Running the DAQ

Warm Start

• In the CODA window on URS1, you'll find the Run Control user interface.
• Check that the run Configuration (see Static Parameters just under the Run Control Buttons) reads run_config.
• If the Configuration is not run_config, click the Configure button and select run_config from the pull down menu and click Ok, then click the StartRun button.
• If the Configuration is run_config, click the Start Run button.
• Respond to the runLog window by clicking the appropriate choices, entering comments, and/or accepting the defaults, and then clicking the OK button.
• Check that the run has begun by noting the increasing event count in the Run Progress, Events this Run, window, the event Rates, and/or the updating event rate plot.

Cold Start

• In an open terminal or xterm session, within the /home/radphi directory, type source dosetupcoda.
• Type codamaster; a small window will appear listing all possible run configurations.  Choose run_config.
• A more colorful user interface will appear.  Pull down the Config menu and select Enable Buttons.
• If the CMLOG display is missing, click Message Logger.  When the display appears, select Connect in the File pull-down menu and then Update in the Options pull-down menu.  Just click OK in response to the message that appears.
• Returning to the Coda Master display, click Event Builder and then Event Recorder.  Two small, yellow screens should appear in the lower right of your session.
• Click Run Control on the Coda Master display.  After a minute or two, the Run Control GUI will appear (abruptly).
• On the Run Control GUI, click Connect, in the upper left corner.  The GUI will eventually transform its appearance into a control interface.
• Follow the directions for Warm Start, above, noting that the run Configuration must be selected.

End Run

• Click End Run to end the run.  The should be done whenever the number of events reaches the maximum recommended, the file size reaches the maximum recommended, the beam and/or experimental conditions change or are about to change dramatically, or when the data disk is very nearly full.  If the data disk is almost full, be sure to follow the procedure below to change data disks.

Trouble Shooting

Startup Failure

•  Click OK to the failure notification, click Reset in the Control box at the upper left, and OK to the ensuing message.  Try starting again.
• If the second attempt fails, note the failure in the log book, including relevant information from the server messages at the bottom of the GUI.
• If a server message like "ROCm has not reported status for n seconds", where 1< m < 6 and n >10, try logging into the ROC and rebooting:

 

ROCm	CPU Name
ROC2	urvme2
ROC3	urfb2
ROC4	urfb1
ROC5	urfb3

Open a window and type telnet cpuname, where cpuname corresponds to the ROC listed in the table above.  If the login was successful, you should find a right-pointing arrow as your prompt.  Type reboot.  Your cursor will freeze and you will get no response.  Type ^ \ (back-slash while holding the control key down) to escape to the telnet> prompt and then q <cr> to exit telnet.  In a few moments, the notice "ROCm" has now "booted" will appear in CMLOG.  Reset Run Control and try starting again.  If the failure persists, call an expert.
• If you cannot telnet to the ROC, try  remote reset  and, after CMLOG confirmation that the reboot was successful, start again.  If this is unsuccessful, call an expert.

Event Rate Changes Dramatically

• If the event rate changes dramatically, typically going to zero or to a very, very large number, stop the run if the condition continues for a significant fraction of the time data has already been taken on the current run.   Check for high voltage trips, changes in the beam or experimental situation (perhaps CLAS personnel are performing a harp scan or the accelerator has gone down), and unusual rates in the visual scalers.  Try to understand the situation, perhaps correcting it, before resuming data taking.

Data Disk Filled

• As part of your shift duties, you should check Disk_Monitor that ought to be running in the CODA window (if it is not, open a terminal or xterm session, type source /home/radphi/setup, and then type Disk_Monitor.
• If a data disk is very nearly full, be sure to stop the current run and note in the logbook that you have done so, and why.
• In a terminal or xterm session, in the /home/radphi directory, type rm data, and then type ln -s next-disk data, where next-disk is the next disk to which data will be written.  You may choose from /data1, /data2, /data3, /data4, /data5, /data6, /data7; Disk_Monitor will inform you as to which of these is empty.  If no disks are available, call an expert.