Home Experience & Tips Sharing Standard Post

Standard Post

VMRack VPS Review: $13/month Los Angeles 2-core 2GB three-network route real test, can it replace BandwagonHost?

In Naiba's VPS recommendation article, for machines with route requirements, BandwagonHost has always been the main recommendation. However, BandwagonHost is relatively expensive, and average users may not be able to afford it. Recently, Naiba came across VMRack, which claims to be directly supplied from its own data center in Los Angeles, offering CN2 GIA, AS9929, and CMIN2 premium routes for all three networks, with a 1-core…

Updated on October 28, 2026 Approximately 81 minutes read
VMRack VPS 评测:$13/月洛杉矶 2 核 2G 三网线路实测,能替代搬瓦工吗?

In Naiba'sVPS RecommendationsIn the article, for machines with route requirements, BandwagonHost has always been the main recommendation. However, BandwagonHost is relatively expensive, and average users may not be able to afford it. Recently, Naiba came across VMRack, which claims to be directly supplied from its own data center in Los Angeles, offering CN2 GIA, AS9929, and CMIN2 premium routes for all three networks, with a 1-core 1G machine costing only $11.5 per month. How are VMRack's machines actually, and can they serve as a cheap CN2 GIA route machine alternative to BandwagonHost? Let's actually test and see the results.

This test uses a custom configuration of VMRack with 2 cores and 2G, priced at $13/month. Prices, stock, routes, and package conditions may change; before purchasing, please refer toVMRack official websitethe checkout page for accuracy.

First, the conclusion

The advantages of this VMRack machine are clear: for $13/month, you get 2 vCPU, 2 GB RAM, 20 GB system disk, 500 Mbps bandwidth, and 1000 GB/month traffic. The international bandwidth test is basically close to 500 Mbps, and the 4K random read/write in the Rongheguai test also reaches about 25.5k IOPS.

In terms of network, the Rongheguai return route test shows that CN2-BackBone appears in the Telecom direction, AS9929/CNC-BACKBONE in the Unicom direction, and CMIN2-NET in the Mobile direction, so the routes do have some highlights. However, the latency of the three-network nearby speed test is about 267–350ms, and the download speed ranges from about 5Mbps to 165Mbps, so you cannot conclude that access is fast in all regions just by looking at the route names.

However, it cannot be called a „complete replacement for BandwagonHost“ based on this single test alone. The return route performance is good, but actual speed tests are greatly affected by test nodes, direction, and time period; the machine„s uptime is also not long, so it is impossible to judge long-term stability, evening peak congestion, after-sales service, and fault handling. A more accurate conclusion is:

If you need a Los Angeles node, value mainland China routes, and can accept maintaining the server yourself, this VMRack $13/month configuration is worth a monthly test; if it is for important business, you still need to first conduct continuous observation, backups, and prepare for fault migration.

Configuration and price of this VMRack VPS

This test uses a custom configuration from the VMRack official website, with the service type as cloud server and the region selected as Los Angeles. You can also first visitVMRack official websiteto view current packages and prices.

ItemConfiguration
CPU2 vCPU
Memory2 GB
System disk20 GB
Bandwidth typeThree-network premium
Bandwidth500 Mbps
Monthly traffic1000 GB/month
IPv41 basic IPv4
OSDebian 13 x64
Price$13/month

This configuration is not positioned for high-concurrency websites or large applications, but is more suitable for low-traffic WordPress, personal blogs, foreign trade showcase sites, development and testing environments, regular network tools other than proxies, and small services.

2 GB RAM running Debian, Nginx, PHP, MariaDB, and a low-traffic WordPress is usually sufficient, but it is not recommended to run many heavyweight services simultaneously. The 20 GB system disk is also relatively tight; after installing panels, logs, backups, and Docker images, pay attention to disk space. Important data should not be stored only on this machine; at least prepare independent backups.

Performance test results

This test used bench.sh and YABS introduced in the VPS testing script article. Readers who need to retest can first readComplete Collection of VPS Testing Scripts: Performance, Network Speed, and Route Testing Methods, connect via SSH, and execute the commands in the article.

Basic Information

  • CPU: Intel Xeon Processor (Skylake, IBRS), 2 cores, approx. 2.49 GHz
  • Virtualization: KVM
  • Memory: approx. 1.9 GiB
  • Disk: approx. 19.6 GB available capacity
  • IPv4: Normal
  • IPv6: Not enabled in this test
  • OS: Debian GNU/Linux 13
  • CPU AES-NI: Enabled
  • Swap: 0

The server had only been running for a few hours when testing began, and the average load was very low. Therefore, the following results are more suitable as „initial performance after provisioning“ and should not be taken as a long-term stability report.

Rongheguai full test results

This time, the Rongheguai GO version full test was used again, with the test date being October 28, 2026. For the complete script and how to run it, please refer to this site'sComprehensive VPS Testing Scripts。

  • CPU: Intel Xeon Processor (Skylake, IBRS), 2 vCPU; sysbench single-core 688.60 points, multi-core 1345.70 points
  • Memory: about 1.93 GB; Stream test about 9.46–11.72 GB/s
  • Disk: 4K read about 25.5k IOPS, write about 25.6k IOPS; 64K read/write about 9k IOPS
  • System: Debian 13.4, KVM virtualization, BBR + fq, IPv6 not enabled this time

From the hardware test, the CPU, memory, and disk performance of this machine can meet the needs of low-traffic WordPress, personal blogs, and development testing. The 4K random read/write of about 25.5k IOPS is more informative than simply looking at the sequential read/write of bench.sh; however, this is still a single test in the early stage of activation and cannot represent stability under long-term load.

