Click to edit Master text styles

1
(No Transcript)
2
DAQ Software Group
Icecube Collaboration Meeting Uppsala Oct. 6-11
Click to edit Master text styles Second
level Third level Fourth level Fifth level

• Members working at LBNL, UW (mostly PSL), PSU,
  Bartol, and Mainz.
• New additions since last collaboration meeting
• Keith Beatie (LBNL)
• Dan Wharton (UW)
• Pat Toale (PSU)
• Primary responsibilies
• DOM MB testing software
• DOM communications software
• FAT test software (TestDAQ and DOMtest)
• Spole DAQ system design and implementation

3
DOM Communications Software Improvements

• New software version of comm. protocol with CRC
  32 and error correction/retransmission.
• Production testing software for DOR rev. 0 cards.
• Used in testing of initial production run (50)
  and recent production run of 100 cards.
• Major rewrite of comm. protocol and firmware test
  suite.
• Firmware/driver revision validation tool.
• Will be used for system/cable tests _at_PSL
  (Oct./Nov.)
• Will be part of post deployment testing procedure
  at spole.

Icecube Collaboration Meeting Uppsala Oct. 6-11
Click to edit Master text styles Second
level Third level Fourth level Fifth level
4
Design Verification and FAT Test

• Integration of TestDAQ software, DOMapp through
  data collector, at PSL.
• Development of system level testing tools for
  individual TestDAQ components at important system
  interfaces (e.g. DOMhub output/data collector
  input).
• DOMtest analysis working. Ongoing development
  and testing to meet FAT test goals.

Icecube Collaboration Meeting Uppsala Oct. 6-11
Click to edit Master text styles Second
level Third level Fourth level Fifth level
5
Important Lessons (so far..)

• Test, test, test.then test again (or old code
  is good code).
• Subtle problems only appear with time.
• highly dynamic hardware availability for
  testing has been a BIG problem.
• In depth testing requires resources (e.g. DOMs)
  to be continuously available.
• First users are extension of test team.
• FAT users as beta testers.
• Constant communications with end user critical.

Icecube Collaboration Meeting Uppsala Oct. 6-11
Click to edit Master text styles Second
level Third level Fourth level Fifth level
6
Important Lessons (contd.)

• Hardware based testing situation improving
  hardware availability improving.
• Graveyard DOM MBs usable for much of DAQ
  testing.
• Greater availability of rev. 0 DOR cards to
  software group.
• Full DOM Hub (30 DOM MB, 8 DOR cards) now
  installed in software lab.

Icecube Collaboration Meeting Uppsala Oct. 6-11
Click to edit Master text styles Second
level Third level Fourth level Fifth level
7
Reviews and Workshops

• IceCube DAQ software internal review (May, LBNL).
• Some recommendatons
• Increased focus on spole DAQ implementation
  (transition from TestDAQ).
• Develop specific goals for first year
  functionality.
• Dedicated integration and testing programming
  effort.
• DAQ workshops held at PSU and LBNL.

Icecube Collaboration Meeting Uppsala Oct. 6-11
Click to edit Master text styles Second
level Third level Fourth level Fifth level
8
System Design Strategy

• Functional Description
• Data Flow Req. and Characteristics
• Optimizations and Implementation Goals
• System Architecture

9
Functional Mapping DOM, Hub, StringProcessor
10
Functional MappingSub detector and global
triggers
11
Functional MappingEvent builder
12
Online
Global Trigger
AMANDA
Experiment Control
IceCube DAQ Architecture
...
DAQ Control
InIce Trigger
IceTop Trigger
...
String Processor
...
DOM Hub Application
DOR Driver
HAL
...
DOMApp
13
DOM firmware status
Icecube Collaboration Meeting Uppsala Oct. 6-11
Click to edit Master text styles Second
level Third level Fourth level Fifth level

• All tests to date (FAT) done with design called
  testFPGA.
• Primarily used for DOM hardware verification.
• Behavior of local coinc. aimed at cable and
  circuitry testing not ATWD triggering.
• ATWD readout performed under program control.
  Can be triggered, but does not run continuously.
• Most other low level functions similar to those
  to be found in final FPGA design.

