[WHIRR-590] ssh port timeout fails the whole cluster

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 0.7.1
• Fix Version/s: 0.7.2
• Assignee: Doug Daniels
• Reporter: Doug Daniels
• Created: 10/May/12 23:15
• Updated: 28/May/12 19:53
• Resolved: 13/May/12 18:05

Description

If the ssh port does not open before the jclouds.compute.timeout.port-open timeout, the whole cluster fails to launch rather than starting a replacement for the failed node and continuing provisioning. You can simulate this by setting your jclouds.compute.timeout.port-open to something very small, such as:

jclouds.compute.timeout.port-open=5

I believe that the failure happens because the node winds up both the successfulNodes and the lostNodes reported by the RunNodesException. I filed a bug for this in jclouds: https://code.google.com/p/jclouds/issues/detail?id=923