Expand to view the complete Rongheguai test results (including routes, IP quality, and three-network speed tests)
---------------------------------VPS融合怪测试----------------------------------
 版本:v0.2.17
 测评频道: https://t.me/+UHVoo2U4VyA5NTQ1
 Go项目地址:https://github.com/oneclickvirt/ecs
 Shell项目地址:https://github.com/spiritLHLS/ecs
----------------------------------系统基础信息----------------------------------
 CPU 型号            : Intel Xeon Processor (Skylake, IBRS) @ 2494.140 MHz
 CPU 数量            : 2 Virtual CPU(s)
 CPU 缓存            : L1: 128 KB / L2: 8 MB / L3: 32 MB
 GPU 型号            : Virtio 1.0 GPU
 AES-NI              : ✔️ Enabled
 VM-x/AMD-V/Hyper-V  : ✔️ Enabled
 内存                : 284.92 MB / 1.93 GB
 气球驱动            : ✔️ Enabled
 内核页合并          : ❌ Undetected
 虚拟内存 Swap       : [ no swap partition or swap file detected ]
 硬盘空间            : 1.14 GB / 19.49 GB [5.9%] /dev/vda1 - /
 启动盘路径          : /dev/vda1
 系统                : debian 13.4 [x86_64]
 内核                : 6.12.74+deb13+1-amd64
 系统在线时间        : 0 days, 05 hours, 31 minutes
 时区                : Etc/UTC (UTC, +0000)
 负载                : 0.00 / 0.00 / 0.00
 虚拟化架构          : KVM
 NAT类型             : Full Cone
 TCP加速/队列        : bbr / fq
 TCP接收缓冲         : 4 KiB/128 KiB/6 MiB
 TCP发送缓冲         : 4 KiB/16 KiB/4 MiB
 Cgroup版本          : v2
 Cgroup CPU集合      : 0-1
 Cgroup内存使用      : 37 MiB
 BIOS厂商            : SeaBIOS
 BIOS版本            : 1.16.1-1.el9
 BIOS日期            : 04/01/2014
 PCI设备数量         : 11
 PCI驱动             : ata_piix,piix4_smbus,uhci_hcd,virtio-pci
 GPU设备数量         : 1
 GPU驱动             : virtio-pci
 NUMA/DIMM           : 1 / 1
 HugePages           : 总数 0 / 空闲 0 / 大小 2 MiB
 物理盘 1            : 协议 virtio / 健康 unsupported
 IPV4 ASN            : AS46783 EASY LINK LLC
 IPV4 Location       : Los Angeles / California / United States
----------------------------CPU测试-通过sysbench测试----------------------------
 1 线程测试(单核)得分: 688.60
 2 线程测试(多核)得分: 1345.70
 压力测试            : 20.022s / 2 线程 / 161.22 次/秒 / 3228 次
----------------------------内存测试-通过stream测试-----------------------------
 Function    Best Rate MB/s  Avg time     Min time     Max time
 Copy:            9461.3     0.017887     0.016911     0.021173
 Scale:          10703.9     0.016103     0.014948     0.020123
 Add:            11621.7     0.021960     0.020651     0.025771
 Triad:          11723.3     0.022346     0.020472     0.026972
------------------------------硬盘测试-通过fio测试------------------------------
 测试路径         块大小       读测试(IOPS)            写测试(IOPS)            总和(IOPS)            
 /                 4k        102.11 MB/s(25.5k)      102.38 MB/s(25.6k)      204.48 MB/s(51.1k)     
 /                 64k       581.82 MB/s(9090)       584.88 MB/s(9138)       1.17 GB/s(18.2k)       
 /                 512k      465.99 MB/s(910)        490.74 MB/s(958)        956.73 MB/s(1868)      
 /                 1m        465.92 MB/s(455)        496.95 MB/s(485)        962.88 MB/s(940)       
----------------------------------跨国平台解锁----------------------------------
===========[ IPV4 跨国平台 ]============
 Amazon Prime Video        YES (Region: US) [Native]
 Apple                     YES (Region: USA) [Native]
 Bilibili Anime            YES (Region: US) [Native]
 BingSearch                YES (Region: US)
 ChatGPT                   YES (Region: US) [Native]
 Claude                    Banned
 Coze                      YES (Region: US)
 DeepSeek                  YES (Region: US)
 Disney+                   YES (Region: US) [Native]
 Dola AI                   YES (Region: US)
 Gemini                    YES (Region: US) [Native]
 Google Play Store         YES (Region: US) [Native]
 GoogleSearch              YES
 Grok                      YES (Region: US) [Native]
 Instagram Licensed Audio  Rate Limited (HTTP 429)
 IQiYi                     YES (Region: US) [Native]
 Kimi                      YES (Region: OVERSEA) [Native]
 KOCOWA                    YES [Native]
 MathsSpot                 YES (Region: US) [Via DNS]
 MetaAI                    YES [Native]
 Microsoft Copilot         YES (Region: US) [Native]
 Mistral AI                NO (Region: US)
 Netflix                   YES (Region: US) [Native]
 Netflix CDN               YES (Region: US) [Native]
 OneTrust                  YES (Region: US CALIFORNIA) [Native]
 ParamountPlus             YES [Native]
 Perplexity AI             Banned
 Poe                       YES (Region: US) [Via DNS]
 Reddit                    YES
 SonyLiv                   Error
 Sora                      YES (Region: US)
 Spotify Registration      YES (Region: US) [Native]
 Steam Store               YES (Community Available) (Region: US)
 TikTok                    YES (Region: US) [Native]
 TVBAnywhere+              YES (Region: US) [Native]
 Viu.com                   YES [Native]
 WeTV                      YES (Region: AE) [Native]
 Wikipedia Editability     Banned
 X (formerly Twitter)      YES (Region: US)
 YouTube CDN               YES (Youtube Video Server) (Region: LAX) [Native]
 YouTube Region            YES [Native]
-----------------------------------IP质量检测-----------------------------------
 以下为各数据库编号,输出结果后将自带数据库来源对应的编号
 ipinfo数据库  [0] | scamalytics数据库 [1] | virustotal数据库   [2] | abuseipdb数据库   [3] | ip2location数据库    [4]
 ip-api数据库  [5] | ipwhois数据库     [6] | ipregistry数据库   [7] | ipdata数据库      [8] | db-ip数据库          [9]
 ipapiis数据库 [A] | ipapicom数据库    [B] | bigdatacloud数据库 [C] | dkly数据库        [D] | ipqualityscore数据库 [E]
 ipintel数据库 [F] | ipfighter数据库   [G] | fraudlogix数据库   [H] | cloudflare数据库  [I] |
 IPV4:
 安全得分:
 社区投票-无害: 0 [2] 
 社区投票-恶意: 0 [2]
 欺诈得分(越低越好): 0 [1] 
 滥用得分(越低越好): 0 [3] 
 威胁级别: low [B] 
 黑名单记录统计:(有多少黑名单网站有记录):
 无害记录数: 0 [2]  恶意记录数: 0 [2]  可疑记录数: 0 [2]  无记录数: 92 [2] 
 安全信息:
 使用类型: unknown [8] hosting [0 3 7 C]
 公司类型: business [0] isp [7]
 是否云提供商: Yes [7] 
 是否数据中心: Yes [0 1 C] No [5 8]
 是否移动设备: No [5 C] 
 是否代理: No [0 1 4 5 7 8 B C] 
 是否VPN: No [0 1 7 C] 
 是否TorExit: No [1 7] 
 是否Tor出口: No [1 7] 
 是否网络爬虫: No [B] 
 是否匿名: No [1 7 8] 
 是否攻击者: No [7 8] 
 是否滥用者: No [7 8 C] 
 是否威胁: No [7 8 C] 
 是否中继: No [0 7 8 C]
 是否Bogon: No [7 8 A C] 
 DNS-黑名单: 301(Total_Check) 254(Clean) 10(Blacklisted) 37(Other) 
----------------------------------邮件端口检测----------------------------------
 Platform  SMTP  SMTPS POP3  POP3S IMAP  IMAPS
 LocalPort ✔     ✔     ✔     ✔     ✔     ✔    
 QQ        ✔     ✔     ✔     ✘     ✔     ✘    
 163       ✔     ✔     ✔     ✘     ✔     ✘    
 Sohu      ✔     ✔     ✔     ✘     ✔     ✘    
 Yandex    ✔     ✔     ✔     ✘     ✔     ✘    
 Gmail     ✔     ✔     ✘     ✘     ✘     ✘    
 Outlook   ✔     ✘     ✔     ✘     ✔     ✘    
 Office365 ✔     ✘     ✔     ✘     ✔     ✘    
 Yahoo     ✔     ✔     ✘     ✘     ✘     ✘    
 MailCOM   ✔     ✔     ✔     ✘     ✔     ✘    
 MailRU    ✔     ✔     ✘     ✘     ✔     ✘    
 AOL       ✔     ✔     ✘     ✘     ✘     ✘    
 GMX       ✔     ✔     ✔     ✘     ✔     ✘    
 Sina      ✔     ✔     ✔     ✘     ✔     ✘    
 Apple     ✘     ✔     ✘     ✘     ✘     ✘    
 FastMail  ✘     ✔     ✘     ✘     ✘     ✘    
 ProtonMail✘     ✘     ✘     ✘     ✘     ✘    
 MXRoute   ✔     ✘     ✔     ✘     ✔     ✘    
 Namecrane ✔     ✔     ✔     ✘     ✔     ✘    
 XYAMail   ✘     ✘     ✘     ✘     ✘     ✘    
 ZohoMail  ✘     ✔     ✘     ✘     ✘     ✘    
 Inbox_eu  ✔     ✔     ✔     ✘     ✘     ✘    
 Free_fr   ✘     ✔     ✔     ✘     ✔     ✘    
-------------------------------上游及回程线路检测-------------------------------
 北京电信v4 219.141.140.10  电信CN2GIA [精品线路]
 北京联通v4 202.106.195.68  联通9929   [优质线路]
 北京移动v4 221.179.155.161 移动CMIN2  [精品线路] 移动CMI    [普通线路]
 上海电信v4 202.96.209.133  电信CN2GIA [精品线路]
 上海联通v4 210.22.97.1     联通9929   [优质线路]
