sysbench 202310
MacBook Pro 2019 13 英寸
1.4 GHz 四核 Intel Core i5
8 GB 2133 MHz LPDDR3
250.69 GB SSD
Intel Iris Plus Graphics 645 1536 MB
MacOS 14.3.1
CPU
sysbench cpu --threads=8 run
sysbench 1.0.20 (using system LuaJIT 2.1.0-beta3)
Running the test with following options:
Number of threads: 8
Initializing random number generator from current time
Prime numbers limit: 10000
Initializing worker threads...
Threads started!
CPU speed:
events per second: 22878757.01
General statistics:
total time: 10.0001s
total number of events: 228821179
Latency (ms):
min: 0.00
avg: 0.00
max: 9.54
95th percentile: 0.00
sum: 14468.53
Threads fairness:
events (avg/stddev): 28602647.3750/164786.01
execution time (avg/stddev): 1.8086/0.08内存
sysbench memory --threads=8 run
sysbench 1.0.20 (using system LuaJIT 2.1.0-beta3)
Running the test with following options:
Number of threads: 8
Initializing random number generator from current time
Running memory speed test with the following options:
block size: 1KiB
total size: 102400MiB
operation: write
scope: global
Initializing worker threads...
Threads started!
Total operations: 104857600 (12536141.16 per second)
102400.00 MiB transferred (12242.33 MiB/sec)
General statistics:
total time: 8.3632s
total number of events: 104857600
Latency (ms):
min: 0.00
avg: 0.00
max: 7.43
95th percentile: 0.00
sum: 41067.49
Threads fairness:
events (avg/stddev): 13107200.0000/0.00
execution time (avg/stddev): 5.1334/0.01硬盘
sysbench fileio --threads=8 --file-test-mode=seqrewr run
sysbench 1.0.20 (using system LuaJIT 2.1.0-beta3)
Running the test with following options:
Number of threads: 8
Initializing random number generator from current time
Extra file open flags: (none)
128 files, 16MiB each
2GiB total file size
Block size 16KiB
Periodic FSYNC enabled, calling fsync() each 100 requests.
Calling fsync() at the end of test, Enabled.
Using synchronous I/O mode
Doing sequential rewrite test
Initializing worker threads...
Threads started!
File operations:
reads/s: 0.00
writes/s: 68869.01
fsyncs/s: 88252.89
Throughput:
read, MiB/s: 0.00
written, MiB/s: 1076.08
General statistics:
total time: 10.0035s
total number of events: 1570934
Latency (ms):
min: 0.00
avg: 0.05
max: 294.46
95th percentile: 0.10
sum: 76610.04
Threads fairness:
events (avg/stddev): 196366.7500/2956.36
execution time (avg/stddev): 9.5763/0.00AMD 7900 台式机
AMD 7900 12 核
宏基掠夺者 16GB*2 DDR5 6000MHz EXPO C30
梵想 S790 1TB PCIe4 TLC 7450MB/s
Windows 10 企业版 LTSC 21H2
CPU
sysbench cpu --threads=24 run
sysbench 1.0.20 (using system LuaJIT 2.1.0-beta3)
Running the test with following options:
Number of threads: 24
Initializing random number generator from current time
Prime numbers limit: 10000
Initializing worker threads...
Threads started!
CPU speed:
events per second: 79302.59
General statistics:
total time: 10.0003s
total number of events: 793080
Latency (ms):
min: 0.16
avg: 0.30
max: 6.68
95th percentile: 0.32
sum: 239870.91
Threads fairness:
events (avg/stddev): 33045.0000/187.14
execution time (avg/stddev): 9.9946/0.00内存
sysbench memory --threads=24 run
sysbench 1.0.20 (using system LuaJIT 2.1.0-beta3)
Running the test with following options:
Number of threads: 24
Initializing random number generator from current time
Running memory speed test with the following options:
block size: 1KiB
total size: 102400MiB
operation: write
scope: global
Initializing worker threads...
Threads started!
Total operations: 104857584 (12493380.85 per second)
102399.98 MiB transferred (12200.57 MiB/sec)
General statistics:
total time: 8.3927s
total number of events: 104857584
Latency (ms):
min: 0.00
avg: 0.00
max: 6.82
95th percentile: 0.00
sum: 185249.97
Threads fairness:
events (avg/stddev): 4369066.0000/0.00
execution time (avg/stddev): 7.7187/0.01硬盘
sysbench fileio --threads=24 --file-test-mode=seqrewr run
sysbench 1.0.20 (using system LuaJIT 2.1.0-beta3)
Running the test with following options:
Number of threads: 24
Initializing random number generator from current time
Extra file open flags: (none)
128 files, 16MiB each
2GiB total file size
Block size 16KiB
Periodic FSYNC enabled, calling fsync() each 100 requests.
