1sockperf(3Version) sockperf(3Version)
2
6 sockperf - SockPerf is a tool for network performance measurement
7 written in C++.
8
10 SockPerf is a network testing tool oriented to measure network latency
11 and also spikes of network latency. Tool can create UDP/TCP data
12 streams and measure the throughput and latency of a network that is
13 carrying them. SockPerf allows the user to define different parameters
14 that can be used for testing a network, or alternately for optimizing
15 or tuning a network. Tool provides a client and server functionality,
16 and can measure the throughput and latency between the two end-points,
17 either unidirectionally or bidirectionally. This utility can be used in
18 Linux systems.
19
21 People are often concerned about measuring the maximum data throughput
22 rate of a communications link or network access. A typical method of
23 performing a measurement is to transfer a 'large' file and measure the
24 time taken to do so. The throughput is then calculated by dividing the
25 file size by the time to get the throughput in megabits, kilobits, or
26 bits per second. Unfortunately, the results of such an exercise will
27 result in the goodput which is less than the maximum throughput,
28 leading to people believing that their communications link is not
29 operating correctly. In fact, there are many overheads accounted for in
30 good case in addition to transmission overheads, including latency, TCP
31 Receive Window size and machine limitations, which means the calculated
32 goodput does not reflect the maximum achievable throughput.
33 Another important thing of tool capacity is latency measurement.
35 Latency - is the time it takes packet to go from user space program on
36 one machine to user space program on another machine. Being able to
37 quantify latency in terms other than millisecond response time is
38 important when determining the quality of a network. One of available
39 tool that can help administrators do just that is Sockperf.
40 SockPerf works as an on-demand client and server test. How this works
42 is that one system runs the Sockperf server over a specified port and
43 another system functions as a client running the Sockperf client. The
44 binaries are the same, and there is an option to have the role of
45 client or server, so the roles can easily be reversed if necessary.
46
48 Sockperf tests UDP/TCP network connection and maintenance following
49 functionality:
50 • Measure latency;
52 • Measure TX/RX bandwidth;
54 • Measure packet loss;
56 • Multicast;
58 • Multi-threaded;
60 features:
62 • Measure the RTT of packets in descrete way;
64 • Provide full log of packet times;
66 • Provide few modes to monitor multiple file descriptors as
68 recvfrom/select/poll/epoll;
69 • Improved CPU utilization;
71 Initially the tool was developed to demonstrate advantages of
73 Mellanox's Messaging Accelerator (VMA). VMA is a socket API based,
74 dynamically linked, user space Linux library which serves to
75 transparently enhance the performance of Multicast/UDP/TCP networking
76 heavy applications over the InfiniBand and Ethernet network. More
77 interested user can read detail information at http://www.mellanox.com.
78 Actually Sockperf can be used natively, or with VMA acceleration and
79 see the benefit of VMA.
80 SockPerf operates by sending packets from the client to the server,
82 which then sends the packets back to the client. This measured round
83 trip time is the route trip time (RTT) between the two machines on a
84 specific network path. The average RTT is calculated by dividing the
85 total number of packets that perform this round trip by some fixed
86 period of time. The average latency for a given one-way path between
87 the two machines is the average RTT divided by two.
88 SockPerf can work as server or execute under-load, ping-pong, playback
90 and throughput tests and be a server or a client.
91 SockPerf can be launched in single point manner that is name as the
93 first mode and using special formatted feed file named as the second
94 mode.
95 Mode One:
97 $sockperf server -i 224.18.7.81 -p 5001
99 sockperf: == version #3.5-no.git ==
100 sockperf: [SERVER] listen on:
101 [ 0] IP = 224.18.7.81 PORT = 5001 # UDP
102 sockperf: Warmup stage (sending a few dummy messages)...
103 sockperf: [tid 4701] using recvfrom() to block on socket(s
104 Mode Two:
107 $sockperf server -f conf.file -F e
109 sockperf: == version #3.5-no.git ==
110 sockperf: [SERVER] listen on:
111 [ 0] IP = 5.2.1.3 PORT = 6671 # TCP
112 [ 1] IP = 5.2.1.3 PORT = 6672 # TCP
113 [ 2] IP = 5.2.1.3 PORT = 6673 # TCP
114 [ 3] IP = 5.2.1.3 PORT = 6674 # TCP
115 [ 4] IP = 5.2.1.3 PORT = 6675 # TCP
116 sockperf: Warmup stage (sending a few dummy messages)...