 上海移动v4 211.136.112.200 移动CMIN2  [精品线路] 移动CMI    [普通线路]
 广州电信v4 58.60.188.222   电信CN2GIA [精品线路]
 广州联通v4 210.21.196.6    联通9929   [优质线路]
 广州移动v4 120.196.165.24  移动CMIN2  [精品线路] 移动CMI    [普通线路]
 成都电信v4 61.139.2.69     电信CN2GIA [精品线路]
 成都联通v4 119.6.6.6       联通9929   [优质线路] 联通4837   [普通线路]
 成都移动v4 211.137.96.205  移动CMIN2  [精品线路] 移动CMI    [普通线路]
 准确线路自行查看详细路由,本测试结果仅作参考
 同一目标地址多个线路时,检测可能已越过汇聚层,除第一个线路外,后续信息可能无效
--------------------------------三网回程路由检测--------------------------------
 [NextTrace API] preferred API IP - 185.218.7.242 - 249.86ms - Saltyfish.SJC (Relay) → Misaka.LAX
 广州电信 - ICMP v4 - traceroute to 58.60.188.222, 30 hops max, 52 byte packets
 *
 0.46 ms      *          *                  *
 1.74 ms      AS4134     [CHINANET-US]      美国, 加利福尼亚, 洛杉矶, en.chinatelecom.com.cn  电信
 142.62 ms    *          [CN2-BackBone]     中国, 广东, 广州, chinatelecom.cn  电信
 149.62 ms    *          [CN2-BackBone]     中国, 广东, 广州, chinatelecom.cn  电信
 145.99 ms    *          [CN2-BackBone]     中国, 广东, 广州, chinatelecom.cn  电信
 *
 149.39 ms    AS4134     [CHINANET-GD]      中国, 广东, 深圳, en.chinatelecom.com.cn  电信
 *
 广州联通 - ICMP v4 - traceroute to 210.21.196.6, 30 hops max, 52 byte packets
 *
 0.72 ms      *          *                  *
 2.29 ms      AS10099    *                  中国, 香港, chinaunicomglobal.com 
 140.12 ms    AS10099    [CUG-BACKBONE]     中国, 上海, chinaunicomglobal.com  联通
 137.18 ms    AS9929     [CNC-BACKBONE]     中国, 上海, chinaunicom.cn  联通 CUII
 159.56 ms    AS9929     [CNC-BACKBONE]     中国, 广东, 广州, chinaunicom.cn  联通 CUII
 160.40 ms    *          [CNC-BACKBONE]     中国, 广东, 广州
 *
 164.80 ms    AS4837     [CU169-BACKBONE]   中国, 广东, 广州, chinaunicom.cn 
 *
 171.02 ms    AS17623    [APNIC-AP]         中国, 广东, 深圳, chinaunicom.cn  联通
 161.46 ms    AS17623    *                  中国, 广东, 深圳, chinaunicom.cn  联通
 广州移动 - ICMP v4 - traceroute to 120.196.165.24, 30 hops max, 52 byte packets
 *
 0.64 ms      *          *                  *
 0.90 ms      AS58807    [CMIN2-NET]        中国, cmi.chinamobile.com  移动
 128.43 ms    AS58807    [CMIN2-NET]        美国, 加利福尼亚, 洛杉矶, cmi.chinamobile.com  移动
 127.12 ms    AS58807    [CMIN2-NET]        中国, 上海, cmi.chinamobile.com  移动
 127.76 ms    AS9808     [CMNET]            中国, 上海, chinamobileltd.com  移动
 128.19 ms    AS9808     [CMNET]            中国, 上海, chinamobileltd.com  移动
 128.16 ms    AS9808     [CMNET]            中国, 上海, chinamobileltd.com 
 153.65 ms    AS9808     [CMNET]            中国, 北京, chinamobileltd.com  移动
 154.42 ms    AS9808     [CMNET]            中国, 北京, chinamobileltd.com  移动
 *
 156.74 ms    AS56040    [APNIC-AP]         中国, 广东, 深圳, gd.10086.cn  移动
-----------------------------------PING值检测-----------------------------------
 电信上海             127 | 电信南京             135 | 电信安徽             139 | 
 电信武汉             139 | 电信浙江             140 | 电信湖北             145 | 
 电信天津             149 | 电信福建             150 | 电信山东             153 | 
 电信新疆             154 | 电信河南             154 | 电信四川             158 | 
 电信吉林             159 | 电信海南             161 | 电信重庆             162 | 
 电信内蒙古           164 | 电信江苏             164 | 电信江西             166 | 
 电信贵州             166 | 电信广西壮族         167 | 电信广东             167 | 
 电信辽宁             169 | 电信云南             171 | 电信黑龙江           172 | 
 电信宁夏             174 | 电信甘肃             175 | 电信青海             176 | 
 电信山西             185 | 电信西藏             187 | 电信Suzhou5G        9999 | 
 电信Zhenjiang5G     9999 | 电信北京            9999 | 电信宁波            9999 | 
 电信成都            9999 | 电信扬州            9999 | 电信杭州            9999 | 
 电信河北            9999 | 电信湖南            9999 | 电信苏州            9999 | 
 电信陕西            9999 | 
 移动上海             127 | 移动江苏             139 | 移动安徽             140 | 
 移动浙江             141 | 移动山东             145 | 移动江西             147 | 
 移动湖北             149 | 移动北京             153 | 移动河南             154 | 
 移动山西             156 | 移动内蒙古           157 | 移动河北             160 | 
 移动甘肃             160 | 移动四川             163 | 移动福建             163 | 
 移动贵州             166 | 移动黑龙江           166 | 移动广西壮族         167 | 
 移动成都             168 | 移动辽宁             169 | 移动宁夏             170 | 
 移动海南             172 | 移动吉林             173 | 移动重庆             174 | 
 移动陕西             179 | 移动青海             181 | 移动新疆             182 | 
 移动西藏             217 | 移动Suzhou          9999 | 移动云南            9999 | 
 移动南京            9999 | 移动天津            9999 | 移动广东            9999 | 
 移动湖南            9999 | 移动福州            9999 | 
 联通浙江             137 | 联通上海             142 | 联通山东             144 | 
 联通上海5G           145 | 联通安徽             146 | 联通江苏             148 | 
 联通湖北             151 | 联通河南             152 | 联通山西             154 | 
 联通甘肃             159 | 联通福建             160 | 联通宁夏             163 | 
 联通内蒙古           163 | 联通太原市           164 | 联通江西             164 | 
 联通四川             164 | 联通河北             165 | 联通辽宁             167 | 
 联通天津             167 | 联通贵州             170 | 联通重庆             172 | 
 联通广西壮族         174 | 联通青海             176 | 联通海南             178 | 
 联通黑龙江           188 | 联通吉林             191 | 联通新疆             192 | 
 联通西藏             203 | 联通Beijing         9999 | 联通云南            9999 | 
 联通北京            9999 | 联通广东            9999 | 联通杭州            9999 | 
 联通沈阳            9999 | 联通湖南            9999 | 联通福州            9999 | 
 联通陕西            9999 | 
 TG-DC3 MIA USA        92 | TG-DC1 MIA USA        96 | TG-DC4 AMS NL        176 | 
 TG-DC2 AMS NL        181 | TG-DC5 Singapore     199 | 
 CNN                    8 | StackOverflow         11 | Azure                 15 | 
 MetaAI                17 | DigitalOcean          18 | Apple                 26 | 
 NYTimes               27 | BBC                   29 | eBay                  36 | 
 Twitch                38 | Claude                48 | Sora                  49 | 
 Wikipedia             62 | AWS                   84 | Amazon                88 | 
 OpenAI                90 | Docker Hub           106 | Microsoft            109 | 
 Reddit               126 | Gemini               128 | Steam                133 | 
 AliExpress           173 | Google               174 | Bing                 187 | 
 DisneyPlus           191 | GitLab               192 | Google Cloud         198 | 
 ViuTV                208 | Bilibili             212 | Instagram            214 | 
 Facebook             216 | Spotify              225 | GitHub               243 | 
 YouTube              261 | TikTok               376 | PrimeVideo           388 | 
 iQIYI                388 | Twitter/X            414 | TVB Anywhere         721 | 
 Netflix             9999 | 
----------------------------------就近节点测速----------------------------------
 位置            上传速度        下载速度        延迟            丢包率          
 Speedtest.net   0.00 Mbps       0.00 Mbps       76.879438ms     0.00% (Sent: 328/Dup: 0/Max: 327)
 日本东京        26.32 Mbps      41.92 Mbps      232.867774ms                    
 联通Shanghai    20.30 Mbps      5.17 Mbps       267.11 ms       N/A             
 联通Beijing     31.02 Mbps      125.57 Mbps     329.25 ms       N/A             
 电信Shanghai    N/A             86.09 Mbps      267.81 ms       N/A             
 电信Beijing     31.98 Mbps      139.71 Mbps     349.67 ms       N/A             
 移动Nanjing     38.47 Mbps      108.94 Mbps     280.58 ms       N/A             
 移动Shanghai    45.18 Mbps      164.87 Mbps     298.22 ms       N/A             
--------------------------------------------------------------------------------
 花费          : 6 分 58 秒
 时间          : Wed Oct 28 12:24:59 UTC 2026
--------------------------------------------------------------------------------

