Sometimes certain operations are not completely reliable - whether just because
the software is buggy, depends on unreliable network connections, hits
transient resource starvation (eg. out of memory, file handles, etc).
Having the whole pipeline terminate because of a transient issue is annoying
and can waste a lot of time. It would be useful if Bpipe could have an option
automatically retry a pipeline stage, or perhaps even the whole pipeline when
it encounters such a failure.
Original issue reported on code.google.com by ssade...@gmail.com on 21 Oct 2012 at 12:47
Original issue reported on code.google.com by
ssade...@gmail.comon 21 Oct 2012 at 12:47