Beam Commissioning Software Developed by Fermilab for the LHC

1
Beam Commissioning Software Developed by Fermilab
for the LHC

• Dave McGinnis
• Fermilab
• September 28, 2006

2
Goal

• Fermilab wants to be involved in the
  commissioning and operations of the LHC.
• For example, we are making a substantial
  investment in remote operations
• It is difficult for outside visitors from other
  labs to make important contributions once beam
  commissioning has started.
• Unfamiliarity with the control system
• Critical problems are assigned to in-house staff.
  
• Fermilab will be much more welcomed at CERN if
  it
• Can bring real resources to the table
• Has the ability to solve operational problems
• One possibility is to develop a suite of software
  products to enable Fermilab accelerator
  physicists to make key contributions to the beam
  commissioning of the LHC.

3
Accelerator Software

• Fermilab has a wealth of experience in the
  software issues of running a collider complex
• The Fermilab control system based on Java is
  similar to the LHC Java based control system
• SDA and OSDA are completely written in Java which
  should be straightforward to port to the LHC
  control system
• Fermilab has a large pool of Java software
  expertise to draw on.
• Fermilab is already collaborating with CERN on a
  number of software projects and already has
  access to the CERN control system
• Role based access
• SDA
• Sequencer development
• A number of LARP hardware deliverables will
  require substantial software investment to make
  them useable.
• It might be unlikely that the software effort
  needed for these projects will receive a high
  priority from CERN

4
Lessons Learned from Run II

• The importance of flexible control room
  software during Run 2 cannot be overstated.
• It is impossible to foresee all the software
  applications necessary for commissioning before
  commissioning begins.
• A framework is required that permits quick and
  flexible application development by machine
  physicists.
• A mature shot data analysis framework must be
  available on Day 1 of LHC beam commissioning.
• There is little need to diagnose the accelerator
  when things are working well.
• It is difficult for outside visitors from other
  labs to make important contributions once beam
  commissioning has started.
• Unfamiliarity with the control system
• Critical problems are assigned to in-house staff.

5
Possible Software Products

• Role Based Access
• Sequencers
• SDA and OSDA

6
Software Products

• Role Based Access
• This software is crucial for any outside
  collaboration to work on the LHC
• Because CERN realizes how vulnerable the LHC is
  to external settings modifications, this software
  has become critical to maintain central control
  of the LHC while still preserving the flexibility
  of remote monitoring and diagnostics.
• Already in progress by Fermilab
• Requirements Document prepared and coordinated by
  Fermilab (S. Gysin)
• Requirements Document has been enthusiastically
  welcomed by CERN and is currently under review by
  the CERN IT Safety and Security Group

7
Software Products

• Sequencer
• It is practically impossible to run a complex
  hadron collider without a sequencer.
• Fermilab has developed a sophisticated sequencer
• Forms the backbone of collider operations
• Fermilab understands the strengths and pitfalls
  of sequencers
• CERN wants a sequencer
• LHC controls has already asked for input from
  Fermilab
• Fermilab can collaborate with CERN to design a
  sophisticated sequencer influenced by experiences
  with Run II.
• Fermilab has already started technical
  discussions with CERN on specifications
• Or Fermilab could collaborate to write a
  sequencer in Java under CERNs direction

8
Software Products

• SDA and OSDA
• A very important product.
• CERN has already expressed a strong interest in
  collaborating with Fermilab on SDA
• Fermilabs SDA and OSDA are written Java and
  might be straightforward to adapt to the CERN
  control system
• Might have to develop the concept of states or
  cases at CERN

9
LHC Accelerator Fermilab Software (LAFS) Team
Organization

• LAFS will be a part of LHC_at_FNAL
• The charge to LAFS team will be determined by the
  LHC_at_FNAL Integration Task Force which will
  provide LARP and other organizations a direct
  avenue for input.
• The LAFS team leader (D. McGinnis)
• Reports directly to the leader of LHC_at_FNAL (E.
  Gottschalk)
• Is a member of LHC_at_FNAL Integration Task Force.
• The LAFS team will be organized along project
  lines with direct line management organization
• LAFS will exist of three projects initially
• Role Based Access (Leader - S. Gysin)
• Shot Data Analysis (Leader - E. McCrory)
• Sequencer Development (Leader J. Patrick)
• Membership in LAFS team will be open to people
  across Fermilab (with approval from their
  supervisors) AND outside institutions
• A request for volunteers for the LAFS team will
  be made at an Accelerator Division seminar where
  the project concept and goals will be presented.

10
LAFS Team Composition

• Computer Professionals
• 1.5 FTEs from AD
• 1.0 FTEs from CD
• Operational Experts
• Can write code
• 1.0 FTEs from AD
• Accelerator Physicists
• Can write code
• 1.0-1.5 FTEs
• With 0.5 FTE as the team leader
• This project would last 2-3 years
• At least until the LHC is commissioned
• The workload will change as LHC commissioning
  progresses
• Demand for computer professional effort will be
  high at the beginning of the project.
• Demand for operational specialist and accelerator
  physicist effort will increase as the LHC
  commissioning matures

11
Initial LAFS Team Tasks

• Develop expertise on the CERN control system
• CERN is asking for experienced Java programmers
  to assist them with their current programming
  loads
• see last LHC-Mac presentation by P. Charrue
• We should assign at least 1.5 FTEs to help CERN
• Spend 35 of their time at CERN
• These FTEs in return become the liaisons with
  CERN
• Knowledgeable on the CERN control system
• Knowledgeable on CERN needs
• Builds good will with CERN
• Further define
• software products
• scope of work
• resources required

12
Collaboration with the LARP

• We would like to see if it is possible to develop
  this project inside the framework of LARP
• Take advantage of LARPs contacts and
  interactions with CERN
• A significant portion of LARP hardware
  deliverables will require substantial software
  investment to make them useable.
• without having to be a formal part of the LARP
  collaboration.
• Do not want to detract or compete for resources
  of LARPs major activities
• LARP has kept, until recently, a distance from
  software projects
• It might be difficult to find a common software
  vision across different institutions that
  comprise LARP.
• A possible strategy could be to have this
  software effort fulfill all or part of the
  Fermilab commitment to the LARP beam
  commissioning team
• Presents a very well defined task and
  organization to Fermilab management
• The software effort will require a constant
  Fermilab presence at CERN.

13
Summary

• Fermilab wants to be involved in the
  commissioning and operations of the LHC.
• Fermilab will be much more welcomed at CERN if it
  can bring real resources to the table and has the
  ability to solve operational problems.
• Fermilab has a wealth of experience in the
  software issues of running a collider complex
• The Fermilab control system based on Java is
  similar to the LHC Java based control system and
  has a large pool of Java software expertise to
  draw on.
• Fermilab is already collaborating with CERN on a
  number of software projects
• Developing a suite of software products to enable
  Fermilab accelerator physicists to make key
  contributions to the beam commissioning of the
  LHC.
• A modest team (5 FTEs or 10 people) of computer
  professionals, operational experts, and
  accelerator physicists can make a substantial
  contribution to the LHC