117 sockperf: [tid 4805] using epoll() to block on socket(s)
118 Every line in feed file should have following format as
121 or
123 or
125 where
127 • [U|T] - UDP or TCP protocol;
129 • address - Internet Protocol (IP) address or host name (IPv6 addresses
131 must be enclosed in square brackets);
132 • port - Port number;
134 • mc_src_addr - Optional multicast source IP address or host name.
136 • PATH - absolute path for UNIX Domain Socket Linux - line must start
138 with '/' Windows - line must start with absoulte path including
139 directory, e.g 'c:\tmp\test'
140 3.1 Available options
142 The following table describes Sockperf options, and their possible
143 values:
144 -h,-? --help,--usage -Show the help message and exit.
146 --tcp --stream -Use stream socket/TCP protocol (default dgram socket/UDP protocol).
147 -i --ip --addr -Listen on/send to ip <ip> or address <name>.
148 -p --port -Listen on/connect to port <port> (default 11111).
149 -f --file -Read list of connections from file (used in pair with -F option).
150 -F --iomux-type -Type of multiple file descriptors handle [s|select|p|poll|e|epoll|r|recvfrom|x|socketxtreme](default epoll).
151 --timeout -Set select/poll/epoll timeout to <msec>, -1 for infinite (default is 10 msec).
152 -a --activity -Measure activity by printing a '.' for the last <N> messages processed.
153 -A --Activity -Measure activity by printing the duration for last <N> messages processed.
154 --tcp-avoid-nodelay -Stop/Start delivering TCP Messages Immediately (Enable/Disable Nagel). Default is Nagel Disabled except in Throughput where the default is Nagel enabled.
155 --tcp-skip-blocking-send
156 -Enables non-blocking send operation (default OFF).
157 --tos -Allows setting tos
158 --mc-rx-ip,--mc-rx-if -Use mc-rx-ip (IPv4) / mc-rx-if (IPv6). Set ipv4 address / interface index of interface on which to receive multicast messages (can be other then route table).
159 --mc-tx-ip,--mc-tx-if -Use mc-tx-ip (IPv4) / mc-tx-if (IPv6). Set ipv4 address / interface index of interface on which to transmit multicast messages (can be other then route table).
160 --mc-loopback-enable -Enables mc loopback (default disabled).
161 --mc-ttl -Limit the lifetime of the message (default 2).
162 --mc-source-filter -Set address <ip, hostname> of multicast messages source which is allowed to receive from.
163 --uc-reuseaddr -Enables unicast reuse address (default disabled).
164 --lls -Turn on LLS via socket option (value = usec to poll).
165 --buffer-size -Set total socket receive/send buffer <size> in bytes (system defined by default).
166 --nonblocked -Open non-blocked sockets.
167 --recv_looping_num -Set sockperf to loop over recvfrom() until EAGAIN or <N> good received packets, -1 for infinite, must be used with --nonblocked (default 1).
168 --dontwarmup -Don't send warm up messages on start.
169 --pre-warmup-wait -Time to wait before sending warm up messages (seconds).
170 --zcopyread,--vmazcopyread
171 -Use VMA's zero copy reads API (See VMA's readme).
172 --daemonize -Run as daemon.
173 --no-rdtsc -Don't use register when taking time; instead use monotonic clock.
174 --load-vma -Load VMA dynamically even when LD_PRELOAD was not used.
175 --load-xlio -Load XLIO dynamically even when LD_PRELOAD was not used.
176 --rate-limit -use rate limit (packet-pacing), with VMA must be run with VMA_RING_ALLOCATION_LOGIC_TX mode.
177 --set-sock-accl -Set socket acceleration before run (available for some of Mellanox systems)
178 -d --debug -Print extra debug information.
179 3.2 Server
182 Server options are:
183 --threads-num -Run <N> threads on server side (requires '-f' option).
185 --cpu-affinity -Set threads affinity to the given core ids in list format (see: cat /proc/cpuinfo).
186 --rxfiltercb,--vmarxfiltercb
187 -Use VMA's receive path message filter callback API (See VMA's readme).
188 --force-unicast-reply -Force server to reply via unicast.
189 --dont-reply -Server won't reply to the client messages.
