We added a kubernetes cloud profile. The node test launch works fine.

We also added a attribute and a resource which is matching the attribute.

Testing a configuration where we require that resource only shows WAITING_NODE.

Output is just when I activate trace logging:

2019-01-15 11:18:41,907 [Thread-74] TRACE com.pmease.quickbuild.DefaultBuildEngine - Allocating step resource (configuration: root/test-config, build: not assigned, step: master)...
2019-01-15 11:18:41,907 [Thread-74] TRACE com.pmease.quickbuild.DefaultBuildEngine - No candiate nodes found, checking cloud profiles...
2019-01-15 11:18:41,907 [Thread-74] TRACE com.pmease.quickbuild.DefaultBuildEngine - Constructing cloud profile virtual nodes...
2019-01-15 11:18:41,907 [Thread-74] TRACE com.pmease.quickbuild.DefaultBuildEngine - Calculating resource count...
2019-01-15 11:18:41,908 [Thread-74] TRACE com.pmease.quickbuild.DefaultBuildEngine - Calculating virtual node candidicates...
2019-01-15 11:18:41,908 [Thread-74] TRACE com.pmease.quickbuild.DefaultBuildEngine - No matching cloud profile found
2019-01-15 11:18:41,908 [Thread-74] TRACE com.pmease.quickbuild.DefaultBuildEngine - Cloud profile checking completed

So trace logging does not show anything why it would not match the cloud profile although I would expect trace logging to really add trace information. In my opinion this is just debug logging not trace logging.


Also tried just setting the node selection on the master step directly to look for a node where the attribute is set without using a resource. No luck either.