Disk I/O

The results of three bench.sh tests are:

  • First run: 630 MB/s
  • Second run: 563 MB/s
  • Third run: 523 MB/s
  • Average: 572 MB/s

This result is sufficient for WordPress, small databases, and general web applications; at least it does not show obvious low-speed disk issues. However, this is mainly a sequential I/O reference value and does not equal real-world performance under high-concurrency databases. To judge database experience, additional 4K random read/write tests, sustained load, and multiple tests at different times are needed.

International Bandwidth

The multi-location results from bench.sh are as follows; upload and download directions may be affected by test node load:

Test NodeUploaddownloadLatency
Los Angeles508 Mbps490 Mbps0.94 ms
Dallas520 Mbps490 Mbps31.90 ms
Paris531 Mbps530 Mbps135.70 ms
Amsterdam570 Mbps548 Mbps151.74 ms
Singapore552 Mbps548 Mbps192.77 ms
Tokyo520 Mbps522 Mbps127.61 ms

The iperf3 results from YABS are also fairly close to this range: to London, send/receive is about 503/505 Mbps; to Amsterdam, about 501/479 Mbps; to Singapore, send is about 490 Mbps and receive about 174 Mbps.

Overall, the 500 Mbps international bandwidth is largely delivered in this test. The differences in receive speeds between different nodes show that a nominal bandwidth value does not mean it can be fully saturated in any region at any time; the load and direction of the test node itself will also affect the results.

Network focus test: How are the routes of the three major Chinese networks, really?

VMRack's official website advertises premium routes such as Los Angeles CN2 GIA, AS9929, and CMIN2, and claims that latency to China can be as low as 120 ms. Official website claims can only serve as pre-purchase information; whether it suits you still depends on the actual IP's routing and your access region.

Routing and latency

In addition to using Rongheguai's return route detection, this time traceroute was also executed independently from the server, and the results were compared withCN2, 9929, CMIN2 route descriptionsto make a judgment. Note that the middle part of the route may show asterisks due to carrier restrictions; you cannot judge whether a route is good or bad solely based on whether a certain hop responds.

Carrier directionReturn route test resultsLatency reference
China TelecomCN2-BackBone appears on the Guangzhou return route; AS4134 CHINANET-US is visible in the earlier segmentApprox. 143–150 ms
China UnicomAS9929 / CNC-BACKBONE appears on the Guangzhou return route, and AS4837 also appears in the later segmentApprox. 137–171 ms
China MobileAS58807 CMIN2-NET appears on the Guangzhou return route, then traffic enters CMNETApprox. 127–157 ms

Judging from the return route results, this VMRack machine does indeed exhibit the premium three-network line characteristics advertised by the vendor: CN2 on the Telecom side, 9929 on the Unicom side, and CMIN2 on the Mobile side. However, this only reflects the routing performance for this test target and time period; it cannot guarantee that the same routes will be maintained permanently across all regions, all IPs, and all time periods.

  • Telecom public DNS 223.5.5.5: 59.43.* and 101.95.* appear in the path; responsive node latency is approx. 129–136 ms.
  • 119.29.29.29: 59.43.* appears in the path; responsive node latency is approx. 146–150 ms.
  • Mobile target 211.136.17.107: 223.120.* and 221.183.* appear in the path; responsive node latency is approx. 127–156 ms.

The independent route test and the fusion-monster return route test are broadly consistent in direction, but some targets stop responding at later nodes. Route tests are better suited for judging the general path; actual website access should also take into account TCP connections, HTTP downloads, packet loss, and peak-evening performance.

Test NodeUploaddownloadLatency
Telecom ShanghaiN/A86.09 Mbps267.81 ms
Telecom Beijing31.98 Mbps139.71 Mbps349.67 ms
Unicom Shanghai20.30 Mbps5.17 Mbps267.11 ms
Unicom Beijing31.02 Mbps125.57 Mbps329.25 ms
China Mobile Nanjing38.47 Mbps108.94 Mbps280.58 ms
China Mobile Shanghai45.18 Mbps164.87 Mbps298.22 ms

Streaming and Overseas Platform Unlock Test

For foreign trade users, overseas social media users, or those who need to access foreign AI, streaming, and development platforms, whether the IP can log in and be used normally is equally important. This fusion monster full test also checked the IPv4 unlock status of common platforms, and the results show that this VMRack test machine's US attributes are relatively normal.

Platform or ServiceTest Result
ChatGPT, Gemini, Grok, Copilot, DeepSeekAvailable in US region or shows overseas region
Netflix, Disney+, Prime Video, YouTubeShows US region or US CDN
X, Reddit, TikTok, Meta AIAvailable
Apple, Google Play, InstagramApple and Google Play show US; Instagram audio authorization API returns HTTP 429 rate limit
Claude, Perplexity, Mistral AIClaude and Perplexity are restricted, Mistral AI is not unlocked

Therefore, this IP is used to log in to common overseas social media, AI services, and streaming platforms, and the overall performance is good; but not all platforms can guarantee unlocking, and services such as Claude and Perplexity have already shown restrictions. Platform policies, IP risk control, and account regions will change, and test results can only represent the current IP and current time, and cannot promise long-term stable unlocking. When used for foreign trade business, it is recommended to actually log in to your own account to confirm that verification codes, payments, advertising backends, and business systems can all be used normally.