190 -m --msg-size -Set maximum message size that the server can receive <size> bytes (default 65507).
191 -g --gap-detection -Enable gap-detection.
192 3.3 Client
195 Sockperf supports different scenarios to run itself as a client. There
196 are under-load, ping-pong, playback and throughput subcommands to
197 select one of the scenarios.
198 • under-load - run sockperf client for latency under load test;
200 • ping-pong - run sockperf client for latency test in ping pong mode;
202 • playback - run sockperf client for latency test using playback of
204 predefined traffic, based on timeline and message size;
205 • throughput - run sockperf client for one way throughput test;
207 General client options are:
209 --sender-affinity -Set sender thread affinity to the given core ids in list format (see: cat /proc/cpuinfo).
211 --receiver-affinity -Set receiver thread affinity to the given core ids in list format (see: cat /proc/cpuinfo).
212 --full-log -Dump full log of all messages send/receive time to the given file in CSV format.
213 --full-rtt -Show results in round-trip-time instead of latency.
214 --giga-size -Print sizes in GigaByte.
215 --increase_output_precision
216 -Increase number of digits after decimal point of the throughput output (from 3 to 9).
217 --dummy-send -Use VMA's dummy send API instead of busy wait, must be higher than regular msg rate.
218 optional: set dummy-send rate per second (default 10,000), usage: --dummy-send [<rate>|max]
219 -t --time -Run for <sec> seconds (default 1, max = 36000000).
220 -n --number-of-packets -Run for n packets sent and received (default 0, max = 100000000).
221 --client_port -Force the client side to bind to a specific port (default = 0).
222 --client_addr -Force the client side to bind to a specific address in IPv4, IPv6, UNIX domain socket format (default = 0).
223 -b --burst -Control the client's number of a messages sent in every burst.
224 --mps -Set number of messages-per-second (default = 10000 - for under-load mode, or max - for ping-pong and throughput modes; for maximum use --mps=max;
225 support --pps for old compatibility).
226 -m --msg-size -Use messages of size <size> bytes (minimum default 14).
227 -r --range -comes with -m <size>, randomly change the messages size in range: <size> +- <N>.
228 --data-integrity -Perform data integrity test.
229 --ci_sig_level -Normal confidence interval significance level for stat reported. Values are between 0 and 100 exclusive (default 99).
230 --histogram -Build histogram of latencies. Histogram arguments formated as binsize:lowerrange:upperrange
231 3.4 Tools
234 SockPerf package contains few scripts that allow to generate special
235 formatted file to launch tool in different configurations.
236 • filter.awk - can be used for filtering lines from the full log file
238 based on given latency range;
239 • gen1.awk - this awk script generates playback files (it is for stable
241 PPS playback file);
242 • gen2.awk - this awk script generates playback files using the input
244 for this script is file with lines of the format: startTime;
245 duration; startPPS; endPPS; msgSize (it is for linear increased and
246 decreased PPS playback file);
247 create playback file using gen1.awk > pfile
249 generated file:
250 # ==== playback file for sockperf - generated by gen1.awk ====
251 #baseTime=1.000000; PPS=200000; runtime=1.000000; interval=0.000005; NUM_RECORDS=200000
252 # file contains 200000 records
253 1.000005000, 12
254 1.000010000, 12
255 1.000015000, 12
256 1.000020000, 12
257 1.000025000, 12
258 1.000030000, 12
259 1.000035000, 12
260 1.000040000, 12
261 ...
262 1.999950000, 12
263 1.999955000, 12
264 1.999960000, 12
265 1.999965000, 12
266 1.999970000, 12
267 1.999975000, 12
268 1.999980000, 12
269 1.999985000, 12
270 1.999990000, 12
271 1.999995000, 12
272 2.000000000, 12
273 #200000 records were written successfully
274 start server on ipX
275 start client using: ./sockperf ping-pong -i <ip-address> -p <port> --playback=pfile
276 3.5 Usage
278 3.5.1 Running Multicast over IPoIB
279 • Configure the routing table to map multicast addresses to the IPoIB
280 interface on both client and server machines, as follows:
281 route add -net 224.0.0.0 netmask 240.0.0.0 dev ib0
283 In this case, ib0 is the IPoIB interface.
286 • Run the server as follows:
288 $sockperf server -i 224.18.7.81 -p 5001
290 sockperf: == version #3.5-no.git ==
291 sockperf: [SERVER] listen on:
292 [ 0] IP = 224.18.7.81 PORT = 5001 # UDP
293 sockperf: Warmup stage (sending a few dummy messages)...
