We're not doing it an "official" way, we've just sort of come up with our own method.
We have Hudson running and monitoring SVN, and building every time trunk gets updated. It pulls the code, builds it, and publishes it to Archiva as LATEST. Now, we're not using Ivy's "LATEST" we're setting the actual version to the string "LATEST." This is how all of the dev work goes, and as soon as something starts the march through QA, PERF, SRT, etc. it gets tagged, published to Archiva, and added to the ivy.xml for the release, and only tagged code makes it past the DEV environment. At this point we're even using Ivy/Archiva at release time. There is an ivy.xml file that represents everything that should be pushed in the release. It's resolved in a build.xml that is nothing but the resolve, scp and restart targets. Pull the code from Archiva, copy it to the server, and restart jboss. We have confs in this ivy.xml file that represent each destination folder on the server that we need to move files to, which simplifies things even further. Bailey, Darragh wrote: > > > Not a true ivy question, but more of a use case query. I'm curious as to > how most people who use Ivy and an artifact repository manager such as > artifactory/archiva/nexus, integrate between them with snapshot and > release builds. -- View this message in context: http://old.nabble.com/Using-artifactory-archiva-nexus-with-Ivy-tp29945175p29966000.html Sent from the ivy-user mailing list archive at Nabble.com.
