VEGAS Pulsar Project Meeting - Sprint Review: 2017 November 20th 14:00 - 15:00pm ET

Present

  • Ryan, Natalia, Joe, Ray

Sprint Story Overview

  • VE-335: Discuss workaround for LBW Pulsar Matrix (IN PROGRESS)
  • VE-336: Troubleshoot high time resolution modes (IN PROGRESS)
  • VE-341: Test Matrix Classic for release (IN PROGRESS)
  • VE-339: Identify directory to install user tools (IN PROGRESS)
  • VE-348: Work on VEGAS Matrix documentation (IN PROGRESS)
  • VE-350: Test 32-channel modes (IN PROGRESS)
  • VE-349: Support testing (COMPLETE)
  • VE-351: Support Yunan testing (COMPLETE)
  • VE-345: Focus on deploying classic HBW pulsar and spectral line Matrix (IN PROGRESS)
  • VE-346: Bundle software for Yunan (COMPLETE)
  • VE-347: Change latency list (IN PROGRESS)
  • VE-331: Continue troubleshooting packet loss issue in c1500 modes (IN PROGRESS)
  • VE-322: Cannot validate scripts offline (INCOMPLETE)

Story Details

VE-335: Discuss workaround for LBW Pulsar Matrix

  • Ryan sent an email to digital asking about possible workarounds that would allow current guppi_daq to be used with LBW pulsar modes. Still waiting on response
  • Ryan will send a reminder email and meet face-to-face if necessary after the Thanksgiving holiday.

VE-336: Troubleshoot high time resolution modes

  • Natalia is doing a comparison of S/N in each bank with some old data.
  • Natalia analyzed some old data and was able to partially recover the noise diode signal. The levels were low in some banks and Ryan wants to get more data with better vegas.scale values.
  • Ryan is working on getting some more data in RAW mode for Natalia to analyze. He has been trying to get data in the 32 channel modes if possible. This has been slowed by other testing.

VE-341: Test Matrix Classic for release

  • Biggest issue is that managers do not conform new parameters when switching between modes if those managers were not already in use.
    • Switcover works when configuring manually
    • Joe and Ray are adding a conform in the config tool for VEGAS manager to emulate manual behavior
    • Ray is working on functions to program the 10-gig core directly instead of going through the borph server. This should prevent ARP storms from banks that are not in use and fix issues with MAC addresses
  • Hoping to have some of these in place for testing over Thanksgiving

VE-339: Identify directory to install user tools

  • Joe and Ray will create a repo for Ryan to use

VE-348: Work on VEGAS Matrix documentation

  • Joe did some work on this. Has a developers document started. Set aside to work on more pressing issues.

VE-350: Test 32-channel modes

  • Incoherent modes seem to be working
  • Coherent modes have a different packet size than higher channel modes
    • Joe sent Luke binary and decryped packet data
    • Joe verified that the counter field is present and incrementing properly
    • Luke sent a follow-up email identifiying which fields are missing
      • These fields could be filled with dummy values so that packets are the correct size

VE-349: Support testing

  • Joe and Ray have been supporting testing

VE-351: Support Yunan testing

  • Visit is over. Seems to have gone well.

VE-345: Focus on deploying classic HBW pulsar and spectral line Matrix

  • Testing is ongoing.

VE-346: Bundle software for Yunan

  • Done.

VE-347: Change latency list

  • Luke sent Ryan a preliminary summary. Ryan waiting on more information.

VE-331: Continue troubleshooting packet loss issue in c1500 modes

  • Joe thinks some of this will go away by deactivating ROACH's and getting ARP tables set properly.

VE-322: Cannot validate scripts offline

  • No update yet.

VEGAS Pulsar Project Meeting - Sprint Planning: 2017 October 23rd 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.

Ryan will test over Thanksgiving shutdown. Goal is to do a test of all spectral line modes with both total power and frequency switching, and all pulsar modes with various scalings. If Joe and Ray's upcoming work is ready, Ryan will also test handoff between different modes.
  1. Operator will leave L-Band in focus and Ryan in the gateway. Ryan will be online at shutdown to verify that he has proper permissions.
  2. Ray and/or Joe will switch software versions before 9:00 PM on Wednesday.
  3. Ryan will do overnight testing on Wednesay and Thursday morning. Will be done before noon Thursday.
  4. Ryan will send Ray and Joe and email to tell them when they are done, and they will switch back to release by 6:00 AM Friday morning.
  5. Ryan will be on Talk and Draw for the end of shutdown to make sure there are no lingering issues.

Backlog

  • High time resolution modes
  • Investigate packet loss in c1500 modes
  • Continue work on Matrix LBW pulsar modes

Critical Issues

  • Offline script validation
  • MAC addresses for 10-gig ethernet ports are wrong for Banks C and D in VPM coherent modes
  • Identify a source tree for VPM utilities
  • Install VPM utilities in new source tree
  • Add a Task Master task for vpm_auto_plot, automatic VPM bandpass monitor, and a DAQ server used to monitor processes on the VPM HPCs in coherent dedispersion modes.

New Stories

  • Testing overe Thanksgiving shutdown (see above)
-- RyanLynch - 2017-11-20
Topic revision: r1 - 2017-11-20, 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