294 sockperf: [tid 30399] using recvfrom() to block on socket(s)
295 • Run the client as follows:
298 $sockperf ping-pong -i 224.18.7.81 -p 5001 -m 16384 -t 10 --mps=max
300 sockperf: == version #3.5-no.git ==
301 sockperf[CLIENT] send on:sockperf: using recvfrom() to block on socket(s)
302 [ 0] IP = 224.18.7.81 PORT = 5001 # UDP
304 sockperf: Warmup stage (sending a few dummy messages)...
305 sockperf: Starting test...
306 sockperf: Test end (interrupted by timer)
307 sockperf: Test ended
308 sockperf: [Total Run] RunTime=10.000 sec; Warm up time=400 msec; SentMessages=240464; ReceivedMessages=240463
309 sockperf: ========= Printing statistics for Server No: 0
310 sockperf: [Valid Duration] RunTime=9.550 sec; SentMessages=229630; ReceivedMessages=229630
311 sockperf: ====> avg-lat= 20.771 (std-dev=5.266)
312 sockperf: # dropped messages = 0; # duplicated messages = 0; # out-of-order messages = 0
313 sockperf: Summary: Latency is 20.771 usec
314 sockperf: Total 229630 observations; each percentile contains 2296.30 observations
315 sockperf: ---> <MAX> observation = 120.108
316 sockperf: ---> percentile 99.999 = 106.349
317 sockperf: ---> percentile 99.990 = 63.772
318 sockperf: ---> percentile 99.900 = 55.940
319 sockperf: ---> percentile 99.000 = 48.619
320 sockperf: ---> percentile 90.000 = 24.295
321 sockperf: ---> percentile 75.000 = 20.358
322 sockperf: ---> percentile 50.000 = 19.279
323 sockperf: ---> percentile 25.000 = 18.641
324 sockperf: ---> <MIN> observation = 16.748
325 3.5.2 Running TCP over Ethernet
328 • Run the server as follows:
329 $sockperf server -i 22.0.0.3 -p 5001 --tcp
331 sockperf: == version #3.5-no.git ==
332 sockperf: [SERVER] listen on:
333 [ 0] IP = 22.0.0.3 PORT = 5001 # TCP
334 sockperf: Warmup stage (sending a few dummy messages)...
335 sockperf: [tid 1567] using recvfrom() to block on socket(s)
336 • Run the client as follows:
339 $sockperf ping-pong -i 22.0.0.3 -p 5001 --tcp -m 64 -t 10 --mps=max
341 sockperf: == version #3.5-no.git ==
342 sockperf[CLIENT] send on:sockperf: using recvfrom() to block on socket(s)
343 [ 0] IP = 22.0.0.3 PORT = 5001 # TCP
345 sockperf: Warmup stage (sending a few dummy messages)...
346 sockperf: Starting test...
347 sockperf: Test end (interrupted by timer)
348 sockperf: Test ended
349 sockperf: [Total Run] RunTime=10.000 sec; Warm up time=400 msec; SentMessages=553625; ReceivedMessages=553624
350 sockperf: ========= Printing statistics for Server No: 0
351 sockperf: [Valid Duration] RunTime=9.550 sec; SentMessages=528579; ReceivedMessages=528579
352 sockperf: ====> avg-lat= 9.017 (std-dev=4.171)
353 sockperf: # dropped messages = 0; # duplicated messages = 0; # out-of-order messages = 0
354 sockperf: Summary: Latency is 9.017 usec
355 sockperf: Total 528579 observations; each percentile contains 5285.79 observations
356 sockperf: ---> <MAX> observation = 98.777
357 sockperf: ---> percentile 99.999 = 72.628
358 sockperf: ---> percentile 99.990 = 17.980
359 sockperf: ---> percentile 99.900 = 16.824
360 sockperf: ---> percentile 99.000 = 16.193
361 sockperf: ---> percentile 90.000 = 14.731
362 sockperf: ---> percentile 75.000 = 14.301
363 sockperf: ---> percentile 50.000 = 6.222
364 sockperf: ---> percentile 25.000 = 5.759
365 sockperf: ---> <MIN> observation = 4.629
366 3.5.3 Running UDP over Ethernet using VMA
369 • More interested user can read detail information about VMA at
370 http://www.mellanox.com.