VMRack Other Line Type Description

The VMRack official website currently also offers two other network options: the Platinum line is labeled as 163/10099/CMI, and the Global BGP line is labeled as Cogent/Arelion. The official website labels the Premium line where this test machine is located as CN2/9929/CMIN2. You can first visitVMRack official websiteto view currently purchasable lines and inventory.

Line TypeOfficial Website LabelIs this article based on actual testing?
PremiumCN2 / 9929 / CMIN2Tested; the network and unlocking results in this article mainly come from this category
Platinum163 / 10099 / CMINot tested; the results in this article cannot be directly applied
Global BGPCogent / ArelionNot tested; the results in this article cannot be directly applied

In terms of use cases, 163/10099/CMI is more suitable for users who need China mainland optimization but do not have as high line requirements as CN2 GIA/9929/CMIN2; Cogent/Arelion Global BGP is more suitable for foreign trade, overseas websites, and global access scenarios. However, this is only a judgment of usage direction based on line type, not an actual test conclusion for these two line packages. Before actually purchasing, you should still conduct return route, three-network speed test, packet loss, evening peak, and platform login tests for the specific package and specific IP.

BandwagonHost network comparison test

To compare network performance more intuitively, I also used the same version of Rongheguai„s “Network Single Item„ to test on a BandwagonHost Los Angeles server. Only the network is compared here; the CPU, memory, disk, and price of the two machines are not compared. The two tests were completed on the same day, one after the other, but different IPs, node load, and line scheduling may still cause result variations.

DirectionVMRack return route resultsBandwagonHost return route results
China TelecomCN2-BackBone appears, about 143–150 msCN2-BackBone / CN2-Global appears, about 161–168 ms, later segment about 175 ms
China UnicomAS9929 / CNC-BACKBONE appears, AS4837 appears in the later segment, about 137–171 msCN2-BackBone appears in the early segment, then enters AS4837 in the later segment, about 151–166 ms
China MobileAS58807 CMIN2-NET appears, then enters CMNET, about 127–157 msAS58807 CMIN2-NET appears, then enters CMNET, about 126–155 ms

From the return routes, both machines show line characteristics such as CN2 and CMIN2 in the China Telecom and China Mobile directions. VMRack's China Unicom return route shows AS9929 in the early segment, while BandwagonHost's test this time mainly shows CN2-BackBone followed by AS4837, so BandwagonHost's result this time cannot simply be described as full 9929. Overall, VMRack's line performance is not noticeably behind this BandwagonHost test machine, but different IPs and different times still require separate retesting.

Shanghai test nodeVMRack uploadVMRack downloadBandwagonHost uploadBandwagonHost downloadLatency comparison
China TelecomN/A86.09 Mbps9.86 Mbps112.38 Mbps267.81 / 256.45 ms
China Unicom20.30 Mbps5.17 Mbps9.95 Mbps5.32 Mbps267.11 / 264.99 ms
China Mobile45.18 Mbps164.87 Mbps9.22 Mbps161.09 Mbps298.22 / 267.40 ms

During the three-network speed test in Shanghai, BandwagonHost„s China Telecom download was slightly higher than VMRack“s, the China Mobile download was roughly comparable, and the China Unicom download was only about 5 Mbps for both; VMRack„s upload speed was higher in this test. This result shows that the network gap between the two machines is not as simple as the marketing suggests, and cannot be summed up with “BandwagonHost is definitely faster„ or “VMRack completely replaces BandwagonHost.„

The collapsible section below retains the complete output of this BandwagonHost network single-item test for verification. The full raw results can also beviewed on the Fusion Monster share page。

