DMIT圣诞新品sPro套餐测评

DMIT 发布了圣诞促销新Offer,洛杉矶sPro套餐开售。该套餐去程会走Cloudflare Magic Transit,回程走GIA CN2,既能提供高速低延迟,也同时带来CF的高防服务(5T防御)。sPro套餐默认配置为2核,2G内存,20G SSD硬盘,1T流量@100M带宽。使用优惠码后,月付$11.92美元。

本人觉得该套餐性价比不错,因此购买了一款,本文是对该产品的简评。

DMIT圣诞新品sPro套餐测评

首先是基本信息:

----------------------------------------------------------------------
CPU Model             : AMD EPYC 7402P 24-Core Processor
CPU Cores             : 2
CPU Frequency         : 2794.704 MHz
CPU Cache             : 512 KB
Total Disk            : 20.0 GB (2.0 GB Used)
Total Mem             : 1997 MB (65 MB Used)
Total Swap            : 0 MB (0 MB Used)
System uptime         : 0 days, 0 hour 8 min
Load average          : 0.00, 0.11, 0.08
OS                    : CentOS 7.9.2009
Arch                  : x86_64 (64 Bit)
Kernel                : 4.9.215-36.el7.x86_64
TCP CC                : bbr
Virtualization        : KVM
Organization          : AS54574 DMIT Cloud Services
Location              : Los Angeles / US
Region                : California
----------------------------------------------------------------------
I/O Speed(1st run)    : 860 MB/s
I/O Speed(2nd run)    : 882 MB/s
I/O Speed(3rd run)    : 817 MB/s
Average I/O speed     : 853.0 MB/s
----------------------------------------------------------------------
Node Name        Upload Speed      Download Speed      Latency     
Speedtest.net    100.86 Mbps       101.37 Mbps         157.10 ms   
Beijing    CU    4.79 Mbps         97.86 Mbps          338.83 ms   
Shanghai   CT    102.06 Mbps       103.53 Mbps         127.52 ms   
Shanghai   CU    58.08 Mbps        103.96 Mbps         245.13 ms   
Guangzhou  CT    35.93 Mbps        86.35 Mbps          170.05 ms   
Guangzhou  CU    1.20 Mbps         92.29 Mbps          349.77 ms   
Shenzhen   CU    19.49 Mbps        101.12 Mbps         237.88 ms   
Hongkong   CN    104.31 Mbps       100.30 Mbps         165.94 ms   
Singapore  SG    101.78 Mbps       105.22 Mbps         173.36 ms   
Tokyo      JP    104.98 Mbps       8.46 Mbps           115.86 ms   
----------------------------------------------------------------------

套餐使用AMD的CPU,可能是现在邻居较少,硬盘IO很不错。下载和上行数据来看,移动电信都不错,联通比较拉胯。

接着看全国ping数据,全国平均延迟174ms,比他家其他GIA产品要低点。需要注意的是ipip.net的ip库未更新,ip会被定位到香港。

DMIT spro套餐ping数据
DMIT spro套餐ping数据

再来看回程路由:

北京电信
traceroute to 220.181.22.1 (220.181.22.1), 30 hops max, 60 byte packets
1  premium-routing.irb.100.gw0.lax.DMIT.io (154.17.0.1)  11.74 ms  AS174,AS54574  United States California Los Angeles cogentco.com
2  premium-routing.ae10.1000.re0.lax.DMIT.io (193.41.248.225)  0.45 ms  AS54574  United States California Los Angeles dmit.io
3  china-telecom-gia.re0.lax.DMIT.io (193.41.248.130)  0.94 ms  AS54574  United States California Los Angeles dmit.io
4  59.43.189.37  126.55 ms  *  China Shanghai ChinaTelecom
5  59.43.247.141  155.17 ms  *  China Beijing ChinaTelecom
6  59.43.188.73  148.74 ms  *  China Beijing ChinaTelecom
7  59.43.131.253  150.03 ms  *  China Beijing ChinaTelecom
8  bj141-152-146.bjtelecom.net (219.141.152.146)  152.57 ms  AS4847  China Beijing ChinaTelecom
9  *
10  *
11  220.181.22.1  179.62 ms  AS23724  China Beijing ChinaTelecom

