Transcript ILC Control and IHEP Activity

ILC Control and IHEP Activity
Jijiu. Zhao, Gang. Li
IHEP, Nov.5~7,2007
CCAST ILC Accelerator Workshop and
1st Asia ILC R&D Seminar under JSPS Core-University Program
Introduction
• The International Linear Collider (ILC) is a 200- to
500-GeV center-of-mass high-luminosity linear
electron-positron collider, based on 1.3-GHz
superconducting radio frequency accelerating
cavities.
• The machine operates at a pulse repetition rate of
5-Hz, with each 1-ms beam pulse comprising
~3000 bunches.
Introduction (cont’)
• The control system overall design is evolving as
details of the accelerator technical design are
developed.
• The Control System Reference Design serves
these purposes:
– Establish a functional and physical model for costing
purposes
– Establish a starting point for engineering design and
R&D efforts
– Communicate our vision of the control system
Scope of Controls
•
Computing Infrastructure
–
–
–
–
–
–
–
•
Computer Center
Business Computing
Computing Networks
Desktop Support
Engineering Support
Computer Security
Management
Controls System
– Central Computers
– On Site Control Room
– Controls Services
•
•
•
•
•
•
•
Operator Interface
Automation
Logging
databases
Data Archival
Alarms
Diagnostics
•
Interfaces to Technical Systems
– Front Ends
• Hardware
• software
– Cabling
•
•
•
•
ATCA High Availability
LLRF Controls
Beam Feedback System
Protection Systems
• Machine Protection
• Personnel Protection
• Beam Containment
•
Network Infrastructure
•
Assembly and Testing of
Controls Racks
Requirements of Controls
• High Availability
– Controls System allocation
• 2500 hours MTBF (Mean Time Before Failure)
• 5 hours MTTR (Mean Time To Repair)
• 15 hours downtime per year
• CS availability – 99% to 99.9%
• Each system 99.999% available
– Standardization
– Diagnostic Layer
• Scalability
– ~100,000 devices should be controlled, millions of control points
• Automation
– Sequencing, automatic startup, tuning, etc.
– Slow and Fast Beam Based Feedback
• Timing and Synchronization
– Precision RF Phase References
– 0.1% amplitude & 0.1 degree phase stability
• Remote Operation
– Enable Collaborators to participate more fully
Functional Model
Client Tier
•
•
GUIs
Scripting
Services Tier
•
•
•
•
•
•
“Business Logic”
Device abstraction
Feedback engine
State machines
Online models
…
Front-End Tier
•
•
Technical Systems Interfaces
Control-point level
Functional Model (cont’)
• Client Tier:
– Provides applications with which people directly
interact. Applications range from engineer-oriented
control consoles to high-level physics control
applications to system configuration management.
– Engineer-oriented consoles are focused on the
operation of the underlying accelerator equipment.
– High-level physics applications require a blend of
services that combine data from the front-end tier and
supporting data from the relational database in the
context of high-level device abstractions (e.g,
magnets, BPMs)
Functional Model (cont’)
• Services Tier:
– Provides services that coordinate many activities
while providing a well-defined set of public interfaces
(non-graphical).
– Device abstractions such as magnets and BPMs that
incorporate engineering, physics, and control models
are represented in this tier.
– This makes it possible to relate high-level machine
parameters with low-level equipment settings in a
standard way. For example, a parameter save/restore
service can prevent two clients from simultaneously
attempting to restore a common subset of operational
parameters.
– This centralization of controls system provides many
benefits in terms of coordination, conflict avoidance,
security, and optimization.
Functional Model (cont’)
• Front-end Tier:
– Provides access to the field I/O and underlying
dedicated fast feedback systems.
– This tier is configured and managed by the services
tier, but can run autonomously.
– For example, the services tier may configure a
feedback loop in the front-end tier, but the loop itself
runs without direct involvement.
– The primary abstraction in this tier is a channel, or
process variable, roughly equivalent to a single I/O
point.
Physical Model (global layer)
Physical Model (front-end field I/O)
Physical Model (cont’)
• The ILC control system must reliably interact with
more than 100,000 technical system devices that
could collectively amount to several million scalar and
vector Process Variables (PVs) distributed across the
many kilometers of beam lines and facilities at the
ILC site.