Expand to view the complete BandwagonHost Fusion Monster network single-item results
---------------------------------VPS融合怪测试----------------------------------
 版本:v0.2.17
 测评频道: https://t.me/+UHVoo2U4VyA5NTQ1
 Go项目地址:https://github.com/oneclickvirt/ecs
 Shell项目地址:https://github.com/spiritLHLS/ecs
-------------------------------------IP信息-------------------------------------
 IPV4 ASN            : AS25820 IT7 Networks Inc
 IPV4 Location       : Los Angeles / California / United States
-----------------------------------IP质量检测-----------------------------------
 以下为各数据库编号,输出结果后将自带数据库来源对应的编号
 ipinfo数据库  [0] | scamalytics数据库 [1] | virustotal数据库   [2] | abuseipdb数据库   [3] | ip2location数据库    [4]
 ip-api数据库  [5] | ipwhois数据库     [6] | ipregistry数据库   [7] | ipdata数据库      [8] | db-ip数据库          [9]
 ipapiis数据库 [A] | ipapicom数据库    [B] | bigdatacloud数据库 [C] | dkly数据库        [D] | ipqualityscore数据库 [E]
 ipintel数据库 [F] | ipfighter数据库   [G] | fraudlogix数据库   [H] | cloudflare数据库  [I] |
 IPV4:
 安全得分:
 社区投票-无害: 0 [2] 
 社区投票-恶意: 2 [2] 
 欺诈得分(越低越好): 1 [1] 
 滥用得分(越低越好): 4 [3] 
 威胁级别: low [B] 
 黑名单记录统计:(有多少黑名单网站有记录):
 无害记录数: 54 [2]  恶意记录数: 0 [2]  可疑记录数: 2 [2]  无记录数: 36 [2]
 安全信息:
 使用类型: hosting [0 3 7 8 C] 
 公司类型: business [7] hosting [0]
 是否云提供商: Yes [7] 
 是否数据中心: No [1 5] Yes [0 8 C]
 是否移动设备: No [5 C] 
 是否代理: Yes [5] No [0 1 4 7 8 B C]
 是否VPN: No [0 1 C] Yes [7]
 是否Tor: No [0 1 3 7 8 B C] 
 是否Tor出口: No [1 7] 
 是否网络爬虫: No [B] 
 是否匿名: No [1 8] Yes [7]
 是否攻击者: No [7 8] 
 是否滥用者: No [7 8 C] 
 是否威胁: No [7 8 C] 
 是否中继: No [0 7 8 C] 
 是否Bogon: No [7 8 A C] 
 DNS-黑名单: 301(Total_Check) 130(Clean) 8(Blacklisted) 163(Other) 
-------------------------------上游及回程线路检测-------------------------------
 北京电信v4 219.141.140.10  电信CN2GIA [精品线路]
 北京联通v4 202.106.195.68  电信CN2GIA [精品线路]
 北京移动v4 221.179.155.161 移动CMIN2  [精品线路] 移动CMI    [普通线路]
 上海电信v4 202.96.209.133  电信CN2GIA [精品线路]
 上海联通v4 210.22.97.1     电信CN2GIA [精品线路] 联通4837   [普通线路]
 上海移动v4 211.136.112.200 移动CMIN2  [精品线路] 移动CMI    [普通线路]
 广州电信v4 58.60.188.222   电信CN2GIA [精品线路]
 广州联通v4 210.21.196.6    电信CN2GIA [精品线路] 联通4837   [普通线路]
 广州移动v4 120.196.165.24  移动CMIN2  [精品线路] 移动CMI    [普通线路]
 成都电信v4 61.139.2.69     电信CN2GIA [精品线路]
 成都联通v4 119.6.6.6       联通4837   [普通线路]
 成都移动v4 211.137.96.205  移动CMIN2  [精品线路] 移动CMI    [普通线路]
 准确线路自行查看详细路由,本测试结果仅作参考
 同一目标地址多个线路时,检测可能已越过汇聚层,除第一个线路外,后续信息可能无效
--------------------------------三网回程路由检测--------------------------------
 [NextTrace API] preferred API IP - 45.88.195.154 - 11.52ms - DMIT.LAX
 广州电信 - ICMP v4 - traceroute to 58.60.188.222, 30 hops max, 52 byte packets
 22.39 ms     *          *                  *
 16.42 ms     *          *                  *
 1.08 ms      AS4134     [CHINANET-US]      美国, 加利福尼亚, 洛杉矶, en.chinatelecom.com.cn  电信
 162.30 ms    *          [CN2-BackBone]     中国, 广东, 广州, chinatelecom.cn  电信
 161.21 ms    *          [CN2-Global]       中国, 广东, 广州, chinatelecom.cn  电信
 167.79 ms    *          [CN2-BackBone]     中国, 广东, 广州, chinatelecom.cn  电信
 *
 175.02 ms    AS134774   [CHINANET-GD]      中国, 广东, 深圳, chinatelecom.cn  电信
 *
 广州联通 - ICMP v4 - traceroute to 210.21.196.6, 30 hops max, 52 byte packets
 22.46 ms     *          *                  *
 25.23 ms     *          *                  *
 0.58 ms      AS4134     [CHINANET-US]      美国, 加利福尼亚, 洛杉矶, en.chinatelecom.com.cn  电信
 151.58 ms    *          [CN2-BackBone]     中国, 广东, 广州, chinatelecom.cn  电信
 153.18 ms    *          [CN2-Global]       中国, 广东, 广州, chinatelecom.cn  电信
 154.87 ms    *          [CN2-BackBone]     中国, 广东, 广州, chinatelecom.cn  电信
 159.70 ms    AS4837     [CU169-BACKBONE]   中国, 广东, 广州, chinaunicom.cn  联通
 *
 165.73 ms    AS17623    [APNIC-AP]         中国, 广东, 深圳, chinaunicom.cn  联通
 155.86 ms    AS17623    *                  中国, 广东, 深圳, chinaunicom.cn  联通
 广州移动 - ICMP v4 - traceroute to 120.196.165.24, 30 hops max, 52 byte packets
 22.41 ms     *          *                  *
 25.23 ms     *          *                  *
 0.73 ms      AS25820    *                  美国, 加利福尼亚州, 洛杉矶, it7.net 
 *
 126.45 ms    AS58807    [CMIN2-NET]        中国, 上海, cmi.chinamobile.com  移动
 126.88 ms    AS9808     [CMNET]            中国, 上海, chinamobileltd.com  移动
 127.85 ms    AS9808     [CMNET]            中国, 上海, chinamobileltd.com  移动
 129.14 ms    AS9808     [CMNET]            中国, 上海, chinamobileltd.com 
 151.51 ms    AS9808     [CMNET]            中国, 北京, chinamobileltd.com  移动
 153.35 ms    AS9808     [CMNET]            中国, 北京, chinamobileltd.com  移动
 *
 154.96 ms    AS56040    [APNIC-AP]         中国, 广东, 深圳, gd.10086.cn  移动
-----------------------------------PING值检测-----------------------------------
 电信上海             131 | 电信南京             133 | 电信安徽             135 | 
 电信江苏             136 | 电信武汉             138 | 电信浙江             138 | 
 电信福建             141 | 电信湖北             142 | 电信江西             145 | 
 电信新疆             149 | 电信甘肃             150 | 电信河南             153 | 
 电信宁夏             157 | 电信山东             158 | 电信天津             158 | 
 电信四川             158 | 电信内蒙古           159 | 电信重庆             159 | 
 电信山西             161 | 电信辽宁             161 | 电信广东             162 | 
 电信广西壮族         162 | 电信贵州             165 | 电信海南             166 | 
 电信黑龙江           167 | 电信青海             167 | 电信吉林             170 | 
 电信云南             171 | 电信西藏             197 | 电信Suzhou5G        9999 | 
 电信Zhenjiang5G     9999 | 电信北京            9999 | 电信宁波            9999 | 
 电信成都            9999 | 电信扬州            9999 | 电信杭州            9999 | 
 电信河北            9999 | 电信湖南            9999 | 电信苏州            9999 | 
 电信陕西            9999 | 
 移动上海             128 | 移动浙江             131 | 移动江苏             134 | 
 移动江西             141 | 移动山东             144 | 移动湖北             149 | 
 移动河南             150 | 移动北京             153 | 移动山西             154 | 
 移动福建             156 | 移动甘肃             159 | 移动重庆             164 | 
 移动广西壮族         165 | 移动宁夏             165 | 移动黑龙江           166 | 
 移动成都             168 | 移动海南             170 | 移动辽宁             170 | 
 移动吉林             173 | 移动陕西             174 | 移动四川             174 | 
 移动贵州             178 | 移动新疆             182 | 移动青海             185 | 
 移动西藏             210 | 移动Suzhou          9999 | 移动云南            9999 | 
 移动内蒙古          9999 | 移动南京            9999 | 移动天津            9999 | 
 移动安徽            9999 | 移动广东            9999 | 移动河北            9999 | 
 移动湖南            9999 | 移动福州            9999 | 
 联通上海5G           126 | 联通上海             128 | 联通浙江             137 | 
 联通江苏             140 | 联通安徽             145 | 联通江西             146 | 
 联通天津             150 | 联通太原市           153 | 联通湖北             153 | 
 联通山东             154 | 联通内蒙古           158 | 联通宁夏             162 | 
 联通重庆             162 | 联通甘肃             163 | 联通河北             167 | 
 联通福建             169 | 联通河南             169 | 联通四川             170 | 
 联通广西壮族         170 | 联通山西             171 | 联通贵州             172 | 
 联通青海             175 | 联通海南             177 | 联通辽宁             182 | 
 联通吉林             187 | 联通西藏             191 | 联通黑龙江           192 | 
 联通新疆             202 | 联通Beijing         9999 | 联通云南            9999 | 
 联通北京            9999 | 联通广东            9999 | 联通杭州            9999 | 
 联通沈阳            9999 | 联通湖南            9999 | 联通福州            9999 | 
 联通陕西            9999 | 
 TG-DC3 MIA USA        56 | TG-DC1 MIA USA        56 | TG-DC2 AMS NL        141 | 
 TG-DC4 AMS NL        153 | TG-DC5 Singapore     248 | 
 Reddit                 6 | CNN                    8 | Claude                10 | 
 MetaAI                12 | Twitch                18 | OpenAI                19 | 
 Sora                  21 | DigitalOcean          21 | Wikipedia             22 | 
 StackOverflow         22 | NYTimes               24 | BBC                   26 | 
 Azure                 27 | Google                45 | AWS                   86 | 
 Facebook              95 | Gemini               103 | Instagram            123 | 
 Amazon               129 | eBay                 135 | Apple                143 | 
 YouTube              148 | GitHub               159 | Spotify              173 | 
 Google Cloud         178 | Docker Hub           182 | AliExpress           187 | 
 Bilibili             197 | ViuTV                201 | PrimeVideo           245 | 
 Bing                 253 | Twitter/X            258 | GitLab               259 | 
 TikTok               266 | DisneyPlus           287 | Microsoft            341 | 
 Netflix              377 | Steam                379 | iQIYI                434 | 
 TVB Anywhere         711 | 
----------------------------------TCP握手延迟-----------------------------------