Calling fsync() at the end of test, Enabled.
Using synchronous I/O mode
Doing sequential rewrite test
Initializing worker threads...
Threads started!
File operations:
reads/s: 0.00
writes/s: 975.69
fsyncs/s: 1501.56
Throughput:
read, MiB/s: 0.00
written, MiB/s: 15.25
General statistics:
total time: 11.8887s
total number of events: 26380
Latency (ms):
min: 0.00
avg: 9.11
max: 48.32
95th percentile: 24.83
sum: 240198.45
Threads fairness:
events (avg/stddev): 1099.1667/314.30
execution time (avg/stddev): 10.0083/0.01sysbench fileio --threads=2 --file-test-mode=seqrewr run
sysbench 1.0.20 (using system LuaJIT 2.1.0-beta3)
Running the test with following options:
Number of threads: 2
Initializing random number generator from current time
Extra file open flags: (none)
128 files, 16MiB each
2GiB total file size
Block size 16KiB
Periodic FSYNC enabled, calling fsync() each 100 requests.
Calling fsync() at the end of test, Enabled.
Using synchronous I/O mode
Doing sequential rewrite test
Initializing worker threads...
Threads started!
File operations:
reads/s: 0.00
writes/s: 1041.35
fsyncs/s: 1349.63
Throughput:
read, MiB/s: 0.00
written, MiB/s: 16.27
General statistics:
total time: 10.1788s
total number of events: 24082
Latency (ms):
min: 0.00
avg: 0.83
max: 18.81
95th percentile: 1.39
sum: 19996.38
Threads fairness:
events (avg/stddev): 12041.0000/445.00
execution time (avg/stddev): 9.9982/0.00阿里云 ECS t5.small
ecs.t5-lc1m2.small
1 vCPU 2GB
20GB 高效云盘
aliyun_3_9_x64_20G_dengbao_alibase_20231219.vhd
CPU
sysbench cpu --threads=1 run
sysbench 1.0.20 (using system LuaJIT 2.1.0-beta3)
Running the test with following options:
Number of threads: 1
Initializing random number generator from current time
Prime numbers limit: 10000
Initializing worker threads...
Threads started!
CPU speed:
events per second: 889.77
General statistics:
total time: 10.0017s
total number of events: 8901
Latency (ms):
min: 1.08
avg: 1.12
max: 9.90
95th percentile: 1.25
sum: 9993.91
Threads fairness:
events (avg/stddev): 8901.0000/0.00
execution time (avg/stddev): 9.9939/0.00sysbench cpu --threads=2 run
sysbench 1.0.20 (using system LuaJIT 2.1.0-beta3)
Running the test with following options:
Number of threads: 2
Initializing random number generator from current time
Prime numbers limit: 10000
Initializing worker threads...
Threads started!
CPU speed:
events per second: 883.35
General statistics:
total time: 10.0013s
total number of events: 8836
Latency (ms):
min: 1.08
avg: 2.26
max: 1040.54
95th percentile: 1.39
sum: 19979.56
Threads fairness:
events (avg/stddev): 4418.0000/3.00
execution time (avg/stddev): 9.9898/0.01内存
sysbench memory --threads=2 run
sysbench 1.0.20 (using system LuaJIT 2.1.0-beta3)
Running the test with following options:
Number of threads: 2
Initializing random number generator from current time
Running memory speed test with the following options:
block size: 1KiB
total size: 102400MiB
operation: write
scope: global
Initializing worker threads...
Threads started!
Total operations: 40132426 (4012449.34 per second)
39191.82 MiB transferred (3918.41 MiB/sec)
General statistics:
total time: 10.0003s
total number of events: 40132426
Latency (ms):
min: 0.00
avg: 0.00
max: 49.05
95th percentile: 0.00
sum: 8899.84
Threads fairness:
events (avg/stddev): 20066213.0000/39339.00
execution time (avg/stddev): 4.4499/0.24硬盘
sysbench fileio --threads=2 --file-test-mode=seqrewr run
sysbench 1.0.20 (using system LuaJIT 2.1.0-beta3)
Running the test with following options:
Number of threads: 2
Initializing random number generator from current time
Extra file open flags: (none)
128 files, 16MiB each
2GiB total file size
Block size 16KiB
Periodic FSYNC enabled, calling fsync() each 100 requests.
Calling fsync() at the end of test, Enabled.
Using synchronous I/O mode
Doing sequential rewrite test
Initializing worker threads...
Threads started!
