cache previous runs from TES

[kmavrommatis]

Hi,
does Rabix --cache-dir have any effect on workflow submitted to a TES? In my case I use funnel with AWS batch as the back end.

I am facing the following two cases that I would like your feedback:

a. long running workflows are failing because the connection between Rabix and Funnel is broken. The tasks continue and successfully finish, but rabix does not know that. is it possible for rabix to recover from this type of error?

b. on a more typical case some jobs of a workflow are failing, due to random problems on the node. Is it possible to recover from the point of failure?

Thanks in advance for your help