14
DOM firmware status (cont.)
Icecube Collaboration Meeting Uppsala Oct. 6-11
Click to edit Master text styles Second
level Third level Fourth level Fifth level

• Real FPGA design mostly complete.
• ATWD will trigger autonomously (selectable
  modes).
• Data will flow into circular buffer under FPGA
  control.
• Local coinc. will need specification and testing.
• Most other functions identical to test FPGA
• Initial version in test..release to software late
  Nov.-early Dec.

15
DOMapp software plans - this season
Icecube Collaboration Meeting Uppsala Oct. 6-11
Click to edit Master text styles Second
level Third level Fourth level Fifth level

• Availability of real FPGA not certain for start
  of this season purdent to plan on using test
  FPGA.
• Have implemented minimum deadtime continuous
  ATWD readout in software.
• Have implemented/tested software version of FPGA
  data compression algorithm.
• Measured performance currently 1.4KHz. for SPE
  events. About 1KHz. for pedestal events.
• Have requested simple nearest neighbor LC in
  FPGA for this season.will allow full readout
  rates.

16
DAQ software component design
Icecube Collaboration Meeting Uppsala Oct. 6-11
Click to edit Master text styles Second
level Third level Fourth level Fifth level

• DAQ components (DOMhub, String Processor, Inice
  and Icetop Trigger, Global Trigger, Event
  Builder, DAQ Control.
• All DAQ components share common code for
• Configuration access and storage (EJB interface
  to SQL database)
• Control (state machines)
• Common use of Splicer classes
• Common mechanism for exposing monitoring data
  (JMX beans)
• Common mechanism for creating and discovering
  connections to other components
  (connectionLocator classes)

17
DAQ Component States
18
Splicer Pattern Class
19
Schematic view of a DAQ component.
20
DAQ software component status
Icecube Collaboration Meeting Uppsala Oct. 6-11
Click to edit Master text styles Second
level Third level Fourth level Fifth level

• Recent meetings at PSU and LBNL have defined
  interface objects.
• Execution environment at PSL and spole well
  defined, documented and implemented.
• DAQ templates programs in release ver. 1. Initial
  implementation of all component templates
  complete.
• Work on individual components well underway.
  Integration tests early Nov.

21
String Processor Status
Icecube Collaboration Meeting Uppsala Oct. 6-11
Click to edit Master text styles Second
level Third level Fourth level Fifth level

• Time correction algorithm tested.
• String processor trigger functionality defined
  for next seasonsimple parameter driven LC.
• Message formats for communication with inice
  trigger and event builder defined.
• Implementation of above within DAQ templates in
  progress.

22
InIce Trigger Summary

• Trigger Algorithms are defined and understood
• Simple Majority Trigger is implemented in Splicer
  Architecture
• Interface implementation nearly complete
• Management of multiple algorithms is being worked
  out

23
IceTop Trigger Detail
Yr-1
DAQ control
Global Trigger
splice
sorter
Joiner
splice
24
IceTop Message Payload Formats

• CLASS Structure - extensions of Payload Class
• Payload
• UTCTime getPayloadTimeUTC( )
• ID_type getDomID( )
• IDHPayload ..from IDH to
  Icetop Trigger
• UTCTime getPayloadTimeUTC( )
• ID_type getDomID( )
• byte getHitType( )
• IcetopTriggerPayload from Icetop Trigger
  to Global Trigger
• Note UTCTime getPayloadTimeUTC( )
  getPayloadStartTimeUTC( )
• UTCTime getPayloadStartTimeUTC( )
• UTCTime getPayloadStopTimeUTC( )
• byte getTriggerType( )
• Vector HitInfo would contain all the
  information for the hits generating the trigger.
  This includes payloadTime , DomID and Hit Type
• Vector EB instructions

25
Global Trigger First Year Goals

• 1. GT will use only InIce and IceTop trigger
  inputs.
• 2. GT will apply extended time window to issue a
• GT event size of the extended time window
  is
• determined by trigger types from InIce and
  IceTop
• trigger.
• 3. GT will instruct EB to read both InIce and
  IceTop
• even though only InIce or IceTop trigger
  exists
• exception pure calibration
  trigger
• 4. GT will generate random trigger for
  monitoring
• TBD random trigger frequency
• TBD random trigger time window

26
Event Builder Status

• Event builder? string processor messaging
  defined.
• Event builder ? DAQ dispatch interface defined.
• Output data format being documented.
• Implementation in progress.

27
Integration Plans

• Initial development at individual sites.
• First integration on DAQ testbed at LBNL
• Pre-pole integration on test system at PSL