本帖最后由 abudu 于 2022-4-30 23:21 编辑
标准型 SA3 实例采用最新一代 AMD EPYC™ 处理器的实例,提供多种规格,具有超高性价比。此实例提供了平衡的计算、存储、网络资源,是众多应用程序的最佳选择。
其实例采用 AMD EPYC™ Milan 全新处理器,内存采用最新 DDR4,默认网络优化,最高内网收发能力达1900万pps,最高内网带宽可支持100Gbps。
本次测试的服务器为SA3.MEDIUM4,vCPU及内存为2C4G,未安装BBR等加速内核,完整配置见腾讯云官网的 产品介绍页
Bench.sh 测试
Bench.sh 可以查看 Linux 系统信息,还可以测试网络带宽及硬盘读写速率 由于是国内服务器,所以不对网络进行测试了,这里放出系统信息和硬盘速率
- ——————– A Bench.sh Script By Teddysun ——————-
- Version : v2022-02-22
- Usage : wget -qO- bench.sh | bash
- ———————————————————————-
- CPU Model : AMD EPYC 7K83 64-Core Processor
- CPU Cores : 2 @ 2545.218 MHz
- CPU Cache : 512 KB
- AES-NI : Enabled
- VM-x/AMD-V : Disabled
- Total Disk : 78.7 GB (2.1 GB Used)
- Total Mem : 3.5 GB (168.2 MB Used)
- System uptime : 0 days, 2 hour 44 min
- Load average : 0.00, 0.00, 0.06
- OS : Debian GNU/Linux 11
- Arch : x86_64 (64 Bit)
- Kernel : 5.10.0-12-amd64
- TCP CC : cubic
- Virtualization : KVM
- Organization : AS45090 Shenzhen Tencent Computer Systems Company Limited
- Location : Beijing / CN
- Region : Beijing
- ———————————————————————-
- I/O Speed(1st run) : 301 MB/s
- I/O Speed(2nd run) : 278 MB/s
- I/O Speed(3rd run) : 278 MB/s
- I/O Speed(average) : 285.7 MB/s
- ———————————————————————-
复制代码
其SSD速度达到了280MB/s,为普通的SSD水品 LemonBench
限于篇幅,此处只放出CPU&内存&硬盘的测试结果,完整结果见 https://pastebin.ubuntu.com/p/Km9VKwyXjF/plain/
- -> CPU Performance Test (Standard Mode, 3-Pass @ 30sec)
-
- 1 Thread Test: 3671 Scores
- 2 Threads Test: 4152 Scores
-
- -> Memory Performance Test (Standard Mode, 3-Pass @ 30sec)
-
- 1 Thread – Read Test : 46248.03 MB/s
- 1 Thread – Write Test: 26844.49 MB/s
-
- -> Disk Speed Test (4K Block/1M Block, Direct-Write)
-
- Test Name Write Speed Read Speed
- 10MB-4K Block 10.0 MB/s (2431 IOPS, 1.05 s) 20.6 MB/s (5022 IOPS, 0.51 s)
- 10MB-1M Block 135 MB/s (128 IOPS, 0.08 s) 264 MB/s (251 IOPS, 0.04 s)
- 100MB-4K Block 10.0 MB/s (2446 IOPS, 10.47 s) 20.4 MB/s (4971 IOPS, 5.15 s)
- 100MB-1M Block 180 MB/s (171 IOPS, 0.58 s) 385 MB/s (367 IOPS, 0.27 s)
- 1GB-4K Block 9.9 MB/s (2427 IOPS, 105.45 s) 16.1 MB/s (3921 IOPS, 65.29 s)
- 1GB-1M Block 200 MB/s (191 IOPS, 5.23 s) 315 MB/s (300 IOPS, 3.33 s)
复制代码
CPU(EPYC 7K83)单核分数达到了3671分,对比曾经测试的硅谷轻量(Gold 6133 CPU)的905分和成都轻量(EPYC 7K62)的1304分有了不小的提升,可见这颗"芯"还是不错的
内网测试
既然提到了内网,那不测一下怎么行,简单跑了下iperf3
- Connecting to host 172.16.xxx.xxx, port 5201
- [ 5] local 172.16.xxx.xxx port 44176 connected to 172.16.xxx.xxx port 5201
- [ ID] Interval Transfer Bitrate Retr Cwnd
- [ 5] 0.00-1.00 sec 307 MBytes 2.57 Gbits/sec 3144 131 KBytes
- [ 5] 1.00-2.00 sec 200 MBytes 1.68 Gbits/sec 1378 48.3 KBytes
- [ 5] 2.00-3.00 sec 177 MBytes 1.48 Gbits/sec 1795 68.9 KBytes
- [ 5] 3.00-4.00 sec 167 MBytes 1.41 Gbits/sec 790 75.8 KBytes
- [ 5] 4.00-5.00 sec 219 MBytes 1.83 Gbits/sec 1926 97.9 KBytes
- [ 5] 5.00-6.00 sec 154 MBytes 1.30 Gbits/sec 2055 60.7 KBytes
- [ 5] 6.00-7.00 sec 213 MBytes 1.79 Gbits/sec 2433 97.9 KBytes
- [ 5] 7.00-8.00 sec 155 MBytes 1.30 Gbits/sec 1052 262 KBytes
- [ 5] 8.00-9.00 sec 211 MBytes 1.77 Gbits/sec 1478 81.4 KBytes
- [ 5] 9.00-10.00 sec 158 MBytes 1.32 Gbits/sec 2225 24.8 KBytes
-
- —
-
- [ ID] Interval Transfer Bitrate Retr
- [ 5] 0.00-10.00 sec 1.92 GBytes 1.65 Gbits/sec 18276 sender
- [ 5] 0.00-10.00 sec 1.91 GBytes 1.64 Gbits/sec receiver
-
- iperf Done.
