Output Trace Files
The simulator can write two CSV outputs for each run: a simulated-job schedule
and a resource-usage trace. Both currently write numeric timestamps regardless
of --timestamp_format. By default they are truncated to whole seconds;
--msec_output writes three decimal places. See
Command-Line Options.
Simulated-job schedule
Use --outfile to choose the simulated-job schedule file:
${CMAKE_INSTALL_PREFIX}/bin/simulator input.csv --outfile results/jobs_sim.csv
If --outfile is omitted, DR_EVT derives a filename from the input trace
(for example, jobs.csv becomes jobs_sim.csv). Normally, the CSV contains
one row for each scheduled job. When --max_time is set, it contains only
jobs completed at or before that inclusive boundary; jobs still running or
waiting are omitted. In the default ID-input build, an input trace that
explicitly provides q_id retains that column:
job_submit_time,begin_time,end_time,num_nodes,exit_status,q_id,time_limit
When the selected queue field is omitted from input, it is omitted from output too:
job_submit_time,begin_time,end_time,num_nodes,exit_status,time_limit
begin_time and end_time are the schedule produced by simulation.
exit_status is written as 0. Jobs rejected before scheduling are not
written to this output. With
-DDR_EVT_LEGACY_QUEUE_INPUT=ON, the equivalent selected input/output column
is the legacy named queue field instead of q_id.
TODO — user-defined queue names: Preserve user-defined input queue names instead of emitting only DR_EVT’s built-in canonical names.
Resource-usage trace
Use --resource_trace to name the resource-usage file explicitly:
${CMAKE_INSTALL_PREFIX}/bin/simulator input.csv \
--outfile results/jobs_sim.csv \
--resource_trace results/resources.csv
The resource trace records the occupancy after each resource-state change.
With --max_time, no event after the inclusive boundary is recorded; if no
resource-state change occurs exactly at the boundary, no synthetic final row
is added there.
With the default --trace_type standard, the format is:
time,free_nodes,allocated_nodes
0,100,0
0,20,80
10,5,95
40,20,80
100,100,0
With --trace_type pcon, the resource trace additionally reports the
quantities carried by the experimental power-usage trace model:
time,free_nodes,allocated_nodes,avgpcon,minpcon,maxpcon
--trace_type selects the job/resource data model independently of
--trace_format, which selects how the input file is parsed.
Without --resource_trace, DR_EVT writes this output as
<outfile>_resources.csv (for example, jobs_sim_resources.csv).
Replay outputs
The tracer uses the recorded begin_time and end_time values from its
input schedule to reconstruct the resource-usage trace described above when
--resource_trace is requested:
${CMAKE_INSTALL_PREFIX}/bin/tracer \
--infile schedule.csv \
--total_nodes 100 \
--resource_trace replay-resources.csv
The following analysis reports are separate from the resource trace:
--outfileoptionally writes a CSV row for every job, including its wait time, execution time, and the busy-node count at submission. No per-job report is written when this option is omitted.--subfilewrites the submission count and batch-resource availability for each hourly slot in the replay period.--subsumfaggregates submission and availability statistics for each of the 168 hour-of-week slots across the replayed weeks.--datfileoptionally writes detected dedicated-allocation-time (DAT) periods as a text report.
The per-job and submission reports are CSV files. Each report is disabled when
its option is omitted. When neither submission report is requested, tracer
skips the submission-statistics pass entirely.
CLI summary
At the end of a run, simulator prints the resolved configuration, job counts,
current simulation time, node count, average wait and turnaround times,
makespan, average and peak queue lengths, and process wall-clock time.
Wait time: submission to start.
Turnaround time: submission to completion.
Makespan: first submission to last completion.
Average queue length: number of jobs already waiting at each submission, averaged across submitted jobs. Running jobs are not counted.
Peak queue length: largest number of waiting jobs during the run.
Buffer sizing and output-flush controls are documented in Command-Line Options.
See also
Input Trace Files - accepted input formats and columns
Command-Line Options -
--outfile,--resource_trace,--job_flush_interval, and--resource_history_capacityFugaku Power-Usage Experiment - Plotting node and power-usage resource curves