Launch, at minimum, 4 terminals. On three of them ssh into On the other one, ssh into

On all of the terminals, issue this command:
source v3_01_02_01

Now, on one of your terminals on seb-dabhb (hereafter, the SEB terminal), start the seb process:

On the terminal logged into evb (hereafter the ASSEMBLER terminal), start the assembler process:

You won't need to touch these terminals again until you kill everything, so you can push them aside.

On one of the free seb-dabhb terminals (hereafter the RUN CONTROL terminal), start the python script to begin the configuration:
python /uboone/uboonedaq/v3_01_02_01/scripts/ -r 156

The argument on -r is the run number: here, it's 156. Keep track of run numbers in the logbook, or you can check the folders in /data/MRT on seb-dabhb to see what run numbers have already been started.

You should now be configuring the FEMs. While that is happening, go to the last unused terminal on seb-dabhb (your CONFIG terminal), and do the following:

You can now run the config commands here:
configASIC_prod -L 1 -D 1 -P 1 -v
configPULSER -a 0.2 -p 20 (typical pulse)
configPULSER -o (disables output)
The run script will do ALL config commands, so you don't have to do these unless you just want to check connections while the FEMs configure.

Back on RUN CONTROL, when it's done configuring the FEMs, it will pause and wait until configuring the ASICs for the initial test. Press '1' then 'enter' to continue along. Any other character will kill the script (but not the assembler or seb processes). There are a number of pause points throughout the script where this happens, so keep an eye on this. Also, if ASIC or pulser configs fail, it will ask you to respond before resuming, giving you a chance to try reconfiguring on your own.

When the run is finished (at the end of the mapping test), kill RUN CONTROL script by entering something other than 1. Then, CTRL-C the sebApp and assemblerMain processes, AND THEN do the following disgraceful thing on the SEB terminal:
ipcs | grep (username)
Look for a shared memory segment that is 4000000016 bytes. Copy the shmid do
ipcrm -m (your_shmid)
Otherwise, you will crash when you try to run again.

If you need to run some additional tests, do not kill the assembler and seb processes if you quit the run control script---they can sit paused for a long time. Configure the ASICs/pulser as needed, and then on the RUN CONTROL terminal (having exited the python script) or another terminal you've set up in the same way, do
sendRunControlMessage 0 resumerun
That will issue another subrun with 100 events, and then pause automatically. Do that as often as needed.