email from Dave Thomson:
" After some poking around by Glenn and I, we understand how the mis-configured LUCI2 over the previous weekend happened. The AO folks use a remotely executed command to trigger a readout on one LUCI, but in doing so the command injects an obsItem into the LUCI Observer GUI (above the current script cursor/pointer position) and then effectively clicks “Go”. With an existing script in the queue this will also execute the commands on the LUCI2 side. The script that was still in the observer GUI configured LUCI2 with the BrG+clear filters and moved the CalUnit in the beam but then there was an error on the readout that stopped further execution on the LUCI2 side until cleared.

This will be tracked in IT#8844. Glenn found a potential bug and is looking into a way to easily clear the queue with a similar remote command so that it can be incorporated into the normal AO data-taking routines.

Recommendation - AO Folks:

Please be sure to clear the queue in the Observer GUI before running any tests where you trigger commands remotely. This can be done on the Observer GUI by clicking “Load” then navigating to /lbt/lbto/luci in the popup and loading either the SAFE_LUCI.xml or nullItem.xml scripts. You don’t need to do anything else after that. SAFE_LUCI may move a camera in LUCI2 depending on how it was left, but the instrument will be safe. The nullItem script has absolutely no instrument configuration information in it, so nothing will move in the LUCI you are not using.

Note also that you normally use savemode=CUBE when taking data this way, so that you get a single FITS file in the return information from the remote readout command. Last Friday the data were taken in savemode=NORMAL so 10 FITS images were produced for each exposure remotely executed. The save mode is NOT currently settable in the remote readout command, so it must be configured manually prior to starting your tests using the readout manager GUI.

If this is not clear how to do any of this feel free to call me or others in SciOps for help.

Recommendation - SciOps folks:

It will probably be “best practice” going forward to load either the SAFE_LUCI or nullItem script into the Observer GUI. In this particular instance the problem was partially because the script left in the queue moved the CalUnit into the beam. Normal observation scripts would include telescope preset information, so there is the possibility of accidentally executing a preset as well. "

How to do it:

  • From the Observer panel.
  • LOAD
  • Directory: /lbt/lbto/luci
  • Select SAFE_*

-- JuanCarlosGuerra - 15 Mar 2023
Topic revision: r1 - 15 Mar 2023, JuanCarlosGuerra
This site is powered by FoswikiCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Foswiki? Send feedback