Minutes of DAQ/Online Calibration Meeting 12/10/99
Minutes of DAQ/Online Calibration Meeting 12/10/99
Rick St. Denis
==============================
Agenda:
=======
Friday, December 10, 10:00 am, B0 Theater
Please contact Dee Hahn (dhahn@fnal.gov) for Video connections
Minutes of last meeting Rick St. Denis
and Milestone matters 5 minutes
Round Table (5 minutes each)
-----------
COT/Muon: Ashutosh Kotwal
SVX: Steve Nahn
Database: Jack Cranshaw
Calorimeter/DBana: Vaia Papadimitriou
CLC: Alexei Safonov
Consumer: Kaori Maeshima
Access Methods: Rick Vidal
ShowerMax: Jimmy Proudfoot
Presentations
-------------
Proposed Message Protocol Rick Vidal 10
Modification
Joblists Rick St. Denis 10
Discussion
----------
Next meetings
Workshop
Milestones
AOB
Minutes
-------
Minutes of last meeting Rick St. Denis
and Milestone matters 5 minutes
-----------------------------------------
There were no objections to the minutes of last meeting.
Discussion of milestones was delayed and Rick collected them once he
got back to Fermilab.
Round Table (5 minutes each)
-----------
COT/Muon: Ashutosh Kotwal
------------------------------------------
o COT calibrations show a 100ps slewing
o rise/fall time is 800ps
o timing is same as for muons
o test boards are being used to get t0 and slope
o dE/dx is another consideration: it requires 3 params: slope chi2 and
rms
o Switching to new EDM
o making changes
--> This lead to a discussion of whether one should write in EDM2 or
EDM1.
Rick St. Denis suggested that EDM1 is fine since it is familiar
to everyone. Later we can ask a programmer to convert it to EDM2.
This is a general problem that needs to be considered in a broader
context than just calibration.
NOTE: This has been followed up and Rich Glosson can do this. It
is easy for him to do. This needs coordiation with the Online
and offline within the context of the general "edm1 edm2
problem".
SVX: Steve Nahn
-------------------------------------
Tony, Ilya and Steve downloaded constants from database to Silicon (FIB)
but there was a consumer server problem. There was an entry in the
runlist but no data in the tables.
Download works ok. Code was updated this morning. Protocol is sorted
out with Rick Vidal.
Database: Jack Cranshaw
----------------------------------------
o Jack is putting in the SHMX and CLC Roles.
o Fixed date problem -- uses unix time
o Rick Vidal raised the point that Oracle Dates are convenient and easy
to query. Jack responded that this will be done in the future.
Calorimeter/DBana: Vaia Papadimitriou
---------------------------------------------
o Took data Mon (dec 4) and Thrus (dec 9).
o On Monday put date failed, was fixed, but then could not write to DB.
o Looked at old data. DAta in consecutive slot: was 7,14 cent 1 plug but
saw 1,2 and 15. This is fixed. The wedges in the config file were
not current. This was fixed and data were taken. The consumer got
stuck waiting for data. Tony did not understand this problem then the
file was erased. Questions to solve: why is this in an infinite
loop-- is there a problem with the consumer input module?
o Next data, Monday Dec 13.
CLC: Alexei Safonov
-----------------------------------------
o Solved problems
o Next week asseble 20 PM and calibrate
o Question posed: TDC: signed for test pulse and calib signal. Bill
Orejudos volunteered help
o VME crate is in clean room and one is on 1st floor. Needs help from
Bill
o QIE: need card
Consumer: Kaori Maeshima
-----------------------------------------
-- at C++ training, the following email was read:
current status of the consumer infrastructure work is that:
We uncovered a very fundamental problem in ROOT concerning
use of shared memory, and communicating with Rene/Fons (root
team people) to fix the problem. Rene was here last week
and I discussed with him about this.
Two of us (hans and hartmut) are at CERN today/tomorrow meeting with them
to follow up more on this.
This problem must get fixed before we can go on with developing
the consumer infrastructure software. But as far as the user programs
(consumer programs) are concerned, this fix will be totally transparent.
So, I am telling you above just as an information on what we
have been up-to, but really there are no practical concequence
to what I am saying above to consumer users.
We have been busy, trying to identify/fix the problem with
the ROOT team, but there are not much need to talk about this
in length in the meeting, since the problem should get fixed
to the way how we expect ROOT to work. At the current stage,
Rene and Fons realize that this is a very important problem
to be fixed, so I expect the problem will be fixed soon.
(please note that this particular ROOT problem only affects,
in CDF, the online consumer process, not affects offline
software, EDM, etc., etc...., since in consumer process
we make use of a wider capability of ROOT. So, please do not
make other people worry much about this.).
Once this fundamental (but invisual) problem is fixed, we
will go back on working on the consumer infrastructure
which could provide more nice things to the users.
Access Methods: Rick Vidal
-------------------------------------
o Root has TSQL allowing access to ORACLE and mySQL.
This looks like JDBC, Drawback: Only characters -- need to convert
but looks useful in fulfilling need to be able to access other databases.
o Need to sort out account (online account)
ShowerMax: Jimmy Proudfoot
------------------------------------------
Email report:
Basically, little progress except on SMXR sign-off tests
(which have been pretty intense now for a couple of months). This
afternoon my plan is to work with the plug folks to help get them
started on calibrating the plug crate and understanding how to
download to the LUTs. As it happens, we will be using my own scripts
and database to accomplish this: its more important right now for
them to be able to test the hardware and set up to read out a laser
signal. Later (once I cease to be mired in the heart of the SMXR) I
hope to have some some calm time to work more with Steve on the
overall integration.
Presentations
-------------
Proposed Message Protocol Rick Vidal 10
Modification
--------------------------------------------
see transparencies
Joblists Rick St. Denis 10
--------------------------------------------
see pages 3 and 4 of:
http://www-cdf.fnal.gov/internal/upgrades/computing/calibration_db/review991215.ps
Discussion
----------
Next meeting is 7 Jan 2000 at 9am in the CDF Theater.
Workshop: a mini workshop will be held during the 1st on-week: Thurs
13-jan-2000
Milestones will be discussed during the workshop
AOB
---
Steve Nahn asked who is reponsible for alignment. Rick St. Denis
commented that this was an excellent question.