371 • VMA_SPEC=latency is a predefined specification profile for latency.
373 • Run the server as follows:
375 $VMA_SPEC=latency LD_PRELOAD=libvma.so sockperf server -i 22.0.0.3 -p 5001
377 VMA INFO: ---------------------------------------------------------------------------
378 VMA INFO: VMA_VERSION: 8.6.10-0 Development Snapshot built on Jun 27 2018 16:06:47
379 VMA INFO: Cmd Line: sockperf server -i 22.0.0.3 -p 5001
380 VMA INFO: Current Time: Tue Sep 18 08:49:23 2018
381 VMA INFO: Pid: 2201
382 VMA INFO: OFED Version: MLNX_OFED_LINUX-4.4-1.0.0.0:
383 VMA INFO: Architecture: x86_64
384 VMA INFO: Node: r-aa-apollo03.mtr.labs.mlnx
385 VMA INFO: ---------------------------------------------------------------------------
386 VMA INFO: VMA Spec Latency [VMA_SPEC]
387 VMA INFO: Log Level INFO [VMA_TRACELEVEL]
388 VMA INFO: Ring On Device Memory TX 16384 [VMA_RING_DEV_MEM_TX]
389 VMA INFO: Tx QP WRE 256 [VMA_TX_WRE]
390 VMA INFO: Tx QP WRE Batching 4 [VMA_TX_WRE_BATCHING]
391 VMA INFO: Rx QP WRE 256 [VMA_RX_WRE]
392 VMA INFO: Rx QP WRE Batching 4 [VMA_RX_WRE_BATCHING]
393 VMA INFO: Rx Poll Loops -1 [VMA_RX_POLL]
394 VMA INFO: Rx Prefetch Bytes Before Poll 256 [VMA_RX_PREFETCH_BYTES_BEFORE_POLL]
395 VMA INFO: GRO max streams 0 [VMA_GRO_STREAMS_MAX]
396 VMA INFO: Select Poll (usec) -1 [VMA_SELECT_POLL]
397 VMA INFO: Select Poll OS Force Enabled [VMA_SELECT_POLL_OS_FORCE]
398 VMA INFO: Select Poll OS Ratio 1 [VMA_SELECT_POLL_OS_RATIO]
399 VMA INFO: Select Skip OS 1 [VMA_SELECT_SKIP_OS]
400 VMA INFO: CQ Drain Interval (msec) 100 [VMA_PROGRESS_ENGINE_INTERVAL]
401 VMA INFO: CQ Interrupts Moderation Disabled [VMA_CQ_MODERATION_ENABLE]
402 VMA INFO: CQ AIM Max Count 128 [VMA_CQ_AIM_MAX_COUNT]
403 VMA INFO: CQ Adaptive Moderation Disabled [VMA_CQ_AIM_INTERVAL_MSEC]
404 VMA INFO: CQ Keeps QP Full Disabled [VMA_CQ_KEEP_QP_FULL]
405 VMA INFO: TCP nodelay 1 [VMA_TCP_NODELAY]
406 VMA INFO: Avoid sys-calls on tcp fd Enabled [VMA_AVOID_SYS_CALLS_ON_TCP_FD]
407 VMA INFO: Internal Thread Affinity 0 [VMA_INTERNAL_THREAD_AFFINITY]
408 VMA INFO: Thread mode Single [VMA_THREAD_MODE]
409 VMA INFO: Mem Allocate type 2 (Huge Pages) [VMA_MEM_ALLOC_TYPE]
410 VMA INFO: ---------------------------------------------------------------------------
411 sockperf: == version #3.5-no.git ==
412 sockperf: [SERVER] listen on:
413 [ 0] IP = 22.0.0.3 PORT = 5001 # UDP
414 sockperf: Warmup stage (sending a few dummy messages)...
