VEGAS Pulsar Project Meeting - Sprint Review: 2017 April 17th 2:00 - 3:00pm ET

Present

  • Jason, Joe, Richard, Randy, Ryan

Sprint Story Overview

  • VE-214: Continue work on spectral line modes in VPM branch
  • VE-180: HBW accumulator refactoring
  • VE-216: Continue work on Matrix book keeping
  • VE-215: Spectral line regression testing (INCOMPLETE)
  • VE-217: Spectral line FITS writer for Matrix
  • VE-218: Matrix Manager interface
  • VE-219: Fix bug in Matrix accumulator
  • VE-220: 32-channel BOF files
  • VE-221: Update existing BOFs with delay register and wrap behavior improvements
  • VE-222: Upgrade post-processing scripts to handle multi-bank data with overlapping frequency coverage (IN PROGRESS)
  • VE-223: Collect all observing tools/scripts into a single softwrae area (COMPLETE)
  • VE-224: High time resolution modes
  • VE-225: Clone /home/sim for regression test environment (COMPLETE)

Story Details

VE-214: Continue work on spectral line modes in VPM branch

  • Ray not present. Via Joe: Ray has spectral line modes running in development environment. Not clear what the status is on regression test environment.

VE-180: HBW accumulator refactoring

  • Able to acquire HBW FITS files. Header data is as expected.

VE-216: Continue work on Matrix book keeping

  • Paused to work on manager interface, but LBW modes seem to be in good shape.

VE-217: Spectral line FITS writer for Matrix

  • Looks to be in good shape.

VE-218: Matrix Manager interface

  • Interface is hacked in at this point but the interface between the current manager and VEGAS system uses shared memory, FITS files, and FIFO's. Need to bind these into one logical interface so that it is possible to operate in old Manager interface or Matrix keymaster interface automatically. Will enable side-by-side comparison.

VE-219: Fix bug in Matrix accumulator

  • Joe does not remember what this bug was but we assume it was fixed smile

VE-220: 32-channel BOF files

  • Running simulations to validate PFB and FFT as well as transpose to reorder FFTs. All 32 channels have been checked and look good. Also tested new vector accumulator which also looks good. Have some BOF files that they want to verify but they are having problems with the shared memory on dev-hpc1 (formerly srbs-hpc1). Script has a problem connecting to shared memory (complains that no buffer exists with the expected ID). Randy will talk with Jason and Ray and look into this.

VE-221: Update existing BOFs with delay register and wrap behavior improvements

  • Finished on 512 channel modes for both coherent and incoherent 1500 and 800 MHz BW modes. Rest are waiting.

VE-222: Upgrade post-processing scripts to handle multi-bank data with overlapping frequency coverage

  • Ryan is running into some roadblocks related to MPI (which handles multi-processor tasks). If anyone has experience with MPI and can help, that would be appreciated.

VE-223: Collect all observing tools/scripts into a single softwrae area

  • Just need to decide on an installation area.

VE-224: High time resolution modes

  • Natalia asks if the file format for the VEGAS raw modes will be the same as GUPPI raw modes. Answer is yes.
  • Also asks if test data taken on B1937+21 has been moved. Ryan is not aware of this but will look into it.

VE-225: Clone /home/sim for regression test environment

  • This area is now called /home/reg_test. Ryan tried taking some data using the broad band noise source. Astrid and CLEO worked well but there were two immediate issues
    • There are no ADC calibration files for this ROACH (vegasr2-10). These files need to be made and placed in /home/reg_test/etc/config
    • SDFITS is not running in online mode (Ryan is not sure if this is typical behavior for the simulator). After running offline, the resulting data seems reasonable at first glance, though. * This is typical behavior.

VE-215: Spectral line regression testing

  • Waiting on release candidate

VEGAS Pulsar Project Meeting - Sprint Planning: 2017 April 3rd 2:00 - 3:00pm ET

Note: The team's "velocity" has been consistent at 11--13 points per sprint. We should aim for 12 points per sprint for planning purposes.

Backlog

  • Continue work on acquiring hardware for upgraded development environment
    • Two potential HPC's could not recognize the memory expansions. The 4 machines we hoped would work (north, south, east, and west) don't seem to have enough memory. So still need to identify second HPC.
  • Prepare a release candidate for the re-based VPM branch that can run spectral line modes as expected
  • Spectral line regression tests of release candidate
  • Continue work on high time resolution modes
  • Trouble shoot guppi_daq shared memory problems on dev-hpc1
  • Post processing scripts for multi-bank data

Critical Issues

  • M&C system and regression tests
  • Manager parameters and interface with CLEO

New Stories

  • ADC calibration for vegasr2-10
  • Manager parameters and interface with CLEO
    • Ron met with software team to discuss strategies for interfacing new managers with CLEO. For VEGAS, the relevant parameter is 'mode'. A decision was made to implement this as an enum (instead of a string) and to use an array of enums for the Coordinator's mode parameter. This can be done dynamically from looking at vegas.conf. This is in Ray's domain so we will need to talk with him about this.
      • Ryan: Can we also use an array for the 'parfile' parameter? This needs to be a string and stores a path to a file on disk that will not always be the same length. Ryan will talk with Ray and Joe about this offline.
    • Need to modify VPM (and Breakthrough Listen) managers to handle 'mode' in this way.
      • Do we also need parameters for different types of observing (spectral line, coherent pulsar, incoherent pulsar, Breakthrough Listen)?
      • Unclear. Ryan will talk with Ron about this.
  • Continue work on spectral line modes in VPM branch
  • Regression testing if possible
  • High time resolution
  • Matrix work: Interface between matrix and Ray's version of VEGAS manager up and running, as well as LBW spectral line modes
  • Shared memory bug on dev-test
-- RyanLynch - 2017-04-17
Topic revision: r2 - 2017-04-17, RyanLynch
This site is powered by FoswikiCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding NRAO Public Wiki? Send feedback