1OSLAT(8)                    System Manager's Manual                   OSLAT(8)
2
3
4

NAME

6       oslat - OS Latency Detector
7

SYNOPSIS

9       oslat [ -shvz ] [ -b bucket-size ] [ -B bias ] [ -c cpu-list ] [ -C
10             cpu-main-thread ] [ -f rt-prio ] [ --json filename ] [ -m
11             workload-mem ] [-t runtime ] [ -T trace-threshold ] [ -w workload
12             ]
13

DESCRIPTION

15       oslat is an open source userspace polling mode stress program to detect
16       OS level latency.  The program runs a busy loop with no or various
17       workloads, collecting TSC information and measuring the time frequently
18       during the process.
19

OPTIONS

21       -b, --bucket-size=N
22              Specify the number of the buckets (4-1024).
23
24       -B, --bias
25              Add a bias to all the buckets using the estimated mininum.
26
27       -c, --cpu-list=CPULIST
28              Specify CPUs to run on.  For example, '1,3,5,7-15'.
29
30       -C, --cpu-main-thread=CORE
31              Specify which CPU the main thread runs on.  Default is cpu0.
32
33       -f, --rtprio=PRIORITY
34              Using specific SCHED_FIFO priority (1-99).  Otherwise use the
35              default priority, normally it will be SCHED_OTHER.
36
37       --json=FILENAME
38              Write final results into FILENAME, JSON formatted.
39
40       -m, --workload-mem=SIZE
41              Size of the memory to use for the workload (e.g., 4K, 1M).
42              Total memory usage will be this value multiplies 2*N, because
43              there will be src/dst buffers for each thread, and N is the
44              number of processors for testing.
45
46       -D, --duration=TIME
47              Specify test duration, e.g., 60, 20m, 2H (m/M: minutes, h/H:
48              hours, d/D: days).  By default the unit is s/second.
49
50       -T, --trace-threshold=THRESHOLD
51              Stop the test when threshold triggered (in USEC).  At the
52              meantime, print a marker in ftrace and stop ftrace too.
53
54       -w, --workload=WORKLOAD
55              Specify a kind of workload, default is no workload.  Options:
56              "no", "memmove".
57
58       -s, --single-preheat
59              Use a single thread when measuring latency at preheat stage
60              NOTE: please make sure the CPU frequency on all testing cores
61              are locked before using this parmater.  If you don't know how to
62              lock the freq then please don't use this parameter.
63
64       -h, --help
65              Show the help message.
66
67       -v, --version
68              Show the version of the program.
69
70       -z, --zero-omit
71              Don't display buckets in the output histogram if all zeros.
72

AUTHOR

74       oslat was written by Peter Xu <peterx@redhat.com>.
75
76
77
78                                August 17, 2020                       OSLAT(8)
Impressum