Run ongoing and rate expectation

Created by: gladky

I would like to do step back and redefine the conditions for firing failure analysis. Right now 'run ongoing' is an essential part of conditions for failure analysis. 'No rate when expected' (NRWE LM) depends on 'run ongoing' and requires it to be true. And all failure analysis starts from NRWE.

NRWE has to have many exceptions though. Take a look at this example:

In above case we consider run ongoing true when TCDS/L0/DAQ are paused/pausing/etc. Filtering this states it then pushed to NRWE as we don't expect rate during this paused periods. I understand that the idea of 'run ongoing' was to have one such condition per run number. I'm ok with this. We also mentioned that there may be some problems identifying run ongoing because we have to consider state transitions which we can miss polling each 2 seconds. And this makes me worried - it can be error prone.

My proposal is following: We leave 'run ongoing' but replace it as a substep of NRWE with new LM 'expected rate' with simple condition (L0=running & TCDS=running & DAQ=running). This way we will fire failure analysis in well defined conditions and avoid any transition-missed boundary cases.

I would like to ask you to comment on this new LM 'expected rate' idea, especially if it's conditions make sense. Here is how it would look like:

Note that I've recently introduced the LM 'transition'. Here LM NRWE would activate if (expectedRate=true & noRate=true & transition=false)