citywar 大佬有话说 :
contabo vps 新加坡机房 7.49欧的4c8g 50G NVMe硬盘 实测
contabo最近不是上线了 NVMe硬盘。说是硬盘读写提升了很多。
https://p0.meituan.net/dpgroup/1a8c7b5d5acb1558111f5c0aa48fc65825860.png
传送门： contabo官网 vps 配置与价格页面
忍不住想试试NVMe看看效果：
实测 IO speed平均 800 ~ 1100MB/s （相比SSD硬盘限制 IO Speed为 100MB/s 舒服很多）
https://p0.meituan.net/dpgroup/795b303fc2243899337c014f3c33235b33361.png
4核8G 那款VPS 选SSD硬盘是200GB如果选 NVMe硬盘就是只有 50 GB
https://p1.meituan.net/dpgroup/9739d0b01637360aa5140efb7a7b982635095.png
-> CPU Performance Test (Standard Mode, 3-Pass @ 30sec)
1 Thread Test: 1552 Scores
4 Threads Test: 6210 Scores
-> Memory Performance Test (Standard Mode, 3-Pass @ 30sec)
1 Thread – Read Test : 39604.02 MB/s
1 Thread – Write Test: 17064.48 MB/s
-> Disk Speed Test (4K Block/1M Block, Direct-Write)
Test Name Write Speed Read Speed
10MB-4K Block 46.4 MB/s (0.09 IOPS, 0.23s)) 69.3 MB/s (16912 IOPS, 0.15s)
10MB-1M Block 614 MB/s (585 IOPS, 0.02s) 1.0 GB/s (997 IOPS, 0.01s)
100MB-4K Block 56.1 MB/s (0.07 IOPS, 1.87s)) 95.2 MB/s (23250 IOPS, 1.10s)
100MB-1M Block 759 MB/s (723 IOPS, 0.14s) 2.1 GB/s (2016 IOPS, 0.05s)
1GB-4K Block 22.7 MB/s (5539 IOPS, 46.21s) 54.8 MB/s (13389 IOPS, 19.12s)
1GB-1M Block 1.0 GB/s (957 IOPS, 1.04s) 1.7 GB/s (1630 IOPS, 0.61s)
德国 和 西雅图 的用过呢，我就是试试新加坡机房的，
如果没反代机器不推荐购买这个机房。德国机房是最便宜的（套cf玩的直接德国最实惠）。西雅图机房直连的可能到国内延迟低一些
德国纽伦堡IP：161.97.87.1
美国西雅图IP：144.126.144.1
新加坡测试IP : 194.233.67.1
新加坡机房的 延迟挺高的：
CHINAZ PING
线路 最快节点 最慢节点 平均响应
电信 江苏泰州311ms 新疆哈密375ms 354.4ms
多线 浙江扬州309ms 北京407ms 357.8ms
联通 山东枣庄333ms 山东枣庄435ms 374.3ms
移动 安徽合肥430ms 安徽合肥430ms 430ms
海外 新加坡1ms 中国香港322ms 151.1ms
新加坡机房路由的话：电信联通绕美国日本，有的还是绕美国日本后再香港到新加坡。移动全球跑。
去程he.net 回程he.nettelia
但是到香港的延迟非常低 ， 之前有MJJ 用腾讯云香港拉 contabo新加坡 说是效果很好
看了一下 电信cn2 80~100ms左右：
0.CN 117.28.242.0 Loss%Snt LastAvg BestWrstStDev AS Name PTR
1 ??? 100.020 0.0 0.0 0.0 0.0 0.0 –
2 CN 117.28.242.129 0.0% 20 3.9 3.3 2.3 6.5 0.9 4809CN CHINATELECOM-CORE-WAN-CN2 117.28.242.129
3 192.168.1.25 0.0% 20 2.9 3.5 2.7 6.2 0.9 –
4 10.1.0.66 0.0% 20 1.3 1.7 1.2 4.2 0.6 –
5 10.1.0.65 0.0% 20 2.1 14.80.9 131.3 33.6-
6 CN 222.77.78.50 0.0% 20 1.1 1.7 1.0 7.6 1.4 4134CN CHINANET-BACKBONE 222.77.78.50
7 CN 59.43.74.57 0.0% 20 36.329.11.4 50.916.74809CN CHINATELECOM-CN2 59.43.74.57
8 CN 59.43.156.177 0.0% 20 4.3 4.7 4.0 9.1 1.1 4809CN CHINATELECOM-CN2 59.43.156.177
9 CN 59.43.46.213 0.0% 20 18.117.917.618.90.2 4809CN CHINATELECOM-CN2 59.43.46.213
10CN 59.43.130.146 75.0%20 18.418.618.419.30.0 4809CN CHINATELECOM-CN2 59.43.130.146
11CN 59.43.187.94 0.0% 20 19.019.318.821.70.6 4809CN CHINATELECOM-CN2 59.43.187.94
