sPHENIX discussion of software

["Osborn, Joseph" <josborn1 AT bnl.gov>]

Hi all,

 

Here are notes from the discussion today. We will reconvene on Friday at 10 AM ET to discuss progress and follow ups – the indico page is below and a reminder will be sent in a few days.

Needs for event building/pooling

Step 1 - concatenate data stream within each PRDF to recover original data stream from PCIe. This is done packet by packet, sector by sector. Right now there is a jitter of up to 3 BCOs resulting in lost hits and/or event mixing. Martin is working on this, will get a report on Friday.
Step 2 -Synchronize between EBDCs, which time aligns them with a BCO. A working prototype exists for the INTT now, which is the “guinea pig.” The next step will be for the TPC, which is more complex due to sensitivity to many BCOs. Chris is working on this, will get a report on Friday.

 

Event building/pooling results in concatenated, time aligned PRDF (calos) and DST (tracking detectors) containing raw data packets organized by event. These will be read by F4All as input to raw data decoders/unpackers.

MVTX

Raw data decoder (Martin's definition) is still missing, which decodes packets. Yasser is working on it, plan to have it done this week. A data unpacker exists in coresoftware which fills the hitsets. This will have to be updated to reflect the event pooling, e.g. as in the case for the INTT. But the framework exists.

 

INTT

Raw data decoder and data unpacker exist. Event pooling prototype exists as of today. Joe O is checking/confirming that the filling of hitsetkeys looks correct in the unpacker. Joseph B. is going to check the entire workflow to confirm that the output BCOs look the same regardless of using the F4A version or the private version which has been used for INTT data analysis already. Expect a report from Joseph on Friday.

 

TPC/TPOT

Raw data decoder and unpacker exist, as demonstrated by e.g. Christof's privately made DST containing hit sets. What is required is the event pooling from Chris to time synchronize files from different EBDCs to avoid manually building events from the same BCOs.

 

Offline tracking team will be responsible for confirming the filling of hitsets is performed properly in the data unpackers.

 

Small committee met at 2 PM ET today to discuss the hitset, header, and specific needs of each subsystem's unpacker into the offline reconstruction framework. Distributed tasks to develop strawman subsystem specific event header IO objects for additional subsystem needs (e.g. global BCO for each FELIX or PCIe).

 

Meeting at 10 AM ET Friday - Indico page: https://indico.bnl.gov/event/20626/, reminder email with info will be distributed in a few days.

 

Reports to expect on Friday

Martin - Update on event mixing/hit dropping

Chris - Update on TPC time/file synchronization

Joseph B - Update on prototype INTT pooling and hitset filling - are the BCOs the same regardless of using F4A or private INTT code?

Yasser - update on MVTX raw decoder

Offline tracking group - Update on unpacker to hitset interface.

 

___________________________

Joe Osborn, Ph.D

Physics Department

Brookhaven National Laboratory

josborn1 AT bnl.gov

 

 

From: sPHENIX-software-l <sphenix-software-l-bounces AT lists.bnl.gov> on behalf of Anthony Frawley via sPHENIX-software-l <sphenix-software-l AT lists.bnl.gov>
Date: Monday, September 18, 2023 at 4:25 PM
To: sphenix-tracking-l AT lists.bnl.gov <sphenix-tracking-l AT lists.bnl.gov>, sphenix-software-l AT lists.bnl.gov <sphenix-software-l AT lists.bnl.gov>, sphenix-physics-l AT lists.bnl.gov <sphenix-physics-l AT lists.bnl.gov>
Subject: Re: [Sphenix-software-l] Online/offline meeting tomorrow, Tuesday September 19 at 10 am EDT

Hi All,

 

Jin has arranged for the 1008 conference room to be available for locals at BNL.

 

Best regards

Tony

---

From: sPHENIX-physics-l <sphenix-physics-l-bounces AT lists.bnl.gov> on behalf of Anthony Frawley <afrawley AT fsu.edu>
Sent: Monday, September 18, 2023 3:17 PM
To: sphenix-tracking-l AT lists.bnl.gov <sphenix-tracking-l AT lists.bnl.gov>; sphenix-software-l AT lists.bnl.gov <sphenix-software-l AT lists.bnl.gov>; sphenix-physics-l AT lists.bnl.gov <sphenix-physics-l AT lists.bnl.gov>
Subject: [Sphenix-physics-l] Online/offline meeting tomorrow, Tuesday September 19 at 10 am EDT

 

Hi All,

 

The first of our meetings to address getting data from the online world to the offline reconstruction world will be tomorrow, September 19, from 10 am - 1 :00 pm EDT (a second meeting is planned for Friday of this week at 10 am).

 

Let's use this zoom link:

https://fsu.zoom.us/j/91902088638

 

The Indico page is:

https://indico.bnl.gov/event/20614/

 

The goal is to bring together the experts from the DAQ, detector readout and offline reconstruction worlds, so that we can establish what remains to be done before we can run large scale reconstruction of experimental data. 

 

The outcome of the first meeting should be a common understanding of what needs to be done, who will do it, who can/will help them, and a timeline for completion. We should start tomorrow by discussing the most important tasks to be completed.

 

I will try to make good notes of the discussion as we go.

 

Best regards

Tony