File operations:
reads/s: 0.00
writes/s: 5722.37
fsyncs/s: 7343.10
Throughput:
read, MiB/s: 0.00
written, MiB/s: 89.41
General statistics:
total time: 10.0117s
total number of events: 130573
Latency (ms):
min: 0.01
avg: 0.15
max: 32.98
95th percentile: 0.15
sum: 19946.66
Threads fairness:
events (avg/stddev): 65286.5000/582.50
execution time (avg/stddev): 9.9733/0.00阿里云 ECS e-c1m1.large
ecs.e-c1m1.large
2 vCPU 2GB
40GB(2120 IOPS) ESSD Entry 云盘
debian_12_2_x64_20G_alibase_20231012.vhd
CPU
sysbench cpu --threads=1 run
sysbench 1.0.20 (using system LuaJIT 2.1.0-beta3)
Running the test with following options:
Number of threads: 1
Initializing random number generator from current time
Prime numbers limit: 10000
Initializing worker threads...
Threads started!
CPU speed:
events per second: 1053.18
General statistics:
total time: 10.0003s
total number of events: 10534
Latency (ms):
min: 0.94
avg: 0.95
max: 1.09
95th percentile: 0.95
sum: 9996.89
Threads fairness:
events (avg/stddev): 10534.0000/0.00
execution time (avg/stddev): 9.9969/0.00sysbench cpu --threads=2 run
sysbench 1.0.20 (using system LuaJIT 2.1.0-beta3)
Running the test with following options:
Number of threads: 2
Initializing random number generator from current time
Prime numbers limit: 10000
Initializing worker threads...
Threads started!
CPU speed:
events per second: 1762.56
General statistics:
total time: 10.0002s
total number of events: 17629
Latency (ms):
min: 1.11
avg: 1.13
max: 3.95
95th percentile: 1.14
sum: 19994.46
Threads fairness:
events (avg/stddev): 8814.5000/7.50
execution time (avg/stddev): 9.9972/0.00sysbench cpu --threads=4 run
sysbench 1.0.20 (using system LuaJIT 2.1.0-beta3)
Running the test with following options:
Number of threads: 4
Initializing random number generator from current time
Prime numbers limit: 10000
Initializing worker threads...
Threads started!
CPU speed:
events per second: 1763.65
General statistics:
total time: 10.0012s
total number of events: 17641
Latency (ms):
min: 1.12
avg: 2.27
max: 17.16
95th percentile: 7.98
sum: 39993.13
Threads fairness:
events (avg/stddev): 4410.2500/12.30
execution time (avg/stddev): 9.9983/0.00可以看到 2 线程对 events 总数有提升，但是再增加就没有效果。
内存
sysbench memory --threads=2 run
sysbench 1.0.20 (using system LuaJIT 2.1.0-beta3)
Running the test with following options:
Number of threads: 2
Initializing random number generator from current time
Running memory speed test with the following options:
block size: 1KiB
total size: 102400MiB
operation: write
scope: global
Initializing worker threads...
Threads started!
Total operations: 83021696 (8300633.56 per second)
81075.88 MiB transferred (8106.09 MiB/sec)
General statistics:
total time: 10.0001s
total number of events: 83021696
Latency (ms):
min: 0.00
avg: 0.00
max: 0.80
95th percentile: 0.00
sum: 11029.87
Threads fairness:
events (avg/stddev): 41510848.0000/7872.00
execution time (avg/stddev): 5.5149/0.00硬盘
sysbench fileio --threads=2 --file-test-mode=seqrewr run
sysbench 1.0.20 (using system LuaJIT 2.1.0-beta3)
Running the test with following options:
Number of threads: 2
Initializing random number generator from current time
Extra file open flags: (none)
128 files, 16MiB each
2GiB total file size
Block size 16KiB
Periodic FSYNC enabled, calling fsync() each 100 requests.
Calling fsync() at the end of test, Enabled.
Using synchronous I/O mode
Doing sequential rewrite test
Initializing worker threads...
Threads started!
File operations:
reads/s: 0.00
writes/s: 6573.49
fsyncs/s: 8433.32
Throughput:
read, MiB/s: 0.00
written, MiB/s: 102.71
General statistics:
total time: 10.0237s
total number of events: 150189
Latency (ms):
min: 0.01
avg: 0.13
max: 33.04
95th percentile: 0.07
sum: 19975.39
Threads fairness:
events (avg/stddev): 75094.5000/205.50
execution time (avg/stddev): 9.9877/0.00阿里云 ECS e-c1m1.large
ecs.e-c1m1.large
2 vCPU 2GB
40GB(2120 IOPS) ESSD Entry 云盘
aliyun_3_9_x64_20G_dengbao_alibase_20231219.vhd
CPU
sysbench cpu --threads=1 run
sysbench 1.0.20 (using system LuaJIT 2.1.0-beta3)
Running the test with following options:
Number of threads: 1
Initializing random number generator from current time
Prime numbers limit: 10000
Initializing worker threads...
