nghttp2/doc/h2load.1

471 lines
14 KiB
Groff
Raw Normal View History

.\" Man page generated from reStructuredText.
.
2019-03-07 15:22:06 +01:00
.TH "H2LOAD" "1" "Mar 07, 2019" "1.37.0" "nghttp2"
.SH NAME
h2load \- HTTP/2 benchmarking tool
.
.nr rst2man-indent-level 0
.
.de1 rstReportMargin
\\$1 \\n[an-margin]
level \\n[rst2man-indent-level]
level margin: \\n[rst2man-indent\\n[rst2man-indent-level]]
-
\\n[rst2man-indent0]
\\n[rst2man-indent1]
\\n[rst2man-indent2]
..
.de1 INDENT
.\" .rstReportMargin pre:
. RS \\$1
. nr rst2man-indent\\n[rst2man-indent-level] \\n[an-margin]
. nr rst2man-indent-level +1
.\" .rstReportMargin post:
..
.de UNINDENT
. RE
.\" indent \\n[an-margin]
.\" old: \\n[rst2man-indent\\n[rst2man-indent-level]]
.nr rst2man-indent-level -1
.\" new: \\n[rst2man-indent\\n[rst2man-indent-level]]
.in \\n[rst2man-indent\\n[rst2man-indent-level]]u
..
.SH SYNOPSIS
.sp
\fBh2load\fP [OPTIONS]... [URI]...
.SH DESCRIPTION
.sp
2017-12-17 06:16:53 +01:00
benchmarking tool for HTTP/2 server
.INDENT 0.0
.TP
.B <URI>
2015-01-15 16:10:16 +01:00
Specify URI to access. Multiple URIs can be specified.
URIs are used in this order for each client. All URIs
are used, then first URI is used and then 2nd URI, and
so on. The scheme, host and port in the subsequent
URIs, if present, are ignored. Those in the first URI
2015-08-29 17:24:11 +02:00
are used solely. Definition of a base URI overrides all
scheme, host or port values.
.UNINDENT
2015-02-01 10:33:24 +01:00
.SH OPTIONS
.INDENT 0.0
.TP
.B \-n, \-\-requests=<N>
2015-10-19 17:22:03 +02:00
Number of requests across all clients. If it is used
with \fI\%\-\-timing\-script\-file\fP option, this option specifies
the number of requests each client performs rather than
2017-09-20 15:41:56 +02:00
the number of requests across all clients. This option
is ignored if timing\-based benchmarking is enabled (see
\fI\%\-\-duration\fP option).
.sp
Default: \fB1\fP
.UNINDENT
.INDENT 0.0
.TP
.B \-c, \-\-clients=<N>
2015-10-19 17:22:03 +02:00
Number of concurrent clients. With \fI\%\-r\fP option, this
specifies the maximum number of connections to be made.
.sp
Default: \fB1\fP
.UNINDENT
.INDENT 0.0
.TP
.B \-t, \-\-threads=<N>
Number of native threads.
.sp
Default: \fB1\fP
.UNINDENT
.INDENT 0.0
.TP
2015-07-28 17:01:12 +02:00
.B \-i, \-\-input\-file=<PATH>
2015-05-12 03:49:42 +02:00
Path of a file with multiple URIs are separated by EOLs.
2015-01-15 16:10:16 +01:00
This option will disable URIs getting from command\-line.
2015-07-28 17:01:12 +02:00
If \(aq\-\(aq is given as <PATH>, URIs will be read from stdin.
2015-01-15 16:10:16 +01:00
URIs are used in this order for each client. All URIs
are used, then first URI is used and then 2nd URI, and
so on. The scheme, host and port in the subsequent
URIs, if present, are ignored. Those in the first URI
2015-08-29 17:24:11 +02:00
are used solely. Definition of a base URI overrides all
scheme, host or port values.
