for the LHC
Elliott McCrory, Timofei Bolshakov
and Dennis Nicklaus
Outline (From McGinnis)
Why the project is important to CERN?
Why is Fermilab suited to do this
What needs to be done on the project?
What type of people are you looking for
to work on the project?
29 Sept 2006 McCrory, Bolshakov, Nicklaus AD/LAFS 2
Overview (6 slides)
Sequenced Data Acquisition (SDA)
Diagram and Definition (slides 2 & 3)
Functional description (4)
Rip-roaring success at Fermilab
T.B.: Excellent ideas for an “SDA II”
Can think of SDA as a general tool
for any repeated process
Industrial applications, etc.
29 Sept 2006 McCrory, Bolshakov, Nicklaus AD/LAFS 3
Definition of SDA (3/6)
A special purpose data logging system
Organizes accelerator data to facilitate analysis.
Collects and stores data according to the stages of the
accelerator cycles or modes, instead of using wall clock time.
It organizes and presents the data so people and programs can
use it easily.
Defines a temporal abstraction for acquiring and storing data
about a repeated process, like a collider store.
Encapsulates the progression of time through the process by
breaking it down into a specific, repeated series of stages.
Data are collected within each stage and are stored in a
database indexed by the stages.
The stages are called Shot, Case and Set.
The cases have descriptive names, in addition to a numerical
index, facilitating understanding of and communication about
29 Sept 2006 McCrory, Bolshakov, Nicklaus AD/LAFS 5
SDA Functional Description (4/6)
API for both
• Look at and plot shot data
– Java implementation
• Setup data acquisition
Summary Table creation
Offline Shot Data Analysis (OSDA)
• A Java API
– A higher level Java API
29 Sept 2006 McCrory, Bolshakov, Nicklaus AD/LAFS 6
SDA Success at Fermilab (5/6)
Almost everyone uses it
Operators, physicist, engineers, bureaucrats.
Provides consistent and accurate, yet
conversational, data from the complex
• Transmissions, emittance growth
Summary information is readily available
Excel spreadsheet analysis
• Super Table
• Derived Tables
Java Analysis Studio (JAS)/AIDA
• Contact: Derwent
But, requires a notion of STATES
And, probably, a Sequencer!
29 Sept 2006 McCrory, Bolshakov, Nicklaus AD/LAFS 7
SDA II For Fermilab (6/6)
Based on experience with Tevatron Collider
Simpler and cleaner implementation
Better data abstraction
• Class Atom
• Easy to add new types of Atoms
Better XML support
• DB, structure, DAQ and history: All XML
• Simplifies data flow
Easier to add new types of “Shots”
29 Sept 2006 McCrory, Bolshakov, Nicklaus AD/LAFS 8
Why Important for LHC
Enhance and extend their data logging
DAQ on event
Any type of data
• Help with recording fast/dense data streams
Data stored on LHC over-arching events
Create a view of the LHC that can be
interpreted, analyzed and discussed
Simple tools for studying LHC
Shot Data Analysis
• Super Table and other summary tables
• Automatic analysis processes
Automatic creation of summary information about LHC
• LHC Super Table?
29 Sept 2006 McCrory, Bolshakov, Nicklaus AD/LAFS 9
Why Is FNAL Suited?
We have long experience with SDA
A functionality that we live/die by, but they
don’t have yet
We have a concrete beginning
Coupling with Sequencer
Enthusiastic core group
29 Sept 2006 McCrory, Bolshakov, Nicklaus AD/LAFS 10
“Design Thoughts” document sent to CERN
Preliminary response from Mike Lamont
He is enthusiastic
• Suggested concrete DAQ “vertical tests” demos
Existing data summary/analysis schema (from LEP) are “a
• Asynchronous DAQ to Measurement DB or to Timber
• Can record UTC time stamp of “events”
• Correlate these two types of readings
Throughput issue to their data logger, Timber?
• “2000 point snapshot at 40 Hz?”
SDA-II is running at Fermilab
No DAQ, yet
• Real Soon Now
29 Sept 2006 McCrory, Bolshakov, Nicklaus AD/LAFS 11
SDA-II “To Do”
Tevatron LHC (cont’d)
DAQ for SDA II Timeline/STATES definition
Demonstrate full, operating and implementation
system in parallel to “SDA I” • Together with Sequencer
LHC Initially: build SDA on top of
Requirements Document Timber??
• Communicate with LHC folks • Covers maybe 90%
on what they want to have in DAQ for things not covered
order to make this useful to
them by Timber
• CERN will probably require a • XML DB or push into Timber
formal document DB??
– May require new features
Full Functional Specifications for Timber
• Again, CERN may require a Fix up existing StorePhysics
formal document API
Atom Definitions • For use at FNAL and at CERN
Figure out what else is on this
29 Sept 2006 McCrory, Bolshakov, Nicklaus AD/LAFS 12
Who Can Help?
Super Table/Excel dabblers
29 Sept 2006 McCrory, Bolshakov, Nicklaus AD/LAFS 13