----------------------------------------------------------------------
北京移动
traceroute to 221.130.33.1 (221.130.33.1), 30 hops max, 60 byte packets
1  premium-routing.irb.100.gw0.lax.DMIT.io (154.17.0.1)  13.81 ms  AS174,AS54574  United States California Los Angeles cogentco.com
2  premium-routing.ae10.1000.re0.lax.DMIT.io (193.41.248.225)  2.65 ms  AS54574  United States California Los Angeles dmit.io
3  china-telecom-gia.re0.lax.DMIT.io (193.41.248.130)  0.72 ms  AS54574  United States California Los Angeles dmit.io
4  59.43.181.145  125.94 ms  *  China Shanghai ChinaTelecom
5  *
6  *
7  59.43.80.102  128.99 ms  *  China Shanghai ChinaTelecom
8  *
9  *
10  221.183.30.49  206.24 ms  AS9808  China Beijing ChinaMobile
11  221.183.9.197  169.65 ms  AS9808  China Beijing ChinaMobile
12  111.24.2.97  170.17 ms  AS9808  China Beijing ChinaMobile
13  111.24.14.46  169.34 ms  AS9808  China Beijing ChinaMobile
14  *
15  211.136.66.129  167.84 ms  AS56048  China Beijing ChinaMobile
16  221.179.159.238  171.30 ms  AS56048  China Beijing ChinaMobile
17  221.130.33.1  168.46 ms  AS56048  China Beijing ChinaMobile

----------------------------------------------------------------------
宁波联通
traceroute to 60.12.17.1 (60.12.17.1), 30 hops max, 60 byte packets
1  premium-routing.irb.100.gw0.lax.DMIT.io (154.17.0.1)  82.93 ms  AS174,AS54574  United States California Los Angeles cogentco.com
2  premium-routing.ae10.1000.re0.lax.DMIT.io (193.41.248.225)  0.32 ms  AS54574  United States California Los Angeles dmit.io
3  china-telecom-gia.re0.lax.DMIT.io (193.41.248.130)  0.92 ms  AS54574  United States California Los Angeles dmit.io
4  59.43.184.153  126.01 ms  *  China Shanghai ChinaTelecom
5  *
6  59.43.138.49  134.04 ms  *  China Shanghai ChinaTelecom
7  *
8  219.158.113.197  160.40 ms  AS4837  China Shanghai ChinaUnicom
9  219.158.106.94  168.68 ms  AS4837  China Zhejiang Hangzhou ChinaUnicom
10  221.12.35.126  161.55 ms  AS4837  China Zhejiang Ningbo ChinaUnicom
11  *
12  60.12.17.1  161.01 ms  AS4837  China Zhejiang Ningbo ChinaUnicom

----------------------------------------------------------------------
杭州移动
traceroute to 211.140.0.2 (211.140.0.2), 30 hops max, 60 byte packets
1  premium-routing.irb.100.gw0.lax.DMIT.io (154.17.0.1)  18.51 ms  AS174,AS54574  United States California Los Angeles cogentco.com
2  premium-routing.ae10.1000.re0.lax.DMIT.io (193.41.248.225)  0.34 ms  AS54574  United States California Los Angeles dmit.io
3  china-telecom-gia.re0.lax.DMIT.io (193.41.248.130)  1.01 ms  AS54574  United States California Los Angeles dmit.io
4  59.43.189.37  126.64 ms  *  China Shanghai ChinaTelecom
5  59.43.187.65  135.08 ms  *  China Shanghai ChinaTelecom
6  59.43.138.49  134.32 ms  *  China Shanghai ChinaTelecom
7  59.43.80.138  131.50 ms  *  China Shanghai ChinaTelecom
8  202.97.68.138  132.48 ms  AS13030  China Zhejiang Hangzhou ChinaTelecom
9  *
10  221.183.54.37  199.87 ms  AS9808  China Zhejiang Hangzhou ChinaMobile
11  221.183.25.145  173.14 ms  AS9808  China Zhejiang Hangzhou ChinaMobile
12  211.140.0.2  174.07 ms  AS56041  China Zhejiang Hangzhou ChinaMobile

