| chad ruppert Sorry, but why does rolling back suck? It is easy, quick and foolproof (thanks @Jenkins team). This is basically the first thing I usually try before even thinking about looking into our backups. Upgrade, something looks fishy, do not overwrite any configuration, especially do not delete "unreadable data" from Jenkins. Try to keep the executors offline, try to isolate the culprit by rolling back upgraded versions one after another or by divide and conquer - whatever fits your style and the situation best. Of course, having a backup keeps blood pressure down while being at it. And having a test machine to test upgrades before going into production would be the best, but I know from own experience that this is a wish not easily fulfilled, especially if Ops or bosses say "you have a backup if something goes wrong". @all who come here now because of this issue (you shouldn't anymore because the 3.0 version has been pulled back) and see this big list of comments and feel lost - here is your way out: do not delete "unreadable data" and simply revert the config-file-provider-plugin back from version 3.0 to 2.18, then restart Jenkins, and everything is back to normal. |