Description of problem: When the cluster is created with wallaby cluster commands, each failover domain in cluster.conf have defined all nodes where the cluster service should be started. In created cluster.conf are missing both ordered directive and priorities directive for each node. As a result, the all cluster services will be started on one node (or randomly assigned to the nodes). Version-Release number of selected component (if applicable): condor 2.3 How reproducible: 100% Steps to Reproduce: 1. add cluster services with wallaby cluster commands 2. check failover domains settings in generated cluster.conf Actual results: <failoverdomain name="Schedd HASchedd1 Failover Domain" nofailback="0" ordered="0" restricted="1"> <failoverdomainnode name="node1"/> <failoverdomainnode name="node2"/> <failoverdomainnode name="node3"/> </failoverdomain> Expected results: <failoverdomain name="Schedd HASchedd1 Failover Domain" nofailback="0" ordered="1" restricted="1"> <failoverdomainnode name="node1" priority="1"/> <failoverdomainnode name="node2" priority="2"/> <failoverdomainnode name="node3" priority="3"/> </failoverdomain> Additional info: For 2nd failover domains the highest priority should be set for 2nd node and vice-versa to allow cluster load balance services around all nodes in the cluster. It's up on future discussion how to add this functionality into wallaby commands. One option is to add it into cluster-add-node, another add it to cluster-create. I'm not sure which way should be better (or where failoverdomains are created in wallaby cluster commands).
MRG-Grid is in maintenance and only customer escalations will be considered. This issue can be reopened if a customer escalation associated with it occurs.