Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!


Shells Virtual Desktop
BMail.ag - Secure Email Service
Server.net
CPLicense.net
VPS Server
Buy VPN
Vultr
VMs for AI
HostDare
ReliableSite White-Label Dedicated Hosting for Resellers
InterServer VPS
BMail.ag - Secure Email Service
Best VPN
High-Performance Bare Metal Server Solutions
Karvl.com
Server Mania Cloud Hosting
DataWagon Hosting
AlphaVPS Hosting
Evoxt.com
Clouvider
VPS Hosting with NVMe
Residential IPs in the US & 4G Mobile Proxies in EU & US with Unlimited Bandwidth
ReliableSite White-Label Dedicated Hosting for Resellers
Rabisu - Hosting Solutions
Shells Virtual Desktop
New on LowEndTalk? Please Register and read our Community Rules.

All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.

[Seeking Assistance] Hong Kong server with sub-0.5ms latency to Cloudflare.

Requirements: Must not connect through CDN77 (already have). Please provide MTR and provider name for reward if found.
thanks

«1

Comments

  • Sub 0.5 ms? I doubt you can achieve that even if you server is hooked up in the same building as Cloudflare's.

  • @dedipromo said:
    Sub 0.5 ms? I doubt you can achieve that even if you server is hooked up in the same building as Cloudflare's.

    Yes, we are currently using Datapacket, but my boss is not satisfied and requires a faster server. If you can find one, please let me know. Thank you! 😃

  • @dedipromo said:
    Sub 0.5 ms? I doubt you can achieve that even if you server is hooked up in the same building as Cloudflare's.

    Pinging my own router gets me ~0.7ms...

    OP has absolutely no chance (unless they move their services to CF Workers). Bestest of luck when telling your boss that.

  • @dedipromo said:
    Sub 0.5 ms? I doubt you can achieve that even if you server is hooked up in the same building as Cloudflare's.

    May be same rack same switch, even same building packets still need to pass multiple routers and switch too…

  • beanman109beanman109 Member, Host Rep, Megathread Squad

    HostHatch HongKong is close if 0.6-0.7 counts

    root@hk1:~# ping 1.1.1.1
    PING 1.1.1.1 (1.1.1.1) 56(84) bytes of data.
    64 bytes from 1.1.1.1: icmp_seq=1 ttl=58 time=0.739 ms
    64 bytes from 1.1.1.1: icmp_seq=2 ttl=58 time=0.659 ms
    64 bytes from 1.1.1.1: icmp_seq=3 ttl=58 time=0.632 ms
    64 bytes from 1.1.1.1: icmp_seq=4 ttl=58 time=0.659 ms
    
  • @sillycat said:

    @dedipromo said:
    Sub 0.5 ms? I doubt you can achieve that even if you server is hooked up in the same building as Cloudflare's.

    Pinging my own router gets me ~0.7ms...

    OP has absolutely no chance (unless they move their services to CF Workers). Bestest of luck when telling your boss that.

    I think it's hard, too
    But I can only do it
    If you find something new, let me know. I'll pay you

  • @beanman109 said:
    HostHatch HongKong is close if 0.6-0.7 counts

    root@hk1:~# ping 1.1.1.1
    PING 1.1.1.1 (1.1.1.1) 56(84) bytes of data.
    64 bytes from 1.1.1.1: icmp_seq=1 ttl=58 time=0.739 ms
    64 bytes from 1.1.1.1: icmp_seq=2 ttl=58 time=0.659 ms
    64 bytes from 1.1.1.1: icmp_seq=3 ttl=58 time=0.632 ms
    64 bytes from 1.1.1.1: icmp_seq=4 ttl=58 time=0.659 ms
    

    HH is good, but he's over 0.5

  • Datapacket and Cloudflare are in the same data center, so it’ll be hard to find better results than Datapacket..

    Thanked by 2sillycat mandala
  • sh97sh97 Member, Host Rep

    https://globalping.io?measurement=TQtWlCZMIQhQxJlb&display=table

    None below 0.5ms, Leaseweb HK and Hytron are close at 0.55-0.60ms

  • fmxmfmxm Member

    No one can do this better than Datapacket.