----------------------------------------------------------------------
广州电信
traceroute to 14.215.116.1 (14.215.116.1), 30 hops max, 60 byte packets
1  premium-routing.irb.100.gw0.lax.DMIT.io (154.17.0.1)  17.33 ms  AS174,AS54574  United States California Los Angeles cogentco.com
2  premium-routing.ae10.1000.re0.lax.DMIT.io (193.41.248.225)  0.32 ms  AS54574  United States California Los Angeles dmit.io
3  china-telecom-gia.re0.lax.DMIT.io (193.41.248.130)  0.78 ms  AS54574  United States California Los Angeles dmit.io
4  59.43.182.106  160.07 ms  *  China Guangdong Guangzhou ChinaTelecom
5  *
6  59.43.130.121  162.70 ms  *  China Guangdong Guangzhou ChinaTelecom
7  202.97.82.38  162.82 ms  AS4134  China Guangdong Guangzhou ChinaTelecom
8  113.108.208.26  162.78 ms  AS58466  China Guangdong Guangzhou ChinaTelecom
9  183.2.167.229  163.51 ms  AS58466  China Guangdong Guangzhou ChinaTelecom
10  183.56.128.10  157.88 ms  AS58466  China Guangdong Guangzhou ChinaTelecom
11  14.215.116.1  162.80 ms  AS58466  China Guangdong Guangzhou ChinaTelecom

----------------------------------------------------------------------
深圳联通
traceroute to 58.250.0.1 (58.250.0.1), 30 hops max, 60 byte packets
1  premium-routing.irb.100.gw0.lax.DMIT.io (154.17.0.1)  12.97 ms  AS174,AS54574  United States California Los Angeles cogentco.com
2  premium-routing.ae10.1000.re0.lax.DMIT.io (193.41.248.225)  0.33 ms  AS54574  United States California Los Angeles dmit.io
3  china-telecom-gia.re0.lax.DMIT.io (193.41.248.130)  0.78 ms  AS54574  United States California Los Angeles dmit.io
4  59.43.182.106  160.04 ms  *  China Guangdong Guangzhou ChinaTelecom
5  *
6  59.43.130.149  165.51 ms  *  China Guangdong Guangzhou ChinaTelecom
7  *
8  219.158.9.73  161.05 ms  AS4837  China Guangdong Guangzhou ChinaUnicom
9  221.4.0.126  167.50 ms  AS17816  China Guangdong Shenzhen ChinaUnicom
10  58.250.0.1  166.85 ms  AS17623  China Guangdong Shenzhen ChinaUnicom
----------------------------------------------------------------------

回程追踪三网都走了cn2 GIA线路,晚高峰表现有保障。

最后看UnixBench跑分:

========================================================================
   BYTE UNIX Benchmarks (Version 5.1.3)

   System: localhost.localhost: GNU/Linux
   OS: GNU/Linux -- 4.9.215-36.el7.x86_64 -- #1 SMP Mon Mar 2 11:42:52 UTC 2020
   Machine: x86_64 (x86_64)
   Language: en_US.utf8 (charmap="UTF-8", collate="UTF-8")
   CPU 0: AMD EPYC 7402P 24-Core Processor (5589.4 bogomips)
          Hyper-Threading, x86-64, MMX, AMD MMX, Physical Address Ext, SYSENTER/SYSEXIT, AMD virtualization, SYSCALL/SYSRET
   CPU 1: AMD EPYC 7402P 24-Core Processor (5589.4 bogomips)
          Hyper-Threading, x86-64, MMX, AMD MMX, Physical Address Ext, SYSENTER/SYSEXIT, AMD virtualization, SYSCALL/SYSRET
   07:31:28 up 7 min,  1 user,  load average: 0.13, 0.16, 0.09; runlevel 3

------------------------------------------------------------------------
Benchmark Run: Sat Dec 26 2020 07:31:28 - 08:00:48
2 CPUs in system; running 1 parallel copy of tests

