Draft: SLink / ELink Validation Plugins : Analyzing Ph2 IT Hardware Input Data Size

[ldishman]

This work integrates Validation plugins for both SLink and ELink occupancy rates, with configurations and harvesters for plotting and analysis. It is still under review at this time.

PR currently contains code from Si Hyun's unmerged unpacker PR #25. This will be removed (e.g. by rebasing) once his PR is merged.

The SLink physics outputs include:

• TH1F: Full spectrum (each event's entry added individually, then all entries divided by nEvents at the end) SLink occupancy ([data size x trigger rate] / [slink bandwidth]) per-event across all detector areas
• TH1F: Event-averaged SLink occupancy (accumulated over all events and then averaged)
• 36 TH1Fs: Full spectrum SLink occupancy per-event for each individual DTC
• 36 TProfiles: Mean SLink occupancy per DTC, one histo for each DTC (one bar per SLink on each DTC's plot)
• 2DTProfile: Mean SLink occupancy mapped in 2D SLink vs. DTC space (color bar for each "pixel" shows occupancy)
• TH2F: Full spectrum SLink occupancy across all DTCs (color bar for each "pixel" shows per-event SLink entries)
• TProfile: Mean SLink occupancy by DTC, averaged across all DTCs (one bar per DTC on one plot)

The ELink physics outputs include:

• TH1F: Full spectrum (each event's entry added individually, then all entries divided by nEvents at the end) ELink occupancy ([data size x trigger rate] / [elink bandwidth], x sectionScale) per-event across all detector areas
• TH1F: Event-averaged ELink occupancy (accumulated over all events and then averaged)
• 13 TH1Fs: Full spectrum ELink occupancy per-event for individual sections (TBPX_L1–4, TFPX_R1–4, TEPX_R1–5)
• 12 TH1Fs: Full spectrum ELink occupancy per-event for individual subtypes (one per subtype ID present)
• TH2F: Full spectrum ELink occupancy across all sections (color bar for each "pixel" shows per-event ELink entries)
• TH2F: Full spectrum ELink occupancy across all subtypes (color bar for each "pixel" shows per-event ELink entries)
• TProfile: Mean ELink occupancy by section, averaged across all sections (one bar per section on one plot)
• TProfile: Mean ELink occupancy by subtype, averaged across all subtypes (one bar per subtype on one plot)

[tomalin]

The title and description of this PR say that it's only about S-Links data rates, but your talk https://indico.cern.ch/event/1701229/#11-boston says you'll add E-Link data rates to this PR to, so please update title & description.

I suggest also adding temporary text to end of description saying:

1. E-Link data rates currently missing. PR will remain draft until added.
2. PR currently contains code from one of Si-Hyun's unmerged unpacker PRs. This will be removed (e.g. by rebasing) once his PR is merged. Please say which of his PRs this is.

[ljuckett]

If these plots are going to be used in the release validation process, they will eventually have to be added to the standard Validation sequence. This will either be in Validation/SiTrackerPhase2V/python/Phase2TrackerMCHarvesting_cff.py or Validation/SiTrackerPhase2V/python/Phase2TrackerValidationFirstStep_cff.py depending on when your code should run. You can then test that it appears when you run DQM/SiTrackerPhase2/test/dqmstep_phase2tk_cfg.py.
Also, please add "V" to the top folders in your validation plugins. This is how we differenciate the DQM and Validation outputs.

Nice work :-)