.UNINDENT
.INDENT 0.0
2014-10-27 16:35:44 +01:00
.TP
2016-01-11 08:23:32 +01:00
.B \-m, \-\-max\-concurrent\-streams=<N>
Max concurrent streams to issue per session. When
http/1.1 is used, this specifies the number of HTTP
pipelining requests in\-flight.
.sp
2016-01-11 08:23:32 +01:00
Default: \fB1\fP
.UNINDENT
.INDENT 0.0
.TP
.B \-w, \-\-window\-bits=<N>
2015-01-15 16:10:16 +01:00
Sets the stream level initial window size to (2**<N>)\-1.
2015-04-27 15:36:34 +02:00
.sp
Default: \fB30\fP
.UNINDENT
.INDENT 0.0
.TP
.B \-W, \-\-connection\-window\-bits=<N>
2015-01-15 16:10:16 +01:00
Sets the connection level initial window size to
2017-12-17 06:16:53 +01:00
(2**<N>)\-1.
2015-04-27 15:36:34 +02:00
.sp
Default: \fB30\fP
.UNINDENT
.INDENT 0.0
.TP
.B \-H, \-\-header=<HEADER>
2014-10-27 16:35:44 +01:00
Add/Override a header to the requests.
.UNINDENT
.INDENT 0.0
2014-10-27 16:35:44 +01:00
.TP
2015-07-11 05:45:23 +02:00
.B \-\-ciphers=<SUITE>
Set allowed cipher list. The format of the string is
described in OpenSSL ciphers(1).
2017-01-09 06:55:51 +01:00
.sp
2017-03-27 15:48:00 +02:00
Default: \fBECDHE\-ECDSA\-AES256\-GCM\-SHA384:ECDHE\-RSA\-AES256\-GCM\-SHA384:ECDHE\-ECDSA\-CHACHA20\-POLY1305:ECDHE\-RSA\-CHACHA20\-POLY1305:ECDHE\-ECDSA\-AES128\-GCM\-SHA256:ECDHE\-RSA\-AES128\-GCM\-SHA256:ECDHE\-ECDSA\-AES256\-SHA384:ECDHE\-RSA\-AES256\-SHA384:ECDHE\-ECDSA\-AES128\-SHA256:ECDHE\-RSA\-AES128\-SHA256\fP
2015-07-11 05:45:23 +02:00
.UNINDENT
.INDENT 0.0
.TP
.B \-p, \-\-no\-tls\-proto=<PROTOID>
2015-01-15 16:10:16 +01:00
Specify ALPN identifier of the protocol to be used when
accessing http URI without SSL/TLS.
2017-12-17 06:16:53 +01:00
Available protocols: h2c and http/1.1
.sp
2015-05-15 17:10:41 +02:00
Default: \fBh2c\fP
.UNINDENT
.INDENT 0.0
.TP
2015-07-28 17:01:12 +02:00
.B \-d, \-\-data=<PATH>
2015-03-26 16:37:12 +01:00
Post FILE to server. The request method is changed to
2016-06-13 17:26:36 +02:00
POST. For http/1.1 connection, if \fI\%\-d\fP is used, the
maximum number of in\-flight pipelined requests is set to
1.
2015-03-26 16:37:12 +01:00
.UNINDENT
.INDENT 0.0
.TP
2015-07-28 17:01:12 +02:00
.B \-r, \-\-rate=<N>
2015-08-08 18:13:35 +02:00
Specifies the fixed rate at which connections are
2015-07-28 17:01:12 +02:00
created. The rate must be a positive integer,
representing the number of connections to be made per
2015-10-19 17:22:03 +02:00
rate period. The maximum number of connections to be
made is given in \fI\%\-c\fP option. This rate will be
distributed among threads as evenly as possible. For
2016-07-16 12:10:34 +02:00
example, with \fI\%\-t\fP2 and \fI\%\-r\fP4, each thread gets 2
2015-10-19 17:22:03 +02:00
connections per period. When the rate is 0, the program
will run as it normally does, creating connections at
whatever variable rate it wants. The default value for
2018-05-07 14:43:13 +02:00
this option is 0. \fI\%\-r\fP and \fI\%\-D\fP are mutually exclusive.