 汇总 目标:94  握手:282/282  成功率:100.0%  失败:0
 平台           成功/尝试  丢包   Min   Avg   P50   P95   Max    D    R    T    O
 Adobe                3/3  0.0%   1.9   2.1   2.0   2.4   2.5    0    0    0    0
 AliExpress           3/3  0.0%  61.9  62.3  62.1  62.9  63.0    0    0    0    0
 Amazon               3/3  0.0%  55.4  58.8  59.3  61.5  61.7    0    0    0    0
 Apple                3/3  0.0%  61.3  61.7  61.6  62.2  62.2    0    0    0    0
 AWS                  3/3  0.0%   3.1   3.8   3.8   4.4   4.5    0    0    0    0
 Azure                3/3  0.0%  59.1  61.0  61.9  62.0  62.0    0    0    0    0
 Azure                3/3  0.0%   1.7   3.3   2.0   5.7   6.1    0    0    0    0
 BBC                  3/3  0.0%   1.7   1.9   1.9   2.0   2.0    0    0    0    0
 Bilibili             3/3  0.0%   2.6  12.7   2.9  29.6  32.5    0    0    0    0
 Bing                 3/3  0.0%  60.3  62.2  63.0  63.3  63.3    0    0    0    0
 Canva                3/3  0.0%   2.6   3.1   2.8   3.7   3.8    0    0    0    0
 ChatGPT              3/3  0.0%   2.1   2.3   2.3   2.5   2.5    0    0    0    0
 Claude               3/3  0.0%   3.6   6.9   8.4   8.5   8.5    0    0    0    0
 Cloudflare           3/3  0.0%   2.0   2.5   2.4   3.0   3.0    0    0    0    0
 CNN                  3/3  0.0%   1.9   2.1   2.0   2.5   2.5    0    0    0    0
 Coinbase             3/3  0.0%   2.5   2.7   2.6   2.9   3.0    0    0    0    0
 Copilot              3/3  0.0%   2.7   2.9   2.9   3.0   3.0    0    0    0    0
 Coursera             3/3  0.0%   1.7   2.4   2.5   2.8   2.9    0    0    0    0
 DeepL                3/3  0.0%   2.3   2.6   2.6   2.8   2.8    0    0    0    0
 DeepSeek             3/3  0.0%   1.7   1.9   2.0   2.1   2.2    0    0    0    0
 DigitalOcean         3/3  0.0%   2.2   4.1   3.8   6.0   6.3    0    0    0    0
 DigitalOcean         3/3  0.0%   2.6   3.0   2.7   3.4   3.5    0    0    0    0
 Discord              3/3  0.0%   2.3   2.5   2.6   2.7   2.7    0    0    0    0
 DisneyPlus           3/3  0.0%  60.4  68.2  61.9  80.2  82.3    0    0    0    0
 Docker Hub           3/3  0.0%   2.4   3.0   2.4   4.1   4.3    0    0    0    0
 Dropbox              3/3  0.0%   1.8   2.2   2.2   2.6   2.7    0    0    0    0
 eBay                 3/3  0.0%  59.2  60.2  59.2  61.9  62.2    0    0    0    0
 edX                  3/3  0.0%   3.1   3.5   3.4   3.9   4.0    0    0    0    0
 EpicGames            3/3  0.0%   2.2   2.6   2.3   3.0   3.1    0    0    0    0
 Facebook             3/3  0.0%   1.8   4.5   5.3   6.2   6.3    0    0    0    0
 Figma                3/3  0.0%   1.8   2.1   2.2   2.3   2.3    0    0    0    0
 Gemini               3/3  0.0%   6.6  12.9   7.8  22.8  24.5    0    0    0    0
 GitHub               3/3  0.0%   9.5   9.6   9.5   9.9   9.9    0    0    0    0
 GitHub               3/3  0.0%   9.3  10.9  10.0  13.1  13.5    0    0    0    0
 GitLab               3/3  0.0%   2.2   3.1   3.3   3.7   3.8    0    0    0    0
 Gmail                3/3  0.0%   1.8   2.1   1.9   2.4   2.5    0    0    0    0
 Google               3/3  0.0%   5.5  12.6  10.7  20.5  21.6    0    0    0    0
 Google Cloud         3/3  0.0%   2.2   2.4   2.4   2.5   2.6    0    0    0    0
 GoogleCloud          3/3  0.0%   1.8   2.1   2.1   2.3   2.3    0    0    0    0
 Grok                 3/3  0.0%   2.4   2.6   2.8   2.8   2.8    0    0    0    0
 Hetzner              3/3  0.0%   155   208   155   299   315    0    0    0    0
 HuggingFace          3/3  0.0%   3.3   3.3   3.3   3.4   3.4    0    0    0    0
 Instagram            3/3  0.0%   3.7  10.2   4.2  20.8  22.7    0    0    0    0
 iQIYI                3/3  0.0%  59.3  59.3  59.3  59.4  59.4    0    0    0    0
 KhanAcademy          3/3  0.0%   2.1   2.2   2.2   2.3   2.3    0    0    0    0
 Line                 3/3  0.0%   108   113   115   117   117    0    0    0    0
 Linode               3/3  0.0%  59.5  62.1  59.7  66.3  67.1    0    0    0    0
 Medium               3/3  0.0%   2.5   2.6   2.5   2.8   2.9    0    0    0    0
 MetaAI               3/3  0.0%   1.8   3.4   2.9   5.1   5.3    0    0    0    0
 Microsoft            3/3  0.0%  59.5  60.0  60.1  60.4  60.4    0    0    0    0
 Midjourney           3/3  0.0%   2.3   2.3   2.3   2.3   2.3    0    0    0    0
 Netflix              3/3  0.0%   5.1  13.7  10.1  24.3  25.9    0    0    0    0
 Netlify              3/3  0.0%   2.7   9.1   6.4  16.9  18.1    0    0    0    0
 Notion               3/3  0.0%   2.3   2.8   2.7   3.3   3.4    0    0    0    0
 NYTimes              3/3  0.0%   1.3   1.8   1.6   2.3   2.4    0    0    0    0
 OpenAI               3/3  0.0%   4.2  12.0   5.5  24.1  26.2    0    0    0    0
 Oracle               3/3  0.0%  59.5  61.3  61.9  62.3  62.3    0    0    0    0
 Outlook              3/3  0.0%  10.7  13.1  12.6  15.5  15.8    0    0    0    0
 PayPal               3/3  0.0%   1.8   2.2   2.1   2.6   2.7    0    0    0    0
 Perplexity           3/3  0.0%   3.0   3.2   3.0   3.5   3.6    0    0    0    0
 PlayStation          3/3  0.0%  63.1  68.2  70.1  71.3  71.4    0    0    0    0
 Poe                  3/3  0.0%   2.7   3.6   3.6   4.3   4.4    0    0    0    0
 PrimeVideo           3/3  0.0%   4.8   6.0   6.4   6.8   6.8    0    0    0    0
 ProtonMail           3/3  0.0%   144   154   157   161   161    0    0    0    0
 Reddit               3/3  0.0%   3.5   5.9   4.5   9.1   9.6    0    0    0    0
 Salesforce           3/3  0.0%  61.8  62.0  61.9  62.1  62.1    0    0    0    0
 Shopify              3/3  0.0%   3.1   7.4   4.0  14.1  15.2    0    0    0    0
 Signal               3/3  0.0%  10.7  13.8  12.6  17.4  17.9    0    0    0    0
 Skype                3/3  0.0%  40.6  86.2  70.7   140   147    0    0    0    0
 Slack                3/3  0.0%  27.2  29.6  28.4  32.8  33.3    0    0    0    0
 Sora                 3/3  0.0%   2.9   4.2   4.4   5.3   5.4    0    0    0    0
 Spotify              3/3  0.0%   4.4   7.4   8.7   9.0   9.0    0    0    0    0
 StackOverflow        3/3  0.0%   2.3   3.0   3.1   3.4   3.4    0    0    0    0
 Steam                3/3  0.0%  59.1  60.1  59.5  61.5  61.7    0    0    0    0
 Stripe               3/3  0.0%  27.5  29.2  28.2  31.6  32.0    0    0    0    0
 Teams                3/3  0.0%   2.1   4.7   2.4   9.1   9.8    0    0    0    0
 Telegram             3/3  0.0%   144   151   149   160   161    0    0    0    0
 TheGuardian          3/3  0.0%   2.1   9.6   9.8  16.1  16.8    0    0    0    0
 TikTok               3/3  0.0%   5.5  12.3   5.7  23.7  25.7    0    0    0    0
 Trello               3/3  0.0%   1.7   2.6   2.9   3.0   3.0    0    0    0    0
 TVB Anywhere         3/3  0.0%   3.0   9.4   3.9  19.5  21.3    0    0    0    0
 Twitch               3/3  0.0%   4.2   5.5   4.2   7.9   8.3    0    0    0    0
 Twitter/X            3/3  0.0%   2.3   6.6   2.6  13.6  14.8    0    0    0    0
 Twitter/X            3/3  0.0%   3.9   5.2   4.7   6.8   7.1    0    0    0    0
 Udemy                3/3  0.0%   3.1   8.9   4.8  17.4  18.8    0    0    0    0
 Vercel               3/3  0.0%   2.1   7.5   3.3  15.7  17.1    0    0    0    0
 ViuTV                3/3  0.0%   2.3   2.5   2.5   2.6   2.7    0    0    0    0
 Vultr                3/3  0.0%  64.4  71.0  72.3  75.8  76.2    0    0    0    0
 WhatsApp             3/3  0.0%   1.7   1.9   1.9   2.2   2.2    0    0    0    0
 Wikipedia            3/3  0.0%   9.9  10.8  10.0  12.2  12.4    0    0    0    0
 Xbox                 3/3  0.0%  59.5  60.6  60.1  62.1  62.3    0    0    0    0
 YahooMail            3/3  0.0%   1.4   1.8   1.8   2.1   2.2    0    0    0    0
 YouTube              3/3  0.0%   5.7  12.9   7.6  23.5  25.3    0    0    0    0
 Zoom                 3/3  0.0%   2.2   2.5   2.6   2.7   2.7    0    0    0    0
----------------------------------就近节点测速----------------------------------
 位置            上传速度        下载速度        延迟            丢包率          
 Speedtest.net   2168.40 Mbps    4789.57 Mbps    2.67847ms       0.00% (Sent: 328/Dup: 0/Max: 327)
 电信Shanghai    9.86 Mbps       112.38 Mbps     256.45 ms       N/A             
 联通Shanghai    9.95 Mbps       5.32 Mbps       264.99 ms       N/A             
 移动Shanghai    9.22 Mbps       161.09 Mbps     267.40 ms       N/A             
--------------------------------------------------------------------------------
 花费          : 3 分 24 秒
 时间          : Wed Oct 28 05:46:36 PDT 2026
--------------------------------------------------------------------------------