415 sockperf: [tid 2201] using recvfrom() to block on socket(s)
416 • Run the client as follows:
419 $VMA_SPEC=latency LD_PRELOAD=libvma.so sockperf ping-pong -i 22.0.0.3 -p 5001 -m 64 -t 10 --mps=max
421 VMA INFO: ---------------------------------------------------------------------------
422 VMA INFO: VMA_VERSION: 8.6.10-0 Development Snapshot built on Jun 27 2018 16:06:47
423 VMA INFO: Cmd Line: sockperf ping-pong -i 22.0.0.3 -p 5001 -m 64 -t 10 --mps=max
424 VMA INFO: Current Time: Tue Sep 18 08:47:50 2018
425 VMA INFO: Pid: 20134
426 VMA INFO: OFED Version: MLNX_OFED_LINUX-4.4-1.0.0.0:
427 VMA INFO: Architecture: x86_64
428 VMA INFO: Node: r-aa-apollo04.mtr.labs.mlnx
429 VMA INFO: ---------------------------------------------------------------------------
430 VMA INFO: VMA Spec Latency [VMA_SPEC]
431 VMA INFO: Log Level INFO [VMA_TRACELEVEL]
432 VMA INFO: Ring On Device Memory TX 16384 [VMA_RING_DEV_MEM_TX]
433 VMA INFO: Tx QP WRE 256 [VMA_TX_WRE]
434 VMA INFO: Tx QP WRE Batching 4 [VMA_TX_WRE_BATCHING]
435 VMA INFO: Rx QP WRE 256 [VMA_RX_WRE]
436 VMA INFO: Rx QP WRE Batching 4 [VMA_RX_WRE_BATCHING]
437 VMA INFO: Rx Poll Loops -1 [VMA_RX_POLL]
438 VMA INFO: Rx Prefetch Bytes Before Poll 256 [VMA_RX_PREFETCH_BYTES_BEFORE_POLL]
439 VMA INFO: GRO max streams 0 [VMA_GRO_STREAMS_MAX]
440 VMA INFO: Select Poll (usec) -1 [VMA_SELECT_POLL]
441 VMA INFO: Select Poll OS Force Enabled [VMA_SELECT_POLL_OS_FORCE]
442 VMA INFO: Select Poll OS Ratio 1 [VMA_SELECT_POLL_OS_RATIO]
443 VMA INFO: Select Skip OS 1 [VMA_SELECT_SKIP_OS]
444 VMA INFO: CQ Drain Interval (msec) 100 [VMA_PROGRESS_ENGINE_INTERVAL]
445 VMA INFO: CQ Interrupts Moderation Disabled [VMA_CQ_MODERATION_ENABLE]
446 VMA INFO: CQ AIM Max Count 128 [VMA_CQ_AIM_MAX_COUNT]
447 VMA INFO: CQ Adaptive Moderation Disabled [VMA_CQ_AIM_INTERVAL_MSEC]
448 VMA INFO: CQ Keeps QP Full Disabled [VMA_CQ_KEEP_QP_FULL]
449 VMA INFO: TCP nodelay 1 [VMA_TCP_NODELAY]
450 VMA INFO: Avoid sys-calls on tcp fd Enabled [VMA_AVOID_SYS_CALLS_ON_TCP_FD]
451 VMA INFO: Internal Thread Affinity 0 [VMA_INTERNAL_THREAD_AFFINITY]
452 VMA INFO: Thread mode Single [VMA_THREAD_MODE]
453 VMA INFO: Mem Allocate type 2 (Huge Pages) [VMA_MEM_ALLOC_TYPE]
454 VMA INFO: ---------------------------------------------------------------------------
455 sockperf: == version #3.5-no.git ==
456 sockperf[CLIENT] send on:sockperf: using recvfrom() to block on socket(s)
457 [ 0] IP = 22.0.0.3 PORT = 5001 # UDP
459 sockperf: Warmup stage (sending a few dummy messages)...
460 sockperf: Starting test...
