IPerf2

Features

• TOO MANY TO LIST SEE THE RELEASE NOTES OR THE MAN PAGE!!
• Fix portability, compile and tested with Linux, Win10, Win7, WinXP, MacOS, Android and some set-top box OSs.
• Require -u for UDP (-b no longer defaults to UDP)
• Improved performance
• Enhanced reporting with -e
• Support smaller report intervals (100 us or greater, configure --enable-fastsampling for high precision interval time output)
• Support SO_RCVTIMEOUT for server reports regardless of no packets
• Support SO_SNDTIMEO on send so socket write won't block beyond -t or -i
• Support SO_TIMESTAMP for kernel level packet timestamping
• Support end/end latency in mean/min/max/stdev format (UDP) (-e required) (assumes client and server clocks synched, e.g by Precision Time Protocol to an OCXO oscillator per Spectracom)
• Fixes to bind so OS can auto assign source port
• Add local port to bind support (-B option) using colon as separator (v4) or brackets (v6) e.g. iperf -c 192.168.100.100 -B 192.168.100.10:6001 (v4) or for v6, iperf -V -c 2001:e30:1401:2:d46e:b891:3082:b939 -B [2001:e30:1401:2:d46e:b891:3082:b940]:6001
• Support TCP rate limited streams (via the -b) using simplified token bucket
• Support packets per second (UDP) via pps as units, (e.g. -b 1000pps)
• Display PPS in both client and server reports (UDP) (-e required)
• Support realtime scheduler as a command line option (--realtime or -z, assumes proper user privileges)
• Improve client tx code path so actual tx offered rate will converge to the -b value
• Improve accuracy of microsecond delay calls (in platform independent manner) (Use of Kalman filter to predict delay errors and adjust delays per predicted error)
• Display target loop time in initial client header (UDP)
• Fix final latency report sent from server to client (UDP)
• Include standard deviation in latency output
• Suppress unrealistic latency output using (-/-/-/-)
• Use clock_gettime() in critical sections, if available, replacing gettimeofday() calls
• TCP write and error counts (TCP retries and CWND for linux) (-e required)
• TCP read count, TCP read histogram (8 bins) (-e required)
• TCP RTT and CWND values in client reports (-e required, Linux only, RTT units microseconds)
• Added support for -t on the Server (Listener) so servers/listener can be set to timeout and exit
• Add ipv6 link local support (e.g. iperf -c fe80::d03a:d127:75d2:4112%eno1)
• Default ipv6 UDP payload to 1450 bytes per one ethernet frame per payload
• -V on server will accept both IPv4 and IPv6 traffic flows
• Isochronous traffic support (via --isochronous) and frame bursts with variable bit rate (vbr) traffic and frame ids
• Multicast SSM support for both v4 and v6 using -H or -ssm-host, e.g. iperf -s -B ff1e::1 -u -V -H fc00::4
• Latency histograms for both packets and frames (e.g. --udp-histogram=10u,200000, 0.03, 99.97)
• Ethernet frame length checks with --l2checks on the client (UDP only and requires systems that support AF_PACKET)
• Server (read) -b support for TCP (via token bucket)
• UDP write counters and write error counters (requires -e)
• Python asyncio code to manage multiple iperf sessions (found in the flows directory)
• Support for timed transmit starts per --txstart-time <unix.epoch time>
• Support for client incrementing the destination ip with -P via --incr-dstip
• Support for varying the offered load using a log normal distribution around a mean and standard deviation (per -b <mean>,<stdev>),
• Honor -T (ttl) for both unicast and multicast
• Added network power to TCP client and UDP server enhanced output (netpower = throughput / RTT or throughput / end2end delay on server)
• Add TCP connect time to connected to message
• Add support for SO_MAX_PACING_RATE socket option using --fq-rate <value>
• Add configure support for --enable-fastsampling, allowing 100 microsecond report intervals
• Add support for --trip-time on client, requires -e on both client and server and synchronized clocks
• UDP uses 64 bit sequence numbers (though still interoperates with 2.0.5 which uses 32b seq no.)

Follow IPerf2

IPerf2 Web Site