• Information must be processed and distributed on a
variety of timescales from microseconds to several
seconds.
• The overall philosophy is to develop an architecture
that can meet the requirements, while leveraging the
cost savings and rapid evolutionary advancements of
commercial off-the-shelf (COTS) components.
Network Infrastructure
• Data collection, issuing and acting on setpoints, and pulseto pulse feedback algorithms are all synchronized to the
pulse repetition rate.
• The controls network must be designed to ensure adequate
response and determinism to support this pulse-to-pulse
synchronous operation, which in turn requires prescribing
compliance criteria for any device attached to this network.
• Additionally, large data sources must be prudently managed
to avoid network saturation.
• Dedicated compute nodes associated with each backbone
network switch service for monitoring, data reduction, and
implementing feedback algorithms.
Network Infrastructure (cont’)
• BCD Control Room Cluster Architecture
ILC test facility and collaboration
• Several accelerator facilities are used as the part of the
ILC test facilities in the wide world:
– LLRF: Fermilab, DESY, KEK, SNS, LBNL, U.Penn and others
– ATCA: DESY is developing a version of their LLRF Simcon
board on ATCA, and several other institutions worldwide are
beginning to explore ATCA, such as SLAC.
– Beam Instrumentation: Fermilab, SLAC, KEK, DESY,
U.Oxford, U. London and others.
• Goal:
– The ILC Control work is highly collaborative, and several
work packages overlap between institutions.
– Researching and Working on the facility to assess the
cost and solve the key technique of ILC Controls.
IHEP Activity
• ATCA (Advanced Telecom Computing
Architecture) is chosen as the Electronics
platform of ILC controls.
– Unique open standard designed specifically for
0.99999 (5-9’s) availability at the crate level
– Core components available from industry
• Crates with intelligent platform management of power, module
type & ID, load shedding & re-routing
• N+1 redundancy options for core controllers, communication,
power, cooling fans
• All serial multi-gigabit communications by wire for short distance
or fiber for long distance
• Controllers, switches and high performance processors
– Ideal for core of controls system
IHEP Activity (cont’)
IHEP Activity (cont’)
•
•
•
Now, IHEP Control Group get budget support.
Plan to set up a prototype of ATCA at the lab.
Do some work at ATCA platform as follows:
– Study performance of ATCA including shelfmanager, redundancy switcher and power supply,
etc.
– Install the EPICS into ATCA system.
– Research and test the EPICS HA at ATCA.
– etc.
IHEP Activity (cont’)
Research field:




ATCA HA
Xen technology
Linux HA
EPICS HA
Shelf Management
Dual Xeon processors
Fig. the prototype of the ATCA
Switch
IHEP Activity (cont’)
Xen Technology
• XenSource company
• A open source virtual machine monitor (VMM), or
hypervisor, for the x86 processor architecture
• Can securely execute multiple virtual machines on a
single physical system with close-to-native performance
• Live migration of running virtual machines between
physical hosts
• Xend: a daemon responsible for managing virtual
machines and providing access to their consoles
IHEP Activity (cont’)
Linux HA
• Putting together a group of computers that trust each
other to provide a service even when system components
fail
• When one machine goes down, others take over its work
• This involves IP address takeover, service takeover, etc.
• New work comes to the “takeover” machine
• Not primarily designed for high performance
• It cannot achieve 100% availability – nothing can
• HA Clustering designed to recover from single faults
IHEP Activity (cont’)
• EPICS HA can be completed via two methods: one is Xen
technology and the other is Linux HA.
Fig1 EPICS HA structure via Xen
Fig2 EPICS HA structure via Linux HA
IHEP Activity (cont’)
Go on:
• Study EPICS HA via HA middleware: OpenClovis or
Goahead.
Fig3 EPICS HA structure via OpenClovis or Goahead
IHEP Activity (cont’)
Conclusion:
• Accumulate more experience based on the
prototype of the ATCA system.
• Hope researching results of ATCA be used into
ILC test facilities as soon as possible and
contribute to the assessment of ILC control
system.
• Improve and strengthen the close cooperation
with the other ILC control teams in the world.
Thanks for your attention!
Some information of this report is referred to ILC control kick-off
meeting, Aug 20~22,2007.