2015-07-28 17:01:12 +02:00
.UNINDENT
.INDENT 0.0
.TP
2015-10-19 17:22:03 +02:00
.B \-\-rate\-period=<DURATION>
Specifies the time period between creating connections.
The period must be a positive number, representing the
length of the period in time. This option is ignored if
the rate option is not used. The default value for this
option is 1s.
2015-07-28 17:01:12 +02:00
.UNINDENT
.INDENT 0.0
.TP
2017-09-20 15:41:56 +02:00
.B \-D, \-\-duration=<N>
Specifies the main duration for the measurements in case
2018-05-07 14:43:13 +02:00
of timing\-based benchmarking. \fI\%\-D\fP and \fI\%\-r\fP are mutually
exclusive.
2017-09-20 15:41:56 +02:00
.UNINDENT
.INDENT 0.0
.TP
.B \-\-warm\-up\-time=<DURATION>
Specifies the time period before starting the actual
measurements, in case of timing\-based benchmarking.
Needs to provided along with \fI\%\-D\fP option.
.UNINDENT
.INDENT 0.0
.TP
2015-10-19 17:22:03 +02:00
.B \-T, \-\-connection\-active\-timeout=<DURATION>
2015-08-29 17:24:11 +02:00
Specifies the maximum time that h2load is willing to
keep a connection open, regardless of the activity on
2015-10-19 17:22:03 +02:00
said connection. <DURATION> must be a positive integer,
specifying the amount of time to wait. When no timeout
value is set (either active or inactive), h2load will
keep a connection open indefinitely, waiting for a
2015-08-29 17:24:11 +02:00
response.
.UNINDENT
.INDENT 0.0
.TP
2015-10-19 17:22:03 +02:00
.B \-N, \-\-connection\-inactivity\-timeout=<DURATION>
2015-08-29 17:24:11 +02:00
Specifies the amount of time that h2load is willing to
2015-10-19 17:22:03 +02:00
wait to see activity on a given connection. <DURATION>
must be a positive integer, specifying the amount of
time to wait. When no timeout value is set (either
active or inactive), h2load will keep a connection open
2015-08-29 17:24:11 +02:00
indefinitely, waiting for a response.
.UNINDENT
.INDENT 0.0
.TP
.B \-\-timing\-script\-file=<PATH>
2015-10-19 17:22:03 +02:00
Path of a file containing one or more lines separated by
EOLs. Each script line is composed of two tab\-separated
fields. The first field represents the time offset from
the start of execution, expressed as a positive value of
milliseconds with microsecond resolution. The second
field represents the URI. This option will disable URIs
getting from command\-line. If \(aq\-\(aq is given as <PATH>,
script lines will be read from stdin. Script lines are
used in order for each client. If \fI\%\-n\fP is given, it must
be less than or equal to the number of script lines,
larger values are clamped to the number of script lines.
If \fI\%\-n\fP is not given, the number of requests will default
to the number of script lines. The scheme, host and
port defined in the first URI are used solely. Values
contained in other URIs, if present, are ignored.
Definition of a base URI overrides all scheme, host or
port values.
2015-08-29 17:24:11 +02:00
.UNINDENT
.INDENT 0.0
.TP
2016-01-19 12:55:08 +01:00
.B \-B, \-\-base\-uri=(<URI>|unix:<PATH>)
2015-08-29 17:24:11 +02:00
Specify URI from which the scheme, host and port will be
used for all requests. The base URI overrides all
values defined either at the command line or inside
2016-01-19 12:55:08 +01:00
input files. If argument starts with "unix:", then the
rest of the argument will be treated as UNIX domain
socket path. The connection is made through that path
instead of TCP. In this case, scheme is inferred from
the first URI appeared in the command line or inside
input files as usual.