Threads started!
CPU speed:
events per second: 1086.43
General statistics:
total time: 10.0006s
total number of events: 10867
Latency (ms):
min: 0.91
avg: 0.92
max: 1.16
95th percentile: 0.94
sum: 9996.74
Threads fairness:
events (avg/stddev): 10867.0000/0.00
execution time (avg/stddev): 9.9967/0.00sysbench cpu --threads=2 run
sysbench 1.0.20 (using system LuaJIT 2.1.0-beta3)
Running the test with following options:
Number of threads: 2
Initializing random number generator from current time
Prime numbers limit: 10000
Initializing worker threads...
Threads started!
CPU speed:
events per second: 1826.13
General statistics:
total time: 10.0003s
total number of events: 18265
Latency (ms):
min: 1.08
avg: 1.09
max: 1.97
95th percentile: 1.10
sum: 19993.89
Threads fairness:
events (avg/stddev): 9132.5000/4.50
execution time (avg/stddev): 9.9969/0.00sysbench cpu --threads=4 run
sysbench 1.0.20 (using system LuaJIT 2.1.0-beta3)
Running the test with following options:
Number of threads: 4
Initializing random number generator from current time
Prime numbers limit: 10000
Initializing worker threads...
Threads started!
CPU speed:
events per second: 1825.26
General statistics:
total time: 10.0009s
total number of events: 18257
Latency (ms):
min: 1.08
avg: 2.19
max: 314.25
95th percentile: 1.12
sum: 39953.39
Threads fairness:
events (avg/stddev): 4564.2500/134.11
execution time (avg/stddev): 9.9883/0.01可以看到 2 线程对 events 总数有提升，但是再增加就没有效果。
内存
sysbench memory --threads=2 run
sysbench 1.0.20 (using system LuaJIT 2.1.0-beta3)
Running the test with following options:
Number of threads: 2
Initializing random number generator from current time
Running memory speed test with the following options:
block size: 1KiB
total size: 102400MiB
operation: write
scope: global
Initializing worker threads...
Threads started!
Total operations: 76222984 (7620841.73 per second)
74436.51 MiB transferred (7442.23 MiB/sec)
General statistics:
total time: 10.0001s
total number of events: 76222984
Latency (ms):
min: 0.00
avg: 0.00
max: 0.34
95th percentile: 0.00
sum: 11625.03
Threads fairness:
events (avg/stddev): 38111492.0000/14938.00
execution time (avg/stddev): 5.8125/0.00硬盘
sysbench fileio --threads=2 --file-test-mode=seqrewr run
sysbench 1.0.20 (using system LuaJIT 2.1.0-beta3)
Running the test with following options:
Number of threads: 2
Initializing random number generator from current time
Extra file open flags: (none)
128 files, 16MiB each
2GiB total file size
Block size 16KiB
Periodic FSYNC enabled, calling fsync() each 100 requests.
Calling fsync() at the end of test, Enabled.
Using synchronous I/O mode
Doing sequential rewrite test
Initializing worker threads...
Threads started!
File operations:
reads/s: 0.00
writes/s: 6602.37
fsyncs/s: 8470.32
Throughput:
read, MiB/s: 0.00
written, MiB/s: 103.16
General statistics:
total time: 10.0101s
total number of events: 150645
Latency (ms):
min: 0.01
avg: 0.13
max: 38.04
95th percentile: 0.07
sum: 19943.57
Threads fairness:
events (avg/stddev): 75322.5000/2302.50
execution time (avg/stddev): 9.9718/0.00总结
对比测试数据可以看出，即使是老款 Mac 表现依然出色，性能全面领先，DDR3 内存竟然不输给台式机的 DDR5，苹果的产品稳定、可靠、耐用。
入门级云服务器 阿里云 ECS t5.small 性能羸弱不出所料，但是搭载 AMD 最新旗舰处理器 R9 7900 与 PCIe4.0 7450MB/s 固态硬盘 的台式机，硬盘性能竟然远远不如入门级 ECS 的非固态硬盘，怎么也说不过去吧？其各项数据可能都受到这块硬盘的拖累。
使用 CrystalDiskMark 对台式机测试，结果竟然非常优秀，与 sysbench 截然相反，但是日常使用经常卡顿，文件管理器都打不开，读写速度几百 KB，让人无法质疑 sysbench 的测试结果。
这奇怪的差异到底是怎么造成的呢？无论如何，梵想的硬盘卡到不能做系统盘，只能等双十一看看有没有优惠，入手致态 TiPlus 7100 再测试一遍。
最后更新于