  • OhJohnOhJohn Member
    edited August 2025

    The only one that could pull of a sub 0.5ms to Cloudflare is ... Cloudflare.

    I'm getting around 0.7ms from HH, and 0.6 to 0.7 from LW in HK to 1.1.1.1.

    1. 31.56.62.1                                                                                                                               0.0%   434   46.2  36.0   1.4 297.7  51.2
     2. 103.73.67.2                                                                                                                              0.0%   434    0.4   1.0   0.2  37.3   3.7
     3. e5.hk-eqxhk1-sw4.globalsecurelayer.com                                                                                                   0.0%   434    0.3   0.2   0.2   0.5   0.1
     4. po10.hk-eqxhk1-bb6.globalsecurelayer.com                                                                                                 0.0%   434    0.3   0.2   0.2   0.6   0.1
     5. 160.202.164.119                                                                                                                          0.0%   434    7.2   3.4   0.8  26.3   4.2
     6. 103.22.203.231                                                                                                                           0.0%   433    1.4   2.4   1.2  26.8   3.1
     7. one.one.one.one                                                                                                                          0.0%   433    0.7   0.7   0.7   1.0   0.1
    
    1. (oneprovider on leaseweb)                                                                                                                           0.2%   399    1.0   0.4   0.3   7.3   0.5
     2. ae-2.br01.hkg-12.hk.leaseweb.net                                                                                                         0.0%   399    0.4   0.9   0.3  21.5   1.9
     3. ae-101.bb11.hkg-12.leaseweb.net                                                                                                          0.0%   399    0.6   0.8   0.4  19.4   1.8
     4. xe-2-3-5.bb04.ams-01.leaseweb.net                                                                                                        0.0%   399    0.4   1.7   0.4  27.8   4.0
     5. 13335.hkg.equinix.com                                                                                                                    0.0%   399   19.1   4.3   1.1  33.4   4.9
     6. 103.22.203.23                                                                                                                            0.0%   399    1.6   4.5   0.9  72.4   8.9
     7. one.one.one.one                                                                                                                          0.2%   399    1.0   0.9   0.7   4.3   0.2
    

    As @sh97 showed (and doing it the right way btw), the best would be Hytron, but even that is >0.5ms but below 0.6ms.

    Thanked by 3nopint3 sh97 mandala
  • i don’t know if it’s a personal thing or some business related thing,
    in my opinion, get a very close network location to cf and have a high traffic volume through them, why not have a talk with cloudflare, they may offer beer lantency results.
    if it’s just for personal use, in the same dc is good enough.

  • @sh97 said:
    https://globalping.io?measurement=TQtWlCZMIQhQxJlb&display=table

    None below 0.5ms, Leaseweb HK and Hytron are close at 0.55-0.60ms

    We have tried leaseweb and hytron.
    We didn't choose them because they were larger than 0.5 ms.
    If you can find something better, you can also tell us and I'll pay you.

  • sh97sh97 Member, Host Rep

    @zhongyang said:

    @sh97 said:
    https://globalping.io?measurement=TQtWlCZMIQhQxJlb&display=table

    None below 0.5ms, Leaseweb HK and Hytron are close at 0.55-0.60ms

    We have tried leaseweb and hytron.
    We didn't choose them because they were larger than 0.5 ms.
    If you can find something better, you can also tell us and I'll pay you.

    Like the others said, I think it's impossible.

    As you can see from the globalping test I did, over 75 probes were used, none of them were able to go below 0.5ms, the closest being Hytron and Leaseweb.

    Have a look with Azure/Google Cloud/AWS, in case they are able to achieve that.

    Good luck in your search.

  • I mean at this point why not just get Cloudflare PNI yourself?

  • MannDudeMannDude Patron Provider, Veteran
    edited August 2025

    I'm genuinely curious as to the case where 0.6ms is no good for you use-case but 0.5ms or 0.4ms would make a noticeable difference.

    Is this some sort of algorithmic trading setup or something?

