New MD stage not triggering after EX

[SrinivasMushnoori]

Lin 475 in sliding window is as follows:

    def _after_ex(self):
        '''
        after an ex cycle, trigger replica resumption
        '''
        self._check_res(self)

self._check_res(self) is basically _check_resume. In line 319, i.e.:

for _r in replica.exchange_list

_check_resume requires a replica.exchange_list. However, this is a NoneType object that causes the new MD stage trigger to fail. Line 362 in the Replica class initializes the exchange_list as self._ex_list=None which is then reassigned as self._ex_list=exchange_list in Line 442

However this assignment is not communicated to the Exchange object that tries to trigger the resume method. Unsure how to set up the MD trigger to pull the list from the replica itself.

Investigating.

[SrinivasMushnoori]

This has been resolved and can be closed.

[andre-merzky]

What was it?

[SrinivasMushnoori]

This was because of incorrect indentation. This, however, is a separate issue from this EnTK issue where pipeline.resume() appears to fail. That has the same symptom of a certain subset of MD tasks to never spawn, but unrelated to this ticket.

[andre-merzky]

Ah, of course - I thought of radical-cybertools/radical.entk/issues/353 indeed...