Project

General

Profile

Feature #20388

Create one or more simple_test_configs that are as similar to protoDUNE DAQ that we can get and stress test them on the Mu2e Pilot cluster

Added by Kurt Biery about 1 year ago. Updated about 1 year ago.

Status:
Work in progress
Priority:
High
Assignee:
-
Category:
-
Target version:
-
Start date:
07/21/2018
Due date:
% Done:

0%

Estimated time:
Experiment:
-
Co-Assignees:
Duration:

Description

It should have as many BoardReaders as is feasible, 8-16 EventBuilders, a Dispatcher, and online monitoring.

It should use as many Mu2edaq cluster nodes as is feasible, ideally 4-6 or more.

It should include the RoutingMaster.

It should support 100MB+ events at tens of Hz.

Stress tests should include:
  • running for days
  • starting and stopping runs without reconfiguring, for days
  • times of extreme backpressure
  • instances of BoardReaders dropping out
  • killing (or sigstop-ing) a BoardReader process and seeing that the system continues to run, and Stops (end-run) cleanly
  • killing and/or stopping an EventBuilder process and seeing that the system continues to run, and Stops (end-run) cleanly

History

#1 Updated by Kurt Biery about 1 year ago

  • Subject changed from Create a simple_test_config that is as similar to protoDUNE DAQ that we can get and stress test it on the Mu2e Pilot cluster to Create one or more simple_test_configs that are as similar to protoDUNE DAQ that we can get and stress test them on the Mu2e Pilot cluster

#2 Updated by Kurt Biery about 1 year ago

  • Tracker changed from Bug to Feature

#3 Updated by Kurt Biery about 1 year ago

  • Priority changed from Normal to High

#4 Updated by Ron Rechenmacher about 1 year ago

At this point the simple_test_config/mediumsystem_with_routing_master is the primary "simple config"

#5 Updated by Kurt Biery about 1 year ago

  • Status changed from New to Work in progress

branch in artdaq_daqinterface called mu2edaq_stresstest

the simple_config name is protodune_mock_system.
24 components ( 1 in push, 4kfragments; even numbered components 1 MB, rest 15 MB fragments)
16 EBs
Routing Master
25 Hz
total rate of 5000 MB/s ; will saturate 4 EB nodes
no disk writing

without disk writing, there should be no back-pressure in this system.



Also available in: Atom PDF