Revision as of 19:19, 23 June 2019

In the process of editing and updating this page. Please contact Juliette Mammei crowder@jlab.org with questions or suggestions.

PREX Main<< Information for Shift Takers << Shift Leader

Experimental Python Alarm Handler

The new python alarm handler lives (for now) in a SBU JLab GitHub Repository

Start a new instance

From within the apar account, execute the ~/bin/ command

AlarmHandler

Hall A/Target Alarm Handler

Figure 1 - Click on the red toolbox in the lower left of the hacweb7 screen (4th icon from the left).

Figure 2 - The alarm handler window when there is a red alarm.

Figure 3 - The window that opens when you click on the alarm handler button. There is a red alarm

Start a new instance of the Hall A Alarm handler

• From the EPICS control computer (hacweb7), click on the red toolbox in the lower left corner (see Figure 1)
• Click on "AlarmHandler" in the popup menu
• The window shown in Figure 2 should appear
• Move the alarm handler window to the upper right of the screen
• Pin the window to all of the desktops
  • Click on the "AlarmHandler" in the upper left corner
  • Mouse over "To Desktop"
  • Choose "All Desktops" at the top of the drop down menu
• Open the Current Alarm History
  • Click the "View" context menu and select "Current Alarm History Window"
  • This is most useful for getting detailed information about timing and values of alarms as they accumulate
  • Information about alarms is also accessible by double clicking on items in the right half-panel

Start a new instance of the Target Alarm Handler

• Make sure the target gui is running on the poltarac computer

cd ~/Desktop/PRex
./tgtgui
Select the ALARMS button

• It should be similar to the Hall A alarm handler, but with difference tracked EPICS variables

Responding to an (audible) alarm

• Click on the alarm handler window in the upper right of the hacweb7 monitor (see Figure 2)
• This will open a window that lists the various pieces of equipment (see Figure 3)
• Please feel free to make strip charts out of any interesting EPICS variables that you find in the Alarm Handler or other GUIs. See Strip chart for more information.
• If an alarm is white then it is a communications failure from an EPICS Input/Output controller (IOC) and should rectify itself in a minute or less (otherwise call the RC)
• If an alarm is red it means that some important parameter is out of range
  • Please keep track of the set points and current value of anything that alarms, and if the alarm isn't transient and trivial please make a HALOG entry about it
  • If the alarm continues for more than a few seconds please try to understand the nature of the issue and contact the RC or other nearby experts if it isn't immediately clear how to resolve it