2015-08-29 17:24:11 +02:00
.UNINDENT
.INDENT 0.0
.TP
2015-09-23 17:17:14 +02:00
.B \-\-npn\-list=<LIST>
Comma delimited list of ALPN protocol identifier sorted
in the order of preference. That means most desirable
protocol comes first. This is used in both ALPN and
NPN. The parameter must be delimited by a single comma
only and any white spaces are treated as a part of
protocol string.
.sp
2017-02-26 10:35:58 +01:00
Default: \fBh2,h2\-16,h2\-14,http/1.1\fP
2015-09-23 17:17:14 +02:00
.UNINDENT
.INDENT 0.0
.TP
2015-11-26 14:25:22 +01:00
.B \-\-h1
Short hand for \fI\%\-\-npn\-list\fP=http/1.1
\fI\%\-\-no\-tls\-proto\fP=http/1.1, which effectively force
http/1.1 for both http and https URI.
.UNINDENT
.INDENT 0.0
.TP
2016-10-08 04:46:03 +02:00
.B \-\-header\-table\-size=<SIZE>
Specify decoder header table size.
.sp
Default: \fB4K\fP
.UNINDENT
.INDENT 0.0
.TP
.B \-\-encoder\-header\-table\-size=<SIZE>
Specify encoder header table size. The decoder (server)
specifies the maximum dynamic table size it accepts.
Then the negotiated dynamic table size is the minimum of
this option value and the value which server specified.
.sp
Default: \fB4K\fP
.UNINDENT
.INDENT 0.0
.TP
2019-01-18 13:20:47 +01:00
.B \-\-log\-file=<PATH>
Write per\-request information to a file as tab\-separated
columns: start time as microseconds since epoch; HTTP
status code; microseconds until end of response. More
columns may be added later. Rows are ordered by end\-of\-
response time when using one worker thread, but may
appear slightly out of order with multiple threads due
to buffering. Status code is \-1 for failed streams.
.UNINDENT
.INDENT 0.0
.TP
.B \-v, \-\-verbose
Output debug information.
.UNINDENT
.INDENT 0.0
.TP
.B \-\-version
Display version information and exit.
.UNINDENT
.INDENT 0.0
.TP
.B \-h, \-\-help
Display this help and exit.
.UNINDENT
2015-10-19 17:22:03 +02:00
.sp
2016-10-08 04:46:03 +02:00
The <SIZE> argument is an integer and an optional unit (e.g., 10K is
10 * 1024). Units are K, M and G (powers of 1024).
.sp
2015-10-19 17:22:03 +02:00
The <DURATION> argument is an integer and an optional unit (e.g., 1s
is 1 second and 500ms is 500 milliseconds). Units are h, m, s or ms
(hours, minutes, seconds and milliseconds, respectively). If a unit
is omitted, a second is used as unit.
2015-02-01 10:19:16 +01:00
.SH OUTPUT
.INDENT 0.0
.TP
.B requests
.INDENT 7.0
.TP
.B total
The number of requests h2load was instructed to make.
.TP
.B started
The number of requests h2load has started.
.TP
.B done
The number of requests completed.
.TP
.B succeeded
The number of requests completed successfully. Only HTTP status
code 2xx or3xx are considered as success.
.TP
.B failed
The number of requests failed, including HTTP level failures
(non\-successful HTTP status code).
.TP
.B errored
The number of requests failed, except for HTTP level failures.
2015-08-14 16:59:47 +02:00
This is the subset of the number reported in \fBfailed\fP and most
likely the network level failures or stream was reset by
RST_STREAM.
2015-08-29 17:24:11 +02:00
.TP
.B timeout
The number of requests whose connection timed out before they were
completed. This is the subset of the number reported in
\fBerrored\fP\&.
2015-02-01 10:19:16 +01:00
.UNINDENT
.TP
.B status codes
The number of status code h2load received.
.TP
.B traffic
.INDENT 7.0
.TP
.B total
The number of bytes received from the server "on the wire". If
2015-11-09 03:40:35 +01:00
requests were made via TLS, this value is the number of decrypted
2015-02-01 10:19:16 +01:00
bytes.