12CN 59.43.184.122 0.0% 20 56.156.456.157.00.0 4809CN CHINATELECOM-CN2 59.43.184.122
13SG 210.57.30.21 0.0% 20 61.161.560.470.92.2 4637HK TELSTRA unknown.telstraglobal.net
14US 216.218.221.1410.0% 20 58.163.257.279.48.5 6939US HURRICANE v131.core1.sin1.he.net
15US 74.82.46.230 0.0% 20 58.058.157.858.90.0 6939US HURRICANE contabo-asia-private-limited.e
16DE 194.233.67.XX 0.0% 20 94.796.383.9103.9 5.6 141995 SG CAPL vmi656312.contaboserver.net
https://p1.meituan.net/dpgroup/0825acd06939f2fd0cdda82a89803b2699366.png
https://p0.meituan.net/dpgroup/ddafe0d491c5c511c0d9bb829d35442d16443.png
unixbencha跑分，单核 1500分 4核3100分左右刚刚测的。
看来新加坡机房用户少一些。 我西雅图机房 4核好像是 2700分。
Benchmark Run: Tue Aug 31 2021 20:50:12 – 21:19:16
4 CPUs in system; running 1 parallel copy of tests
Dhrystone 2 using register variables 35070771.6 lps (10.0 s, 7 samples)
Double-Precision Whetstone 3339.7 MWIPS (16.2 s, 7 samples)
Execl Throughput 4780.6 lps (30.0 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks 956326.4 KBps(30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks 253835.6 KBps(30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks 2051107.5 KBps(30.0 s, 2 samples)
Pipe Throughput 1727885.3 lps (10.0 s, 7 samples)
Pipe-based Context Switching 228128.8 lps (10.0 s, 7 samples)
Process Creation 8050.5 lps (30.0 s, 2 samples)
Shell Scripts (1 concurrent) 7246.8 lpm (60.0 s, 2 samples)
Shell Scripts (8 concurrent) 2387.6 lpm (60.0 s, 2 samples)
System Call Overhead 2156497.4 lps (10.0 s, 7 samples)
System Benchmarks Index Values BASELINE RESULT INDEX
Dhrystone 2 using register variables 116700.0 35070771.6 3005.2
Double-Precision Whetstone 55.0 3339.7 607.2
Execl Throughput 43.0 4780.6 1111.8
File Copy 1024 bufsize 2000 maxblocks 3960.0 956326.4 2415.0
File Copy 256 bufsize 500 maxblocks 1655.0 253835.6 1533.7
File Copy 4096 bufsize 8000 maxblocks 5800.0 2051107.5 3536.4
Pipe Throughput 12440.0 1727885.3 1389.0
Pipe-based Context Switching 4000.0 228128.8 570.3
Process Creation 126.0 8050.5 638.9
Shell Scripts (1 concurrent) 42.4 7246.8 1709.1
Shell Scripts (8 concurrent) 6.0 2387.6 3979.4
System Call Overhead 15000.0 2156497.4 1437.7
========
System Benchmarks Index Score 1501.7
————————————————————————
Benchmark Run: Tue Aug 31 2021 21:19:16 – 21:48:28
4 CPUs in system; running 4 parallel copies of tests
Dhrystone 2 using register variables 140213897.9 lps (10.0 s, 7 samples)
Double-Precision Whetstone 13448.9 MWIPS (16.1 s, 7 samples)
