Project

General

Profile

Special run requests


Michelle - 11/18/15 (pending on code fix by Nuno)
4 hour run with PTB only to check counter rates, disk write ON.


Michelle and Gabriel - 11/18/15
(pending setup by John F?)
Disk write on.
RCE readout only with 16 RCEs.
Trigger rate as low as possible (0.05 Hz is low enough, ok to try at 1 Hz)
- 3 "Pedestal" runs - 1 event each with trigger that activates 200 microslices.
- 1 "Pedestal" run - 5 events with trigger that activates 200 microslices
- 3 "Pedestal" runs - 1 event each with trigger that activates 400 microslices.
- 1 "Pedestal" run - 5 events with trigger that activates 400 microslices


Michelle - 11/17/15
(pending code change by Matt G. to make N a fcl parameter)
(First part finished 11/18, see below)
- Run in triggered mode, 1 Hz, with 4 RCEs only. Increase number of microslices
activated by each trigger (N) to N=40. No disk write. 30 minutes or until it crashes
- Repeat with disk write on.
Default value is currently N=10.


Michelle - 11/19/15
(No need to wait for PTB code fix from Nuno)
- Run with SSPs in external trigger mode. Readout SSPs and PTB, disk write on
- Run with SSPs in internal trigger mode. Readout SSPs and PTB, disk write on
Trigger frequency not important, looking to see how the SSP triggers are recorded in the PTB data stream
and get them into the online monitoring, need data files to play with.
5 minutes of data is sufficient.
------------------------------------------------------------------------------------------------

Completed

Requested: Michelle - 11/18/15
- Run in triggered mode, 1 Hz, all 16 RCEs. Increase number of microslices
activated by each trigger (N) to N=40. No disk write. 30 minutes or until it crashes
- Repeat with 4 RCEs only.
- Repeat with 4 RCEs and disk write on.
Default value is currently N=10.

Completed 11/19/15: Runs 4777-9, 2x,4x,3x data size respectively, 1Hz trigger rate, no-disk-write and all 16 RCEs
From M. Graham: 4x and 3x "The throughput seemed to max out at ~90MBps" and "significant backpressure, was dropping ADC payloads"
for 2x (run 4777) "Ran fine at 80 MBps though we didn’t really sustain 80 MBps for very long…I’d say 50MBps is more representative."

And also from Matt: "The links from boardReader PC->eventBuilder PC-> aggregator PC are all 1 Gbps; so the theoretical max there is 125MBps, but you never actually get that. The tests I did yesterday we got about 90MBps (720 Mbps) without disk writing. I think that’s not too bad…we can probably still do some artDAQ-level optimization, so possibly this can go up to 100MBps. In that case, we may be able to double by writing to two disks in parallel."