Closed ambarb closed 1 year ago
note that the exposure time is under .1 seconds for most of these failures. perhaps it is something related to this. Added 2 second sleep between setting the fccd.exposure
and ct_dark_all()
happened again around 19:00 happened again around 22:42 happened again around 23:19 after resetart this time , noted error but things worked as they should.
Subscription value callback exception (EpicsSignal(read_pv='XF:23IDA-EPS{DP:1-Sh:1}Cmd:In-Cmd', name='inout_state1_cmd', parent='inout', value='None', timestamp=1565493340.553536, pv_kw={}, auto_monitor=False, string=True, write_pv='XF:23IDA-EPS{DP:1-Sh:1}Cmd:In-Cmd', limits=False, put_complete=False))
Traceback (most recent call last):
File "/opt/conda_envs/collection-2019-3.0-csx/lib/python3.7/site-packages/ophyd/ophydobj.py", line 269, in inner
cb(*args, **kwargs)
File "/epics/xf/23id/ipython/profile_collection/startup/csx1/devices/eps.py", line 55, in cmd_retry_cb
time.sleep(.5)
NameError: name 'time' is not defined
and two more times before 1 am .
This seems to happen only when having a short exposure time.
added sleep before setting gain and before launching count. see comments in dark_ff.py list as
#adding because it gets hung some where around here TODO
#adding because it gets hung somewhere around here #8
if this works, then will remove one at a time to determine where the failure occurs
Did not work. Issue in a sub-function.
In [6]: RE(myplan())
Starting procedure to acquire darks 80.0Hz or 0.013s.
Current number of images = 6000.
Setting to 50 images.
/opt/conda_envs/collection-2019-3.0-csx/lib/python3.7/site-packages/epics/pv.py:110: UserWarning: The `context` kwarg for epics.get_pv() is deprecated. New PVs will _not_ be created in the requested context.
'The `context` kwarg for epics.get_pv() is deprecated. New PVs '
Moving gains next.
Gain bit set to 0 for a gain value of auto
Ready to begin data collection next.
Transient Scan ID: 120781 Time: 2019-08-12 03:00:19
Persistent Unique Scan ID: 'ea7fce8a-d6c4-461f-b307-08a4440efec0'
Starting procedure to acquire darks 80.0Hz or 0.013s.
Current number of images = 250.
Setting to 50 images.
A 'deferred pause' has been requested. The RunEngine will pause at the next checkpoint. To pause immediately, hit Ctrl+C again in the next 10 seconds.
Deferred pause acknowledged. Continuing to checkpoint.
^CPausing...
@danielballan @mrakitin @jklynch i meant that this is the change from yesterday that we should try to merge and test from https://github.com/danielballan/xf23id1_profiles/commit/87001ee3f7b573b95f24694d23200d92763e8021 . It may address this issue, but I think it will fix another problem as well. I just need to find it.
P.S. I am happy to test. I just don't know how to pick this very specific thing from Dan's repo. Willing to learn.
The following commands should help:
$ cd <you repo directory>
$ git remote -v
$ git remote add danielballan https://github.com/danielballan/xf23id1_profiles
$ git fetch danielballan
$ git checkout darkframes
Cherry-picking the commit from should then work (don't checkout to Dan's branch (last command above)):
# after you fetch Dan's remote in the previous commands, do the following:
$ git cherry-pick 2e57922
Same can be done with https://github.com/danielballan/xf23id1_profiles/commit/87001ee3f7b573b95f24694d23200d92763e8021.
This is duplicated in #38
The following worked for a long time and then the plan got "hung up" . Had to restart bsui in order to regain control. This is an intermittent issue.