复制代码
达到了1.5~2.0Gbits/sec 符合预期水平
UnixBench
UnixBench 是一款开源的测试 Unix 系统基本性能的工具,是比较通用的测试VPS性能的工具 完整数据见 https://pastebin.ubuntu.com/p/fb8s9V5wpM/plain/
单任务:
- Benchmark Run: Fri Apr 29 2022 22:53:07 – 23:21:05
- 2 CPUs in system; running 1 parallel copy of tests
-
- Dhrystone 2 using register variables 44734686.7 lps (10.0 s, 7 samples)
- Double-Precision Whetstone 7618.1 MWIPS (9.9 s, 7 samples)
- Execl Throughput 5861.2 lps (30.0 s, 2 samples)
- File Copy 1024 bufsize 2000 maxblocks 1217379.1 KBps (30.0 s, 2 samples)
- File Copy 256 bufsize 500 maxblocks 318649.6 KBps (30.0 s, 2 samples)
- File Copy 4096 bufsize 8000 maxblocks 3650531.6 KBps (30.0 s, 2 samples)
- Pipe Throughput 2405640.6 lps (10.0 s, 7 samples)
- Pipe-based Context Switching 50809.9 lps (10.0 s, 7 samples)
- Process Creation 11187.2 lps (30.0 s, 2 samples)
- Shell Scripts (1 concurrent) 12000.6 lpm (60.0 s, 2 samples)
- Shell Scripts (8 concurrent) 1855.0 lpm (60.0 s, 2 samples)
- System Call Overhead 2576839.6 lps (10.0 s, 7 samples)
-
- System Benchmarks Index Values BASELINE RESULT INDEX
- Dhrystone 2 using register variables 116700.0 44734686.7 3833.3
- Double-Precision Whetstone 55.0 7618.1 1385.1
- Execl Throughput 43.0 5861.2 1363.1
- File Copy 1024 bufsize 2000 maxblocks 3960.0 1217379.1 3074.2
- File Copy 256 bufsize 500 maxblocks 1655.0 318649.6 1925.4
- File Copy 4096 bufsize 8000 maxblocks 5800.0 3650531.6 6294.0
- Pipe Throughput 12440.0 2405640.6 1933.8
- Pipe-based Context Switching 4000.0 50809.9 127.0
- Process Creation 126.0 11187.2 887.9
- Shell Scripts (1 concurrent) 42.4 12000.6 2830.3
- Shell Scripts (8 concurrent) 6.0 1855.0 3091.7
- System Call Overhead 15000.0 2576839.6 1717.9
- ========
- System Benchmarks Index Score 1760.1
复制代码
多任务:
- Benchmark Run: Fri Apr 29 2022 23:21:05 – 23:49:06
- 2 CPUs in system; running 2 parallel copies of tests
-
- Dhrystone 2 using register variables 68864489.9 lps (10.0 s, 7 samples)
- Double-Precision Whetstone 13038.2 MWIPS (9.9 s, 7 samples)
- Execl Throughput 7893.6 lps (29.9 s, 2 samples)
- File Copy 1024 bufsize 2000 maxblocks 1514472.5 KBps (30.0 s, 2 samples)
- File Copy 256 bufsize 500 maxblocks 398094.6 KBps (30.0 s, 2 samples)
- File Copy 4096 bufsize 8000 maxblocks 4636873.7 KBps (30.0 s, 2 samples)
- Pipe Throughput 3149205.4 lps (10.0 s, 7 samples)
- Pipe-based Context Switching 439878.2 lps (10.0 s, 7 samples)
- Process Creation 19942.3 lps (30.0 s, 2 samples)
- Shell Scripts (1 concurrent) 14120.6 lpm (60.0 s, 2 samples)
- Shell Scripts (8 concurrent) 1881.0 lpm (60.0 s, 2 samples)
- System Call Overhead 3376550.5 lps (10.0 s, 7 samples)
-
- System Benchmarks Index Values BASELINE RESULT INDEX
- Dhrystone 2 using register variables 116700.0 68864489.9 5901.0
- Double-Precision Whetstone 55.0 13038.2 2370.6
- Execl Throughput 43.0 7893.6 1835.7
- File Copy 1024 bufsize 2000 maxblocks 3960.0 1514472.5 3824.4
- File Copy 256 bufsize 500 maxblocks 1655.0 398094.6 2405.4
- File Copy 4096 bufsize 8000 maxblocks 5800.0 4636873.7 7994.6
- Pipe Throughput 12440.0 3149205.4 2531.5
- Pipe-based Context Switching 4000.0 439878.2 1099.7
- Process Creation 126.0 19942.3 1582.7
- Shell Scripts (1 concurrent) 42.4 14120.6 3330.3
- Shell Scripts (8 concurrent) 6.0 1881.0 3135.0
- System Call Overhead 15000.0 3376550.5 2251.0
- ========
- System Benchmarks Index Score 2760.2
复制代码
作为对比,上海地区轻量的得分为单核1478.6,多核2666.8,SA3依然更胜一筹
总结
SA3 作为采用最新一代 AMD EPYC™ 处理器的实例,绝不仅仅是一个“大号轻量”,对比轻量应用服务器更能胜任需要运算性能的环境,但相应的价格也会更高。针对不同的使用环境来选择依然是亘古不变的道理。 如果是你,你会选择SA3还是其他厂家的产品呢
|