In November 1998, we will split the two Hall A Spectrometer DAQ systems.  One spectrometer runs on adaqs2 on the adev account, which people are familiar with.  A second, new system runs on adaqs3 on the atrig account.  These two accounts are very similar, nearly mirror images. What experimenters have already learned from the DAQ overview and the CODA 2.x guide for running CODA is mostly still valid, but those guides need to be amended. Someday I'll rewrite and unify them.  For now, let the following supplement the old guides.

Attributes of the Two DAQ Setups

Attribute
Spectrometer being read out	hadron  arm	electron  arm
Computer where to run	adaqs2	adaqs3
Account from which to run	adev	atrig
ROCs involved	ROC2, ROC14, TS1	ROC1, ROC15,  TS0
BPM Crate	ROC14 (as before)	ROC15 (new, similar)
Disks where data are stored	/adaqs2/dataN, N=1,2...	/adaqs3/dataN, N=1,2...
Run numbers	continue from old sequence	numbers > 20,000
Prescale factors	~adev/prescale/prescale.dat	~atrig/prescale/prescale.dat
Deadtime and Disk monitor	datamon (from adev)	datamon (from atrig)
RunControl configuration	hadron	electron
RunControl config for pedestals	pedrunh  (h means hadron)	pedrune (e means electron)
Location of pedestal files	~adev/ped/ped2.dat	~atrig/ped/ped1.dat
Location of scaler_history.dat	~adev/scaler/	~atrig/scaler/

Additional Notes:

• If using these 2-independent DAQs, do NOT use the "TWOSPECT" nor the "PEDRUN" configuration. Use the configuration corresponding to the spectrometer you are running (electron, pedrune, hadron, pedrunh -- see table above).
• The scheme for resetting and rebooting is similar to before.  kcoda works everywhere.  You reboot only the ROCs associated with the spectrometer that has crashed.  Further details about rebooting are at the old CODA 2.x guide.
• Pedestal runs should work as before.  For the new account, the ped files should appear on ~atrig/ped/ped1.dat (for ROC1).   See README there.
• "find_run" should still work. It tries to look on adaqs3 first for runs numbers > 20,000.
• dplot and dataspy will still work.  If you use "-S adaqs3" it tries adaqs3 to find data (note the capital S).   You can also use "-S adaqs2" to force adaqs2 as the server. If the "-S" option is absent, the code will attempt to connect to DD on the local machine. If it does not find DD there, it will attempt to connect on adaqs2, by default. Of course, "-f filename" still overrides all this. Furthermore, most users dont have to think about any of this, since "dhist" will take be adjusted to have the correct arguments.
• It would be wise not to run too many processes (e.g. ESPACE or PAW) on adaqs2 or adaqs3.  Do as much as possible on other computers, e.g. linux box.
• Switching from this 2-independent DAQ to the previous 2-non-independent setup is easy and takes less than 1 hour, but requires access and must be done by expert.
• The information about raw data structure in the data structure writeup has been amended by the new data in ROC15 (2nd BPM crate) and the helicity information on the Electron Arm. It is imperative to check these new data.