Dhrystone 2 using register variables       35872868.8 lps   (10.0 s, 7 samples)
Double-Precision Whetstone                     3136.8 MWIPS (17.8 s, 7 samples)
Execl Throughput                               3602.6 lps   (30.0 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks        794637.4 KBps  (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks          227902.1 KBps  (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks       2046130.5 KBps  (30.0 s, 2 samples)
Pipe Throughput                             1649601.2 lps   (10.0 s, 7 samples)
Pipe-based Context Switching                  52289.2 lps   (10.0 s, 7 samples)
Process Creation                               8560.3 lps   (30.0 s, 2 samples)
Shell Scripts (1 concurrent)                   7217.3 lpm   (60.0 s, 2 samples)
Shell Scripts (8 concurrent)                   1471.4 lpm   (60.0 s, 2 samples)
System Call Overhead                        1788028.2 lps   (10.0 s, 7 samples)

System Benchmarks Index Values               BASELINE       RESULT    INDEX
Dhrystone 2 using register variables         116700.0   35872868.8   3073.9
Double-Precision Whetstone                       55.0       3136.8    570.3
Execl Throughput                                 43.0       3602.6    837.8
File Copy 1024 bufsize 2000 maxblocks          3960.0     794637.4   2006.7
File Copy 256 bufsize 500 maxblocks            1655.0     227902.1   1377.1
File Copy 4096 bufsize 8000 maxblocks          5800.0    2046130.5   3527.8
Pipe Throughput                               12440.0    1649601.2   1326.0
Pipe-based Context Switching                   4000.0      52289.2    130.7
Process Creation                                126.0       8560.3    679.4
Shell Scripts (1 concurrent)                     42.4       7217.3   1702.2
Shell Scripts (8 concurrent)                      6.0       1471.4   2452.4
System Call Overhead                          15000.0    1788028.2   1192.0
                                                                   ========
System Benchmarks Index Score                                        1193.9

------------------------------------------------------------------------
Benchmark Run: Sat Dec 26 2020 08:00:48 - 08:30:10
2 CPUs in system; running 2 parallel copies of tests

Dhrystone 2 using register variables       70893585.2 lps   (10.0 s, 7 samples)
Double-Precision Whetstone                     6323.3 MWIPS (17.5 s, 7 samples)
Execl Throughput                               7197.6 lps   (29.9 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks        586011.6 KBps  (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks          161224.3 KBps  (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks       1849878.0 KBps  (30.0 s, 2 samples)
Pipe Throughput                             3258777.3 lps   (10.0 s, 7 samples)
Pipe-based Context Switching                 413909.6 lps   (10.0 s, 7 samples)
Process Creation                              17324.3 lps   (30.0 s, 2 samples)
Shell Scripts (1 concurrent)                   9969.8 lpm   (60.0 s, 2 samples)
Shell Scripts (8 concurrent)                   1330.3 lpm   (60.1 s, 2 samples)
System Call Overhead                        2733523.8 lps   (10.0 s, 7 samples)

System Benchmarks Index Values               BASELINE       RESULT    INDEX
Dhrystone 2 using register variables         116700.0   70893585.2   6074.9
Double-Precision Whetstone                       55.0       6323.3   1149.7
Execl Throughput                                 43.0       7197.6   1673.9
File Copy 1024 bufsize 2000 maxblocks          3960.0     586011.6   1479.8
File Copy 256 bufsize 500 maxblocks            1655.0     161224.3    974.2
File Copy 4096 bufsize 8000 maxblocks          5800.0    1849878.0   3189.4
Pipe Throughput                               12440.0    3258777.3   2619.6
Pipe-based Context Switching                   4000.0     413909.6   1034.8
Process Creation                                126.0      17324.3   1374.9
Shell Scripts (1 concurrent)                     42.4       9969.8   2351.4
Shell Scripts (8 concurrent)                      6.0       1330.3   2217.1
System Call Overhead                          15000.0    2733523.8   1822.3
                                                                   ========
System Benchmarks Index Score                                        1876.3

单核跑分1200,两核跑分接近1900,还是不错的。

总结

该款套餐性价比不错,适合建站,也推荐购买。但是目前许多人在进行测速,可能速度不是很理想。

发表回复

您的电子邮箱地址不会被公开。 必填项已用 * 标注