461 sockperf: Test end (interrupted by timer)
462 sockperf: Test ended
463 sockperf: [Total Run] RunTime=10.000 sec; Warm up time=400 msec; SentMessages=5166035; ReceivedMessages=5166034
464 sockperf: ========= Printing statistics for Server No: 0
465 sockperf: [Valid Duration] RunTime=9.550 sec; SentMessages=4951987; ReceivedMessages=4951987
466 sockperf: ====> avg-lat= 0.951 (std-dev=0.034)
467 sockperf: # dropped messages = 0; # duplicated messages = 0; # out-of-order messages = 0
468 sockperf: Summary: Latency is 0.951 usec
469 sockperf: Total 4951987 observations; each percentile contains 49519.87 observations
470 sockperf: ---> <MAX> observation = 4.476
471 sockperf: ---> percentile 99.999 = 1.318
472 sockperf: ---> percentile 99.990 = 1.270
473 sockperf: ---> percentile 99.900 = 1.179
474 sockperf: ---> percentile 99.000 = 1.110
475 sockperf: ---> percentile 90.000 = 0.967
476 sockperf: ---> percentile 75.000 = 0.952
477 sockperf: ---> percentile 50.000 = 0.943
478 sockperf: ---> percentile 25.000 = 0.936
479 sockperf: ---> <MIN> observation = 0.895
480
483 Read 'copying' file in the root place.
484
486 5.1 Requirements
487 What you will need to compile sockperf on Unix systems
488 • perl 5.8+ (used by the automake tools)
490 • GNU make tools: automake 1.7+, autoconf 2.57+, m4 1.4+ and libtool
492 1.4+
493 • A Compiler, among those tested are:
495 • gcc4+ (Ubuntu)
497 • gcc4+ (Red Hat)
499 5.2 Options to compile
501 5.3 How to install
502 Download sockperf-<version>.tar.gz.
503 Uncompress *.tar.gz file in Unix systems in the same folder with the
505 file by runing the following command in the shell:
506 tar -zxvf sockperf-<version>.tar.gz
508 or 2 command:
511 gzip -d ./sockperf-<version>.tar.gz
513 tar -xf ./sockperf-<version>.tar
514 The sockperf package uses the GNU autotools compilation and
517 installation framework.
518 These are generic installation instructions.
520 The `configure' shell script attempts to guess correct values for
522 various system-dependent variables used during compilation. It uses
523 those values to create a `Makefile' in each directory of the package.
524 It may also create one or more `.h' files containing system-dependent
525 definitions. Finally, it creates a shell script `config.status' that
526 you can run in the future to recreate the current configuration, a file
527 `config.cache' that saves the results of its tests to speed up
528 reconfiguring, and a file `config.log' containing compiler output
529 (useful mainly for debugging `configure').
530 If you need to do unusual things to compile the package, please try to
532 figure out how `configure' could check whether to do them, and mail
533 diffs or instructions to the address given in the `README' so they can
534 be considered for the next release. If at some point `config.cache'
535 contains results you don't want to keep, you may remove or edit it.
536 The file `configure.in' is used to create `configure' by a program
538 called `autoconf'. You only need `configure.in' if you want to change
539 it or regenerate `configure' using a newer version of `autoconf'.
540 The simplest way to compile this package is:
542 1. `cd' to the directory containing the package's source code and type
544 `./configure' to configure the package for your system. If you're
545 using `csh' on an old version of System V, you might need to type
546 `sh ./configure' instead to prevent `csh' from trying to execute
547 `configure' itself.
548 Running `configure' takes awhile. While running, it prints some
550 messages telling which features it is checking for.
551 $ ./configure --prefix=<path to install>
553 There are several options to ./config (or ./Configure) to customize the
556 build:
557 To enable test scripts
559 $ ./configure --prefix=<path to install> --enable-test
561 To enable the documentation
564 $ ./configure --prefix=<path to install> --enable-doc
566 To enable the special scripts
569 $ ./configure --prefix=<path to install> --enable-tool
571 To compile with debug symbols and information:
574 $ ./configure --prefix=<path to install> --enable-debug
576 This will define the _DEBUG variable at compile time.
579 Type './configure --help' for a list of all the configure options. Some
581 of the options are generic autoconf options, while the SockPerf
582 specific options are prefixed with 'SOCKPERF:' in the help text.
583 2. Type `make' to compile the package.
585 $ make
587 3. Optionally, type `make check' to run any self-tests that come with
590 the package.
591 4. Type `make install' to install the programs and any data files and
593 documentation.
594 $ make install
596 5. You can remove the program binaries and object files from the
599 source code directory by typing `make clean'. To also remove the
600 files that `configure' created (so you can compile the package for
601 a different kind of computer), type `make distclean'. There is also
602 a `make maintainer-clean' target, but that is intended mainly for
603 the package's developers. If you use it, you may have to get all
604 sorts of other programs in order to regenerate files that came with
605 the distribution.
606sockperf 3.10-no.git" sockperf(3Version)