Deployment 2020-2.0 checklist

mrakitin commented 4 years ago

Previsit

[x] Make sure that the current latest environments are pushed to all the bl machines
[x] Skim the IPython startup files in profile_collection
- Check if it's safe to start bsui remotely (does it touch any hardware?) - use git grep .put or git grep caput
[x] Check for outstanding PRs and issues on beamline repos
[x] Check if beamline specific packages pushed to the BL machines
[x] (Optional) Enable CI

Housekeeping

[x] Add BL staff to the BL GitHub organization as owners
[x] Work with BL staff to commit any un-committed changes to their profiles
[x] Tag the profile as-found as 2019C3.x
[x] Discuss with BL staff which conda envs they want to keep/delete and perform the cleanup
[x] Check/update the beamline's inventory with the BL staff (in https://github.com/NSLS-II/playbooks/blob/master/production)

Test deployment

[x] In ~/.bashrc add the following and remember to source ~/.bashrc afterward.
```
BLUESKY_LOG_FILE=/var/log/bluesky.log
BLUESKY_IPYTHON_LOG_FILE=/var/log/bluesky_ipython.log
```
and (if needed) create those directories, making them world-read/write-able.
[x] Append to 00-startup.py https://gist.github.com/jklynch/a4366b8900ec0c03883403455ae711b2
[x] Run profile against new environment
- BS_ENV=collection-2020-2.0rc7 bsui
[x] Update the profile as needed to run:
- [x] Replace .value -> .get() if it's in a Device, or with yield from bps.read() if in a plan
- [x] history.sqlite issue - see https://blueskyproject.io/bluesky/metadata.html#persistence-between-sessions
- [ ] potential issue with images in databroker V2
[x] Run acceptance tests (if any) and run a representative set of scans for the beamline
[x] Tag the updated profile as 2020C2.0
[x] Update BSUI via ansible to use the new environment (http://nsls-ii.github.io/deployment/upgrading-beamlines.html#change-bsui-s-default-environment)

cryos commented 4 years ago

I see some caput:

plans/03-plan.py:    from epics import caget, caput 
plans/03-plan.py:        caput('XF:02IDA-OP{Mir:1-Ax:4_Pitch}Cmd:Kill-Cmd',1 ) 
plans/03-plan.py:        caput('XF:02IDA-OP{Mir:1-Ax:4_Pitch}Cmd:Kill-Cmd',1 )

Also, .put:

startup/04-epu.py:            self.setpoint.put(self.position)
startup/04-epu.py:        self.setpoint.put(position, wait=False)
startup/21-areadetector.py:            getattr(self, 'roi' + str(roi_num) + '.min_xyz.min_x').put(min_x)
startup/21-areadetector.py:            getattr(self, 'roi' + str(roi_num) + '.size.x').put(size_x)
startup/21-areadetector.py:            getattr(self, 'roi' + str(roi_num) + '.min_xyz.min_y').put(min_y)
startup/21-areadetector.py:            getattr(self, 'roi' + str(roi_num) + '.size.y').put(size_y)
startup/21-areadetector.py:            getattr(self, 'roi' + str(roi_num) + '.min_xyz.min_z').put(min_z)
startup/21-areadetector.py:            getattr(self, 'roi' + str(roi_num) + '.size.z').put(size_z)
startup/22-rixscam.py:        self._acquisition_signal.put(1, wait=False)
startup/November 2018 22-rixscam.pyno:        self._acquisition_signal.put(1, wait=False)

PR #31 contains my proposed .get() changes.

cryos commented 4 years ago

PR #31 is merged, testing now. The .put and caput mentioned above have not been changed.

danielballan commented 4 years ago

Looking at that PV

XF:02IDA-OP{Mir:1-Ax:4_Pitch}Cmd:Kill-Cmd

they have an ophyd Device connected to that hardware, or so I infer based on the matching PV prefix:

startup/10-mirror.py:m1 = M1('XF:02IDA-OP{Mir:1-Ax:4', name='m1')

An M1 is a custom compound device defined in that same file:

from ophyd import Device, EpicsMotor
from ophyd import Component as Cpt

class M1(Device):
    x = Cpt(EpicsMotor, '_Trans}Mtr')
    pit = Cpt(EpicsMotor, '_Pitch}Mtr')
    rol = Cpt(EpicsMotor, '_Roll}Mtr')

but EpicsMotor has no signal connected to the Kill-Cmd PV, as you can verify in https://github.com/bluesky/ophyd/blob/master/ophyd/epics_motor.py.

I do not know off the top of my head if Kill-Cmd is common to all EpicsMotors or particular to these. We should investigate that, and then decide whether to add that to EpicsMotor or a special subclass here.

danielballan commented 4 years ago

startup/04-epu.py:            self.setpoint.put(self.position)
startup/04-epu.py:        self.setpoint.put(position, wait=False)

startup/22-rixscam.py:        self._acquisition_signal.put(1, wait=False)

These are correct because they are called inside methods on custom ophyd Devices.

startup/21-areadetector.py:            getattr(self, 'roi' + str(roi_num) + '.min_xyz.min_x').put(min_x)
startup/21-areadetector.py:            getattr(self, 'roi' + str(roi_num) + '.size.x').put(size_x)
startup/21-areadetector.py:            getattr(self, 'roi' + str(roi_num) + '.min_xyz.min_y').put(min_y)
startup/21-areadetector.py:            getattr(self, 'roi' + str(roi_num) + '.size.y').put(size_y)
startup/21-areadetector.py:            getattr(self, 'roi' + str(roi_num) + '.min_xyz.min_z').put(min_z)
startup/21-areadetector.py:            getattr(self, 'roi' + str(roi_num) + '.size.z').put(size_z)

...and same here, though a minor refactor could make that code a bit prettier.

NSLS-II-SIX / profile_collection