If running a command over >1 node or a group all the 'bunk' entries follow all legitimate ones.
In addition to using up twice as many Job_IDs, this interferes greatly with the node-status command (all the "most recent exit code"s are labeled as None)
an example of such an entry is:
this is commited immediately after the legitimate one, for example:
If running a command over >1 node or a group all the 'bunk' entries follow all legitimate ones.
In addition to using up twice as many Job_IDs, this interferes greatly with the
node-status
command (all the "most recent exit code"s are labeled asNone
)It very likely results from the merging of code involved with parallelization to develop (https://github.com/alces-software/adminware/commit/87cf03d64ea2b25a692dbcf06577389e5e9e3754).