netperf(1)

✨ netperf ✨

okay so bestie, welcome to netperf(1), straight out of the General Commands Manual 💖 this is the page for netperf, the network performance benchmark, and we're gonna go through every single flag together because you deserve to know all of them, no cap 🫶

---

🌸 NAME 🌸

netperf — a network performance benchmark

---

🎀 SYNOPSIS 🎀

netperf [global options] -- [test specific options]

---

💅 DESCRIPTION 💅

So netperf is a benchmark you use to measure all sorts of different aspects of networking performance. Right now its main character energy is focused on bulk data transfer and request/response performance, using either TCP or UDP, over the Berkeley Sockets interface. On top of that, tests for DLPI and Unix Domain Sockets, plus tests for IPv6, may be conditionally compiled-in. iykyk 💫

---

👑 GLOBAL OPTIONS 👑

• -4 — Use AF_INET (aka IPv4) addressing for the control connection and possibly the data connections too.
• -6 — Use AF_INET6 (aka IPv6) addressing for the control connection and possibly the data connections too.
• -a sizespec — Change the send and receive buffer alignments on the local system. This defaults to 8 bytes.
• -A sizespec — Same as -a, but for the remote system.
• -B brandstr — Add brandstr to the output of a test with banners disabled.
• -c [rate] — Request CPU utilization and service demand calculations for the local system. If you also pass the optional rate parameter, netperf will use that instead of calculating the rate itself.
• -C [rate] — Same as -c, but for the remote system.
• -d — Increase the quantity of debugging output displayed during a test (and possibly at the expense of performance, oop 🙈).
• -D [secs,units] () — Display interim results at least every *secs seconds, using units as the initial guess for units per second. Heads up bestie: this is only available when netperf has been configured with --enable-demo.
• -f GMKgmk — Change the units of measure for *_STREAM tests. Capital letters are powers of two, lowercase are powers of ten. 💅
• -F fill_file — Pre-fill the send buffers with data from the named file. The whole point of this is to give you a way to avoid buffers filled with data that's trivially easy to compress. A good choice for a file that should be present on any system is this manpage — netperf.man. Other files may be provided as part of the distribution.
• -h — Display a usage string, and exit. bye 👋
• -H name|ip,family (*) — Set the hostname (or IP address) and address family to use to establish the control connection to the remote system. If you pass a single name with no comma, it'll only set remote_host and will leave selection of the address family for the control connection to the stack, or to a -4 -r -6 command line option.
• -i max,min — Set the maximum and minimum number of iterations when trying to reach certain confidence levels.
• -j — Tell netperf to calculate additional statistics on timing when running an omni test. Whether those stats actually show up depends on whether the corresponding output selectors are present in the output selection. Those selectors are MIN_LATENCY, MAX_LATENCY, P50_LATENCY, P90_LATENCY, P99_LATENCY, MEAN_LATENCY and STDDEV_LATENCY.
• -I lvl,[,intvl] — Specify the confidence level (either 95 or 99 — 99 is the default) and the width of the confidence interval as a percentage (default 10).
• -l testlen — Specify the length of the test (default 10 seconds). A negative value sets the number of request/response transactions, or the number of bytes for a stream test.
• -L name|ip,fam (*) — Set the local name|IP and/or address family for the socket used for the control connection to the remote netserver.
• -n numcpus — Specify the number of CPUs in the system, for those systems where netperf has no way to find the number of CPUs programatically.
• -N — This one tells netperf to not establish a control connection to a remote netserver. Instead it'll try to establish a data connection directly, using only the information supplied by the command line parameters and/or internal defaults. Unless other ports are provided on the command line, by default the data connection goes to the "discard" port for a "STREAM" or "SENDFILE" test, the "echo" port for an "RR" test, or the "chargen" port for a "MAERTS" test.
• -o sizespec — Set an offset from the alignment specified with -a.
• -O sizespec — Same as -o, but for the remote system.
• -p portnum,locport (*) — Direct the control connection to a netserver listening on the specified port, rather than using a "netperf" entry in /etc/services or the internal default (port 12865). If ,locport is specified, the control connection will be established from that local port number. Specifying a single port number with no comma will specify only the remote netserver port number and will leave local port number selection to the stack.
• -P 0|1 — Show (1) or suppress (0) the test banner.
• -S — This option causes an attempt to set SO_KEEPALIVE on the ends of the data connection for tests using BSD Sockets. It'll be done on the netperf side of classic tests, and on both the netperf and netserver side of an omni or migrated test.
• -s seconds — Makes netperf sleep seconds seconds before transferring data over the data connection. lil nap 😴
• -t testname — Specify the test to perform. Valid testnames include, but are not limited to, nor always compiled-in:

TCP_STREAM
TCP_SENDFILE
TCP_MAERTS
TCP_RR
TCP_CRR
UDP_STREAM
UDP_RR
DLCO_STREAM
DLCO_RR
DLCL_STREAM
DLCL_RR
STREAM_STREAM
STREAM_RR
DG_STREAM
DG_RR
SCTP_STREAM
SCTP_STREAM_MANY
SCTP_RR
SCTP_RR_MANY
LOC_CPU
REM_CPU

• -T lcpu,remcpu — Request that netperf be bound to CPU lcpu and/or netserver be bound to CPU rcpu.
• -v verbosity — Set the verbosity level for the test (only with -P).
• -V — Display the netperf version and exit. slay, version check 💫

---

🦋 TEST SPECIFIC OPTIONS 🦋

• -h — Display a usage string based on the test name set with -t, and exit.

Please consult the netperf manual Care and Feeding of Netperf 2.5.X (doc/netperf.[pdf|html|txt]) for more information. Or you can join and send email to netperf-talk@netperf.org. 📬

---

💀 NOTE 💀

okay this part is important, listen up 🫶

For those options taking two parms, at least one must be specified. Specifying one value without a comma will set both parms to that value. Specifying a value with a leading comma will set just the second parm. A value with a trailing comma will set just the first. And if you want to set each parm to unique values, specify both of them and separate them with a comma.

* For these options taking two parms, specifying one value with no comma will only set the first parm and will leave the second at the default value. To set the second value it must be preceded with a comma or be a comma-separated pair. This is to retain previous netperf behaviour.

---

😭 BUGS 😭

ngl there are bound to be bugs. If you think you've found a bug, please mention it in netperf-talk@netperf.org. List membership is required to send email to the list. See http://www.netperf.org/cgi-bin/mailman/listinfo/netperf-talk. And if all else fails, send email to netperf-feedback@netperf.org. 🐛

---

🌈 SEE ALSO 🌈

• netserver(1)
• Care and Feeding of Netperf 2.5.X
• http://www.netperf.org/

---

💜 AUTHORS 💜

• HP Information Networks Division - Networking Performance Team.
• Rick Jones <rick.jones2@hp.com>
• Karen Choy — HP IND
• Dave Shield <daves@csc.liv.ac.uk> (man pages)
• Others too numerous to mention here - see the AUTHORS file

---

and that's the whole thing bestie, we ate 🍓 this page signs off as netperf(1) from the General Commands Manual, section 1, and that footer line is literally just netperf(1) — no version bump, no date drama, just the name holding it down. periodt 💅✨