Hi,
Actually, this is a caching bug in jclouds.
I just issued https://github.com/jenkinsci/jclouds-plugin/pull/104 which circumvents jcloud's image cache.

Cheers
-Fritz

@ nikolay_borisenko

Can you please replace "Exception: blablabla" by a real stacktrace?
This would help me do create a workaround in the plugin, because it turms out that circumventing the cache isn't a viable option either.

Thanks
-Fritz

felfert Want me to merge jclouds-jenkins into jclouds-plugin and make you component lead? Looks like you're the maintainer after all.

Code changed in jenkins
User: Fritz Elfert
Path:
jclouds-plugin/src/main/java/jenkins/plugins/jclouds/compute/JCloudsCloud.java
jclouds-plugin/src/main/java/jenkins/plugins/jclouds/compute/JCloudsSlaveTemplate.java
http://jenkins-ci.org/commit/jclouds-plugin/d636f0910302f8ad65da5539bc7500c058d02287
Log:
Fixes JENKINS-28815

Code changed in jenkins
User: Fritz Elfert
Path:
jclouds-plugin/src/main/java/jenkins/plugins/jclouds/compute/JCloudsSlaveTemplate.java
http://jenkins-ci.org/commit/jclouds-plugin/247b1415cce9e3efdac2cce1764ef8157aec6242
Log:
Fix JENKINS-28815, 2nd attempt

Code changed in jenkins
User: Fritz Elfert
Path:
jclouds-plugin/src/main/java/jenkins/plugins/jclouds/compute/JCloudsCloud.java
jclouds-plugin/src/main/java/jenkins/plugins/jclouds/compute/JCloudsSlaveTemplate.java
http://jenkins-ci.org/commit/jclouds-plugin/c36d74077e0c298e9f0356bc114da4f4060eb086
Log:
Merge pull request #104 from OpenDataSpace/JENKINS-28815

JENKINS-28815 Providing nodes by "Image name" always works as by "Image ID"

Compare: https://github.com/jenkinsci/jclouds-plugin/compare/686a366ff7fd...c36d74077e0c

Changed in git (to some degree).
This is actually caused by caching performed by the underlying jclouds lib. I tried circumventing it as best as possible and it as improved slightly, however it may still need a few minutes or so to actually see changes in the image storage. I talked with jclouds developers about that, and they explained that there is nothing that can be done, because in a big environment (like Amazon for example) there are thousands of images that need to be filtered and it would take ages to iterate over tem. Therfore they do the caching.