1bonnie++(8) System Manager's Manual bonnie++(8)
2
6 bonnie++ - program to test hard drive performance.
7
10 bonnie++ [-d dir] [-c concurrency] [-s
11 size(MiB)[:chunk-size(b)[:seeks[:seekers]]]] [-n num‐
12 ber-to-stat(*1024)[:max-size[:min-size][:num-directo‐
13 ries[:chunk-size]]]] [-m machine-name] [-r ram-size-in-MiB] [-x num‐
14 ber-of-tests] [-u uid-to-use:gid-to-use] [-g gid-to-use] [-q] [-f
15 size-for-char-io] [-b] [-D] [-p processes | -y p|s ] [-z seed-num|-Z
16 random-file]
17
20 This manual page documents briefly the bonnie++, program.
21 Bonnie++ is a program to test hard drives and file systems for perfor‐
23 mance or the lack therof. There are a many different types of file sys‐
24 tem operations which different applications use to different degrees.
25 Bonnie++ tests some of them and for each test gives a result of the
26 amount of work done per second and the percentage of CPU time this
27 took. For performance results higher numbers are better, for CPU usage
28 and latency lower are better (NB a configuration scoring a performance
29 result of 2000 and a CPU result of 90% is better in terms of CPU use
30 than a configuration delivering performance of 1000 and CPU usage of
31 60%).
32 There are two sections to the program's operations. The first is to
34 test the IO throughput in a fashion that is designed to simulate some
35 types of database applications. The second is to test creation, read‐
36 ing, and deleting many small files in a fashion similar to the usage
37 patterns of programs such as Squid or INN.
38 All the details of the tests performed by Bonnie++ are contained in the
40 file /usr/share/doc/bonnie++/readme.html
41
44 For Bonnie++ every option is of the form of a hyphen followed by a let‐
45 ter and then the next parameter contains the value.
46 -d the directory to use for the tests.
48 -c the level of concurrency (default 1). The number of copies of
50 the test to be performed at the same time.
51 -s the size of the file(s) for IO performance measures in
53 megabytes. If the size is greater than 1G then multiple files
54 will be used to store the data, and each file will be up to 1G
55 in size. This parameter may include the chunk size seperated
56 from the size by a colon. The chunk-size is measured in bytes
57 and must be a power of two from 256 to 1048576, the default is
58 8192. NB You can specify the size in giga-bytes or the
59 chunk-size in kilo-bytes if you add g or k to the end of the
60 number respectively. This parameter may also include the number
61 of seeks (default 8192) and the number of seeker processes
62 (default 5).
63 If the specified size is 0 then this test will be skipped.
65 -n the number of files for the file creation test. This is measured
67 in multiples of 1024 files. This is because no-one will want to
68 test less than 1024 files, and we need the extra space on
69 braille displays.
70 If the specified number is 0 then this test will be skipped.
72 The default for this test is to test with 0 byte files. To use
74 files of other sizes you can specify number:max:min:num-directo‐
75 ries:chunk-size where max is the maximum size and min is the
76 minimum size (both default to 0 if not specified). If minimum
77 and maximum sizes are specified then every file will have a ran‐
78 dom size from the range min..max inclusive. If you specify a
79 number of directories then the files will be evenly distributed
80 amoung that many sub-directories.
81 If max is -1 then hard links will be created instead of files.
83 If max is -2 then soft links will be created instead of files.
84 -m name of the machine - for display purposes only.
88 -r RAM size in megabytes. If you specify this the other parameters
91 will be checked to ensure they make sense for a machine of that
92 much RAM. You should not need to do this in general use as it
93 should be able to discover the RAM size. NB If you specify a
94 size of 0 then all checks will be disabled...
95 -x number of test runs. This is useful if you want to perform more
98 than one test. It will dump output continuously in CSV format
99 until either the number of tests have been completed, or it is
100 killed.