.TP
.B headers
2015-11-12 15:14:45 +01:00
The number of response header bytes from the server without
decompression. The \fBspace savings\fP shows efficiency of header
compression. Let \fBdecompressed(headers)\fP to the number of bytes
used for header fields after decompression. The \fBspace savings\fP
is calculated by (1 \- \fBheaders\fP / \fBdecompressed(headers)\fP) *
100. For HTTP/1.1, this is usually 0.00%, since it does not have
2017-12-17 06:16:53 +01:00
header compression. For HTTP/2, it shows some insightful numbers.
2015-02-01 10:19:16 +01:00
.TP
.B data
The number of response body bytes received from the server.
.UNINDENT
.TP
.B time for request
.INDENT 7.0
.TP
.B min
The minimum time taken for request and response.
.TP
.B max
The maximum time taken for request and response.
.TP
.B mean
The mean time taken for request and response.
.TP
.B sd
2015-05-07 13:55:10 +02:00
The standard deviation of the time taken for request and response.
2015-02-01 10:19:16 +01:00
.TP
.B +/\- sd
The fraction of the number of requests within standard deviation
range (mean +/\- sd) against total number of successful requests.
.UNINDENT
2015-05-07 13:55:10 +02:00
.TP
.B time for connect
.INDENT 7.0
.TP
.B min
2018-11-02 07:44:57 +01:00
The minimum time taken to connect to a server including TLS
handshake.
2015-05-07 13:55:10 +02:00
.TP
.B max
2018-11-02 07:44:57 +01:00
The maximum time taken to connect to a server including TLS
handshake.
2015-05-07 13:55:10 +02:00
.TP
.B mean
2018-11-02 07:44:57 +01:00
The mean time taken to connect to a server including TLS
handshake.
2015-05-07 13:55:10 +02:00
.TP
.B sd
The standard deviation of the time taken to connect to a server.
.TP
.B +/\- sd
The fraction of the number of connections within standard
deviation range (mean +/\- sd) against total number of successful
connections.
.UNINDENT
.TP
.B time for 1st byte (of (decrypted in case of TLS) application data)
.INDENT 7.0
.TP
.B min
The minimum time taken to get 1st byte from a server.
.TP
.B max
The maximum time taken to get 1st byte from a server.
.TP
.B mean
The mean time taken to get 1st byte from a server.
.TP
.B sd
The standard deviation of the time taken to get 1st byte from a
server.
.TP
.B +/\- sd
The fraction of the number of connections within standard
deviation range (mean +/\- sd) against total number of successful
connections.
.UNINDENT
.TP
2015-12-23 14:52:11 +01:00
.B req/s
.INDENT 7.0
.TP
.B min
The minimum request per second among all clients.
.TP
.B max
The maximum request per second among all clients.
.TP
.B mean
The mean request per second among all clients.
.TP
.B sd
The standard deviation of request per second among all clients.
server.
.TP
.B +/\- sd
The fraction of the number of connections within standard
deviation range (mean +/\- sd) against total number of successful
connections.
.UNINDENT
2015-02-01 10:19:16 +01:00
.UNINDENT
2015-04-27 15:36:34 +02:00
.SH FLOW CONTROL
.sp
h2load sets large flow control window by default, and effectively
disables flow control to avoid under utilization of server
performance. To set smaller flow control window, use \fI\%\-w\fP and
\fI\%\-W\fP options. For example, use \fB\-w16 \-W16\fP to set default
2017-12-17 06:16:53 +01:00
window size described in HTTP/2 protocol specification.
.SH SEE ALSO
.sp
2016-07-16 12:10:34 +02:00
\fBnghttp(1)\fP, \fBnghttpd(1)\fP, \fBnghttpx(1)\fP
.SH AUTHOR
Tatsuhiro Tsujikawa
.SH COPYRIGHT
2016-01-11 08:23:32 +01:00
2012, 2015, 2016, Tatsuhiro Tsujikawa
.\" Generated by docutils manpage writer.
.