CHEP Summary Items

1
CHEP Summary Items

• D. Petravick
• 10-23-01

2
Dataflow and Systems Architecture

• Track 9, Triggering and Data Acquisition
• All major Data Acquisition groups are exploring
  Ethernet event building.
• Commodified? -- this is no longer fun
• FNAL capability to propose systems based in
  understanding of the architecture of switches.

3
Local Wide Area Networking

• Had computer security talks.
• Gave the FNAL strong authentication talk
• IHEP talk about security infrastructure.
• Begged (in my mind) the observation - I hope
  this is the last CHEP that security seems treated
  as the realm of the specialist
• Perhaps this view is skewed by the sessions I was
  able to attend.

4
Plenary Pique-ers

• Status of Preparation for LHC Computing Manuel
  Delfino(CERN)
• The LHC Experiments' Joint Controls Project
  (JCOP) Wayne Salter (CERN

5
Mixing up Online and Offline

• Gave the talk in Track2 Control Systems Fermilab
  Distributed Monitoring System (NGOP) Tanya
  Levshina (FNAL)
• The session included a lively discussion about
  the functionality of a commercial control system
  tool based on SCADA, and investigated by CERN.
• It seems to subsume the monitoring functions of
  the FNAL NGOP software project. it is being
  pursued by the re-formulated PEM project at CERN
  according to David Meyers (CERN) and Clara Gaspar
  (CERN).

6
Conventional Magtape Storage

• All labs using STK equipment.
• CERN reports a very low error rate on the 9940
  drives, about a leader block problem every
  30,000 mounts).
• Evaluating LTO, could not get a statement of
  direction.
• Castor will read HPSS tapes.
• CERN is discontinuing HPSS

7
Friday Summary Session

• A trend to unify online system software systems
  and offline software systems.
• Use of commodity hardware at all labs in
  triggering and event building systems.
• Use of control system software to monitor and
  control offline systems like farms.
• Use of commodity networking hardware for event
  building and other such triggering applications.
• Use of commodity operating systems like LINUX in
  online systems/ embedded hardware.

8
Hopes for steering the next CHEP

• Seeing security discussed in the context of GRID
  experiments system architecture.
• Better representation of all Labs large
  expenditures in storage.
• Storage system inter-operability, not
  contentious.
• Less presentation of stunts
• Infrastructure reports as labs become endpoints
  on the grid -- a true measure of acceptance.