System-Benchmarks unter Linux: Sinnvoll und vergleichbar
System-Benchmarks unter Linux messen die Leistung von CPU, Speicher, Festplatten und Netzwerk, um Schwachstellen aufzudecken und Optimierungen zu planen. Sie sind sinnvoll, da sie reale Performance-Probleme vor der Inbetriebnahme enthüllen und fundierte Hardware-Upgrades ermöglichen
Mit folgenden Befehlen kann man unter Linux das System auf Geschwindigkeit testen und somit vergleichen.
sudo apt update
sudo apt upgrade -y
sudo apt install sysbench fio bpytop iperf3 util-linux
lscpu
lspci
ping ipv4.ipv64.net
ping ipv6.ipv64.net
iperf3 -c speedtest.myloc.de -p 5200 -P 10 -4
iperf3 -c speedtest.myloc.de -p 5200 -P 10 -4 -R
sysbench cpu run
sysbench memory run
sudo dd if=/dev/zero of=/root/test.iso bs=128k count=10000
## Schreib- / Lesetests
fio --name=write-test --size=1G --filename=/tmp/fio-testfile --bs=128k --rw=write --direct=1 --numjobs=1 --time_based --runtime=30 --group_reporting
fio --name=read-test --size=1G --filename=/tmp/fio-testfile --bs=128k --rw=read --direct=1 --numjobs=1 --time_based --runtime=30 --group_reporting
fio --name=realistic-test --filename=/tmp/fio-testfile --size=2G --bs=64k --rw=randrw --rwmixread=70 --direct=1 --iodepth=16 --numjobs=4 --time_based --runtime=60 --group_reporting
Testergebnisse
Strato VPS 4-8
Produkt: VC 4-8
Preis pro Monat: 6,00 €
Model name: AMD EPYC-Milan Processor
CPU family: 25
Model: 1
Thread(s) per core: 2
Core(s) per socket: 1
Socket(s): 2
IPv4 Ping ~ 15ms
IPv6 Ping ~ 15ms
iPerf3 Speedtest Upload
[SUM] 0.00-10.00 sec 2.21 GBytes 1.89 Gbits/sec 1294 sender
[SUM] 0.00-10.10 sec 2.19 GBytes 1.87 Gbits/sec receiver
iPerf3 Speedtest Download
[SUM] 0.00-10.01 sec 6.85 GBytes 5.87 Gbits/sec 1513 sender
[SUM] 0.00-10.00 sec 6.82 GBytes 5.86 Gbits/sec receiver
CPU Benchmark
CPU speed:
events per second: 3474.25
General statistics:
total time: 10.0001s
total number of events: 34746
Latency (ms):
min: 0.28
avg: 0.29
max: 0.51
95th percentile: 0.30
sum: 9989.57
Threads fairness:
events (avg/stddev): 34746.0000/0.00
execution time (avg/stddev): 9.9896/0.00
Memory Benchmark
Total operations: 55788126 (5578289.31 per second)
54480.59 MiB transferred (5447.55 MiB/sec)
General statistics:
total time: 10.0001s
total number of events: 55788126
Latency (ms):
min: 0.00
avg: 0.00
max: 0.24
95th percentile: 0.00
sum: 3715.04
Threads fairness:
events (avg/stddev): 55788126.0000/0.00
execution time (avg/stddev): 3.7150/0.00
SSD Schreibtests
Nullen schreiben:
1310720000 bytes (1.3 GB, 1.2 GiB) copied, 1.64173 s, 798 MB/s
FIO Write-test
Jobs: 1 (f=1): [W(1)][100.0%][w=477MiB/s][w=3817 IOPS][eta 00m:00s]
write: IOPS=3827, BW=478MiB/s (502MB/s)(14.0GiB/30001msec); 0 zone resets
clat (usec): min=83, max=4004, avg=259.30, stdev=29.97
lat (usec): min=83, max=4006, avg=260.39, stdev=29.93
clat percentiles (usec):
| 1.00th=[ 143], 5.00th=[ 239], 10.00th=[ 247], 20.00th=[ 253],
| 30.00th=[ 255], 40.00th=[ 258], 50.00th=[ 260], 60.00th=[ 262],
| 70.00th=[ 265], 80.00th=[ 269], 90.00th=[ 273], 95.00th=[ 281],
| 99.00th=[ 322], 99.50th=[ 359], 99.90th=[ 424], 99.95th=[ 523],
| 99.99th=[ 1074]
bw ( KiB/s): min=488192, max=586240, per=100.00%, avg=490205.29, stdev=12719.19, samples=59
iops : min= 3814, max= 4580, avg=3829.73, stdev=99.37, samples=59
FIO Read-test
Jobs: 1 (f=1): [R(1)][100.0%][r=355MiB/s][r=2843 IOPS][eta 00m:00s]
read: IOPS=2813, BW=352MiB/s (369MB/s)(10.3GiB/30001msec)
clat (usec): min=69, max=4284, avg=354.39, stdev=63.95
lat (usec): min=69, max=4284, avg=354.51, stdev=63.96
clat percentiles (usec):
