Issue 11897

Set up jenkins to manage datacube

Reporter: trobertson
Type: Improvement
Summary: Set up jenkins to manage datacube
Priority: Major
Resolution: WontFix
Status: Closed
Created: 2012-09-14 08:41:26.41
Updated: 2017-10-06 14:42:39.495
Resolved: 2017-10-06 14:42:39.475
Description: Currently Tim manually builds urban airship datacube and pushes to our repository.  This should be automated *if at all possible*, so we trace snapshots.  Not sure how we could detect released versions though, but I am not sure this is totally necessary as they seem quite strict on version updates.

Assigning to Lars only for his guidance on this.]]>

Created: 2012-10-22 10:45:26.585
Updated: 2012-10-22 10:45:26.585
released versions 1.2.2 and 1.2.3 are present in maven central:

These are the latest releases, github is on 1.2.4-snapshot:

Is that good enough or do we rely on snapshots from the cube and should build it ourselves via jenkins?


Created: 2012-10-22 10:53:42.938
Updated: 2012-10-22 10:53:42.938
The maven central ones are not suitable for our environment (hence we deploy our own with CDH3 classifiers), and we need to inspect all the code commits to verify nothing drastic has changed.  To automate this:
- fork the project into the GBIF github
- commit the changes to use the CDH3 classifier, and to reference the Cloudera repository
- setup jenkins to deploy (this will likely have to be to the GBIF repository in nexus as we can't deploy to 3rd party snapshots)
- determine if we can automatically detect releases and push to the 3rd part repository in nexus (3rd party cannot have snapshots)

I propose this be reduced to minor priority


Comment: not a priority right now
Created: 2012-10-22 10:57:58.536
Updated: 2012-10-22 10:57:58.536

Author: hoefft
Comment: moving away from cubes
Created: 2017-10-06 14:42:39.492
Updated: 2017-10-06 14:42:39.492