Cluster wide defaults for failedBuildsHistoryLimit and successfulBuildsHistoryLimit

1. Proposed title of this feature request
Cluster wide defaults for failedBuildsHistoryLimit and successfulBuildsHistoryLimit

2. What is the nature and description of the request?
To proactively prevent failed or finished builds to use up lots of space in etcd and thus require extensive pruning it would be great to have the possibility to set cluster wide default values for failedBuildsHistoryLimit and successfulBuildsHistoryLimit.

3. Why does the customer need this? (List the business requirements here)
With that in place, nobody would need to bother pruning build objects as they are by default limited and are taking care of the clean-up automatically. It would also allow customization if and where needed, which is currently not possible to all extend when using pruning.

4. List any affected packages or components.
builds.config.openshift.io