Compared with BandwagonHost, can it serve as a low-cost alternative?

BandwagonHost is well-known among Chinese VPS users mainly due to the China-optimized routes of some plans, route product choices, and long-term reputation. The price of this VMRack machine is $13/month, and its configuration and 500 Mbps bandwidth are attractive in terms of price; this routing test also did not show the obvious detours typical of ordinary routes, so it indeed meets the conditions for a „low-cost alternative candidate.“

However, the two cannot be compared solely on CPU, memory, and bandwidth:

Comparison ItemVMRack Test Machine (this time)BandwagonHost Route-Optimized Plans
PriceThis machine is $13/monthSubject to specific plans and promotions
This network testThree-network routing and latency are good, but speed results from China nodes are inconsistentRequires testing based on specific routes, data centers, and time
BackupWithout backup or snapshotSome products have backup or snapshot capabilities; refer to specific plans
StabilityOnly initial data from the setup period is available this timeThis site has long-term usage experience, but differences remain across plans
Target AudienceFor those who can maintain it themselves and seek cost-effectiveness and route testingFor those who value route products and long-term usage experience more

My judgment is:VMRack can serve as a low-cost testing candidate besides BandwagonHost, but it cannot yet be written as a complete replacement for BandwagonHost.If it is for a personal blog, test site, low-traffic WordPress, or development environment, you can consider monthly payment for testing first; if it is for a formal commercial website, backup, monitoring, migration, and provider failure handling should all be included in the cost.

Is this VMRack configuration suitable for purchase?

Relatively suitable for:

  • Low-traffic websites that need a Los Angeles data center and have users in mainland China or Asia
  • WordPress personal blogs, foreign trade showcase sites, development and test sites
  • Users who want to test premium routes but do not want to directly purchase more expensive plans
  • People who can install the Debian environment, configure web services, and handle security updates and backups themselves

Not very suitable for:

  • People who have no server maintenance experience and want to use it directly after activation
  • Critical businesses that require high availability, official backups, and clear SLAs
  • High-concurrency online stores, membership systems, video sites, or large file distribution
  • Projects that require guaranteed high-speed access in Hong Kong and every region of mainland China
  • Plans to use a 20 GB disk simultaneously for the website, logs, backups, and a large number of media files

Recommendations before purchase and use

First, it is recommended to pay monthly at the beginning, and not to pay long-term fees for routes and stability that have not yet been verified. Second, immediately change the SSH key or password after activation, close unnecessary ports, and configure the firewall and automatic security updates. Third, back up website data separately to object storage or another server; do not put the only backup on the same VPS.

If you plan to install WordPress, you can use the one designed specifically for WordPressWP Panel, which can directly manage Nginx, PHP, MariaDB, Redis, and caching. This 2C2G VMRack configuration is fully sufficient for running WP Panel; however, when running a small WordPress site with 2 GB of memory, you should still avoid installing too many heavyweight plugins at the same time, and monitor PHP-FPM, the database, and disk space.

Final evaluation

The core competitiveness of this VMRack Los Angeles 2-core 2G custom VPS is its price, 500 Mbps bandwidth, and the relatively good three-network routing in the initial test. In terms of performance, disk I/O and international bandwidth have no obvious shortcomings; in terms of network, the paths and latency toward China Telecom, China Unicom, and China Mobile are commendable.

However, the speed test results for mainland China nodes show clear divergence, and the machine„s uptime is also too short to prove long-term stability for now. Therefore, I will not directly list it as a “BandwagonHost alternative,„ but rather give it the evaluation of “worth monthly payment for testing, suitable for low-traffic and non-critical projects.„

If, after continuous observation for one to two weeks, repeated tests during evening peak hours still remain stable, and actual download, packet loss, jitter, and return route results for China's three major networks are supplemented, then VMRack will be more qualified to enter the more formal recommendation range for low-price CN2 GIA VPS.

Rate this article post
Previous Website frontend all normal, yet four backdoors lying in the server: a WordPress infection investigation record Continue reading content around the same timeline.

Join the discussion

Welcome to share experiences, ask questions, or point out areas that need updating.