to be build/deployed
Background:
Although we already had used a dockerized approach to test issues, the code itself was build inside the Jenkins server instance, using the configuration "agent:all" . To avoid security issues, 1Team asked us to use "agent:podman" configuration.
Analysis:
To use podman, we need to intall it in a Ubuntu version 20.10 or newer. As our previous Ubuntu version in a WLS Windows system was 20.04, it was needed to upgrade the version. More info here.
Other issue is: Jenkins Cron is based in a Windows System, so, Jenkins file pipeline configuration uses the "Windows batch" command. But Jenkins Polaris is based in a Linux system, so pipeline configuration needs to be handled using sh/bash commands.
The jenkins file translation would be handled by issue bcgov/nr-spar-data-sync#27 .