-
Notifications
You must be signed in to change notification settings - Fork 4
/
Copy pathperf-readme
52 lines (32 loc) · 1.79 KB
/
perf-readme
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
.\" Manpage for perf.sh.
.\" Contact sbandyop@redhat.com to correct errors or typos.
.TH per.sh MAN "13 May 2017" "1.0" "perf.sh man page"
.SH NAME
perf.sh \- Collects generic performance data from all subsystem (CPU, Memory, I/O )
.SH SYNOPSIS
perf.sh [OPTION] [Interval] [Iterations]
.SH DESCRIPTION
Collects generic performance data from all subsystem (CPU, Memory, I/O )
1. By default the script will run in interval of 2 seconds and 30 times.
2. This script can be stopped using ctrl+c and still will capture the data.
3. This will collect data like sar, vmstat, iostat, iotop, ps, top and create a tarball inside current directory which can be helpful finding out
reason of high CPU/Memory usage, high load average, high iowait.
4. This script is tested to work in RHEL/CentOS 5,6,7 and Fedora 23,24. You should have tools like systat, iotop installed for the script to work.
5. Tools such as collectl are available which does this better, however it was created keeping simplycity in mind.
6. It does not capture huge ammount of data like collectl and much easier to use.
.SH OPTIONS
-p : Collect perf data perf(1) , Not to be confused with name of this script - perf.sh
-d : Run in daemon mode, for this the script needs to be run using setsid.
This is useful when the performance issue is sporadic and can not be reproduced at will
$ setsid ./perf.sh -d 2> /dev/null &
To end the data collection kill the script with SIGNAL 15
$ pkill perf.sh
.SH SEE ALSO
sar(1) free(1) ps(1) iostat(1) iotop(8)
https://github.com/shatadru/SGPD
.SH BUGS
No known bugs. Report Bugs at : https://github.com/shatadru/SGPD/issues/new
.SH TODO
Collect network related data apart from stuffs collected by SAR
.SH AUTHOR
Shatadru Bandyopadhyay (sbandyop@redhat.com)