| 1.00th=[ 225], 5.00th=[ 273], 10.00th=[ 289], 20.00th=[ 310],
| 30.00th=[ 326], 40.00th=[ 334], 50.00th=[ 347], 60.00th=[ 359],
| 70.00th=[ 392], 80.00th=[ 408], 90.00th=[ 420], 95.00th=[ 429],
| 99.00th=[ 486], 99.50th=[ 570], 99.90th=[ 824], 99.95th=[ 988],
| 99.99th=[ 1811]
bw ( KiB/s): min=350720, max=376320, per=100.00%, avg=360300.47, stdev=5005.10, samples=59
iops : min= 2740, max= 2940, avg=2814.85, stdev=39.10, samples=59
FIO Realistic-Test
READ: bw=476MiB/s (499MB/s), 476MiB/s-476MiB/s (499MB/s-499MB/s), io=27.9GiB (30.0GB), run=60001-60001msec
WRITE: bw=205MiB/s (215MB/s), 205MiB/s-205MiB/s (215MB/s-215MB/s), io=12.0GiB (12.9GB), run=60001-60001msec
Netcup VPS 4-8
Produkt: VPS 1000 G12 Pro
Preis pro Monat: 6,84 €
Model name: AMD EPYC-Genoa Processor
CPU family: 25
Model: 17
Thread(s) per core: 1
Core(s) per socket: 1
Socket(s): 4
IPv4 Ping ~ 3,5ms
IPv6 Ping ~ 3,5ms
iPerf3 Speedtest Upload
[SUM] 0.00-10.00 sec 3.26 GBytes 2.80 Gbits/sec 25822 sender
[SUM] 0.00-10.01 sec 3.18 GBytes 2.73 Gbits/sec receiver
iPerf3 Speedtest Download
[SUM] 0.00-10.01 sec 361 MBytes 302 Mbits/sec 5284 sender
[SUM] 0.00-10.00 sec 338 MBytes 283 Mbits/sec receiver
CPU Benchmark
CPU speed:
events per second: 3521.35
General statistics:
total time: 10.0002s
total number of events: 35219
Latency (ms):
min: 0.22
avg: 0.28
max: 11.39
95th percentile: 0.30
sum: 9984.42
Threads fairness:
events (avg/stddev): 35219.0000/0.00
execution time (avg/stddev): 9.9844/0.00
Memory Benchmark
Total operations: 52107471 (5210092.81 per second)
50886.20 MiB transferred (5087.98 MiB/sec)
General statistics:
total time: 10.0002s
total number of events: 52107471
Latency (ms):
min: 0.00
avg: 0.00
max: 11.06
95th percentile: 0.00
sum: 3924.03
Threads fairness:
events (avg/stddev): 52107471.0000/0.00
execution time (avg/stddev): 3.9240/0.00
SSD Schreibtests
Nullen schreiben:
1310720000 bytes (1.3 GB, 1.2 GiB) copied, 2.55724 s, 513 MB/s
FIO Write-test
Jobs: 1 (f=1): [W(1)][100.0%][w=6774MiB/s][w=54.2k IOPS][eta 00m:00s]
write: IOPS=55.0k, BW=6877MiB/s (7211MB/s)(201GiB/30001msec); 0 zone resets
clat (usec): min=13, max=11293, avg=17.63, stdev=133.57
lat (usec): min=13, max=11293, avg=17.96, stdev=134.00
clat percentiles (usec):
| 1.00th=[ 14], 5.00th=[ 14], 10.00th=[ 15], 20.00th=[ 15],
| 30.00th=[ 15], 40.00th=[ 15], 50.00th=[ 15], 60.00th=[ 16],
| 70.00th=[ 17], 80.00th=[ 18], 90.00th=[ 19], 95.00th=[ 20],
| 99.00th=[ 27], 99.50th=[ 30], 99.90th=[ 39], 99.95th=[ 50],
| 99.99th=[10290]
bw ( MiB/s): min= 6492, max= 7299, per=100.00%, avg=6885.49, stdev=194.98, samples=59
iops : min=51942, max=58396, avg=55083.95, stdev=1559.83, samples=59
FIO Read-test
Jobs: 1 (f=1): [R(1)][100.0%][r=12.4GiB/s][r=101k IOPS][eta 00m:00s]
read: IOPS=106k, BW=13.0GiB/s (13.9GB/s)(390GiB/30001msec)
clat (usec): min=6, max=11121, avg= 9.17, stdev=95.18
lat (usec): min=6, max=11121, avg= 9.21, stdev=95.35
clat percentiles (usec):
| 1.00th=[ 8], 5.00th=[ 8], 10.00th=[ 8], 20.00th=[ 8],
| 30.00th=[ 8], 40.00th=[ 8], 50.00th=[ 8], 60.00th=[ 8],
| 70.00th=[ 8], 80.00th=[ 9], 90.00th=[ 11], 95.00th=[ 12],
| 99.00th=[ 20], 99.50th=[ 25], 99.90th=[ 42], 99.95th=[ 53],
| 99.99th=[ 269]
bw ( MiB/s): min=11999, max=14564, per=100.00%, avg=13299.22, stdev=523.88, samples=59
iops : min=95992, max=116518, avg=106393.83, stdev=4191.04, samples=59
FIO Realistic-Test
READ: bw=13.7GiB/s (14.8GB/s), 13.7GiB/s-13.7GiB/s (14.8GB/s-14.8GB/s), io=825GiB (886GB), run=60001-60001msec
WRITE: bw=6031MiB/s (6324MB/s), 6031MiB/s-6031MiB/s (6324MB/s-6324MB/s), io=353GiB (379GB), run=60001-60001msec