101 -u user-id to use. When running as root specify the UID to use for
104 the tests. It is not recommended to use root (since the occa‐
105 sion when a Bonnie++ bug wiped out someone's system), so if you
106 really want to run as root then use -u root. Also if you want
107 to specify the group to run as then use the user:group format.
108 If you specify a user by name but no group then the primary
109 group of that user will be chosen. If you specify a user by
110 number and no group then the group will be nogroup.
111 -g group-id to use. Same as using :group for the -u parameter,
114 just a different way to specify it for compatibility with other
115 programs.
116 -q quiet mode. If specified then some of the extra informational
119 messages will be suppressed. Also the csv data will be the only
120 output on standard out and the plain text data will be on stan‐
121 dard error. This means you can run bonnie++ -q >> file.csv to
122 record your csv data.
123 -f size-for-char-io
126 fast mode control, skips per-char IO tests if no parameter, oth‐
127 erwise specifies the size of the tests for per-char IO tests
128 (default 20M).
129 -b no write buffering. fsync() after every write.
132 -p number of processes to serve semaphores for. This is used to
135 create the semaphores for synchronising multiple Bonnie++ pro‐
136 cesses. All the processes which are told to use the semaphore
137 with -ys will start each test with synchronization. Use "-p -1"
138 to delete the semaphore.
139 -y s|p perform synchronization before each test. Option s for sema‐
142 phores and option p for prompting.
143 -D use direct IO (O_DIRECT) for the bulk IO tests
146 -z seed
149 random number seed to repeat the same test.
150 -Z random-file
153 file containing random data in network byte order.
154
156 Run the following commands to run three copies of Bonnie++ simultane‐
157 ously:
158 bonnie++ -p3
160 bonnie++ -y > out1 &
162 bonnie++ -y > out2 &
164 bonnie++ -y > out3 &
166
169 The primary output is plain-text in 80 columns which is designed to fit
170 well when pasted into email and which will work well with Braille dis‐
171 plays.
172 The second type of output is CSV (Comma Seperated Values). This can
174 easily be imported into any spread-sheet or database program. Also I
175 have included the programs bon_csv2html and bon_csv2txt to convert CSV
176 data to HTML and plain-ascii respectively.
177 The "Name:Size etc" field has the name, filesize, IO chunk size, con‐
179 currency, number of seeks, and number of seek processes separated by :
180 characters.
181 For every test two numbers are reported, the amount of work done
183 (higher numbers are better) and the percentage of CPU time taken to
184 perform the work (lower numbers are better). If a test completes in
185 less than 500ms then the output will be displayed as "++++". This is
186 because such a test result can't be calculated accurately due to round‐
187 ing errors and I would rather display no result than a wrong result.
188 Data volumes for the 80 column text display use "K" for KiB (1024
190 bytes), "M" for MiB (1048576 bytes), and "G" for GiB (1073741824
191 bytes). So K/sec means a multiple of 1024 bytes per second.
192
195 This program, its manual page, and the Debian package were written by
196 Russell Coker <russell@coker.com.au>, parts of the program are based on
197 the work of Tim Bray <tbray@textuality.com>.
198 The documentation, the Perl scripts, and all the code for testing the
200 creation of thousands of files was written by Russell Coker, but the
201 entire package is under joint copyright with Tim Bray.
202
205 Handles SIGINT and does a cleanup (which may take some time), a second
206 SIGINT or a SIGQUIT will cause it to immediately die.
207 SIGXCPU and SIGXFSZ act like SIGINT.
209 Ignores SIGHUP.
211
215 The random file sizes will add up to different values for different
216 test runs. I plan to add some code that checks the sum and ensures
217 that the sum of the values will be the same on seperate runs.
218
221 The source is available from http://www.coker.com.au/bonnie++ .
222 See http://etbe.coker.com.au/category/benchmark for further informa‐
224 tion.
225
228 zcav(8), getc_putc(8), bon_csv2html(1), bon_csv2txt(1)
229 bonnie++(8)