Execl Throughput 13015.7 lps (30.0 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks 693622.1 KBps(30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks 156764.9 KBps(30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks 2030795.9 KBps(30.0 s, 2 samples)
Pipe Throughput 6872537.1 lps (10.0 s, 7 samples)
Pipe-based Context Switching 739827.7 lps (10.0 s, 7 samples)
Process Creation 33973.4 lps (30.0 s, 2 samples)
Shell Scripts (1 concurrent) 19674.9 lpm (60.0 s, 2 samples)
Shell Scripts (8 concurrent) 2745.3 lpm (60.0 s, 2 samples)
System Call Overhead 3963391.0 lps (10.0 s, 7 samples)
System Benchmarks Index Values BASELINE RESULT INDEX
Dhrystone 2 using register variables 116700.0140213897.912014.9
Double-Precision Whetstone 55.0 13448.9 2445.2
Execl Throughput 43.0 13015.7 3026.9
File Copy 1024 bufsize 2000 maxblocks 3960.0 693622.1 1751.6
File Copy 256 bufsize 500 maxblocks 1655.0 156764.9 947.2
File Copy 4096 bufsize 8000 maxblocks 5800.0 2030795.9 3501.4
Pipe Throughput 12440.0 6872537.1 5524.5
Pipe-based Context Switching 4000.0 739827.7 1849.6
Process Creation 126.0 33973.4 2696.3
Shell Scripts (1 concurrent) 42.4 19674.9 4640.3
Shell Scripts (8 concurrent) 6.0 2745.3 4575.5
System Call Overhead 15000.0 3963391.0 2642.3
========
System Benchmarks Index Score 3103.5
另外contabo 注册如果 假资料开T子 会人工审核，要身份证明。最好不要全乱填。
默认打开是美元价格，
像4核8G 50G这款德国是 6.99美元改成欧元是4.99欧 Paypal付款6.2刀左右
签名第一个链接 有改成 欧元价格付款的说明。
citywar 大佬有话说 :
补一个国外的 curl -sL yabs.sh | bash
https://p1.meituan.net/dpgroup/3be2dbf55891e2dc8ce8a2a614dc7a3c18019.png
Tue Aug 31 22:16:41 +08 2021
Basic System Information:
———————————
Processor: AMD EPYC 7282 16-Core Processor
CPU cores: 4 @ 2794.748 MHz
AES-NI : ✔ Enabled
VM-x/AMD-V : ❌ Disabled
RAM : 7.8 GiB
Swap : 0.0 KiB
Disk : 49.1 GiB
fio Disk Speed Tests (Mixed R/W 50/50):
———————————
Block Size | 4k (IOPS) | 64k (IOPS)
—— | — —-| —- —-
Read | 23.31 MB/s (5.8k) | 331.37 MB/s (5.1k)
Write | 23.32 MB/s (5.8k) | 333.12 MB/s (5.2k)
Total | 46.64 MB/s (11.6k) | 664.49 MB/s(10.3k)
| |
Block Size | 512k (IOPS) | 1m (IOPS)
—— | — —-| —- —-
Read | 3.60 GB/s (7.0k) | 8.08 GB/s (7.8k)
Write | 3.80 GB/s (7.4k) | 8.62 GB/s (8.4k)
Total | 7.41 GB/s (14.4k) | 16.71 GB/s (16.3k)
optimism 大佬有话说 :
写的很好必顶
webjin 大佬有话说 :
不差啊
Faxlok 大佬有话说 :
这性能真不错https://cdn.jsdelivr.net/gh/hishis/forum-master/public/images/patch.gif
louiejordan 大佬有话说 :
他家新加坡如果没有特别需要还是买德国的好，至少不绕路
冲浪麦浪花郎 大佬有话说 :
还是美区和德区香https://cdn.jsdelivr.net/gh/master-of-forums/master-of-forums/public/images/patch.gif
紫玉成烟 大佬有话说 :
这测评详细的很。yc022t
16qf 大佬有话说 :
雀食不错，可惜上车netcup了
6fyC8r 大佬有话说 :
大部分商家的新加坡现在都绕路，