    Curious, because sometimes I obsess over minor details that have no real world benefit. I'm wondering if I need to be obsessing over this, too. :D

  • fmxmfmxm Member

    @MannDude said:
    I'm genuinely curious as to the case where 0.6ms is no good for you use-case but 0.5ms or 0.4ms would make a noticeable difference.

    Is this some sort of algorithmic trading setup or something?

    Curious, because sometimes I obsess over minor details that have no real world benefit. I'm wondering if I need to be obsessing over this, too. :D

    He might be a scalper and needs to compete with peers for items.

  • likely trading related

    Thanked by 1ducpt254
  • Guess it may be related to quantitative trading?

  • no one can find it?
    I can increase the stake to 30U

  • @zhongyang said:
    no one can find it?
    I can increase the stake to 30U

    If you have any new discoveries please send them to my telegrambot @zytwoway_bot

  • That is near impossible to have 0.5ms unless your "server" is running inside of cloudflare servers

  • serverpointserverpoint Member, Patron Provider

    From my experience working with routers for decades now, even with the latest, fastest packet processors there are, sub 0.5ms is unlikely.

    If you were connected to the same switch as Cloudflare's servers, you could get 0.1xx to 0.2.

    I can get 0.2 to 0.4 from one server connected to a switch to a server in another switch, but first crossing through a core switch. That involves several SFP+/qsfp+ modules and at least three switches processing that packet.

    But once that packet crosses a router and has to somehow go into "the public internet", even if it's just a few simple cables here and there within the same building crossing two or three routers and several switches, it will easily jump to beyond 0.5.

    To get sub 0.5ms to a Cloudflare server, you'd have to be connected to the switches owned and operated by Cloudflare.

  • If 0.1ms makes such a big difference for you and your boss then the only way is to talk directly with cloudflare. There's nothing else to do

    Thanked by 1oloke
  • zhongyangzhongyang Member
    edited September 2025

    @serverpoint said:
    From my experience working with routers for decades now, even with the latest, fastest packet processors there are, sub 0.5ms is unlikely.

    If you were connected to the same switch as Cloudflare's servers, you could get 0.1xx to 0.2.

    I can get 0.2 to 0.4 from one server connected to a switch to a server in another switch, but first crossing through a core switch. That involves several SFP+/qsfp+ modules and at least three switches processing that packet.

    But once that packet crosses a router and has to somehow go into "the public internet", even if it's just a few simple cables here and there within the same building crossing two or three routers and several switches, it will easily jump to beyond 0.5.

    To get sub 0.5ms to a Cloudflare server, you'd have to be connected to the switches owned and operated by Cloudflare.

    The service to which our server is going is on cloudflare argo
    Can you do this in Hong Kong?
    As a person, I don't think I can talk to Cloudflare
    No price limit
    Dedicated servers is ok

  • Unicom cloud ?
    Probe IP https://ipinfo.io/118.26.153.43


    PING 1.1.1.1 (1.1.1.1) 56(84) bytes of data.
    64 bytes from 1.1.1.1: icmp_seq=1 ttl=253 time=0.380 ms
    64 bytes from 1.1.1.1: icmp_seq=2 ttl=253 time=0.236 ms
    64 bytes from 1.1.1.1: icmp_seq=3 ttl=253 time=0.230 ms
    64 bytes from 1.1.1.1: icmp_seq=4 ttl=253 time=0.226 ms

    --- 1.1.1.1 ping statistics ---
    4 packets transmitted, 4 received, 0% packet loss, time 2ms
    rtt min/avg/max/mdev = 0.226/0.268/0.380/0.064 ms, ipg/ewma 0.710/0.330 ms

  • @Strikerr said:
    Unicom cloud ?
    Probe IP https://ipinfo.io/118.26.153.43


    PING 1.1.1.1 (1.1.1.1) 56(84) bytes of data.
    64 bytes from 1.1.1.1: icmp_seq=1 ttl=253 time=0.380 ms
    64 bytes from 1.1.1.1: icmp_seq=2 ttl=253 time=0.236 ms
    64 bytes from 1.1.1.1: icmp_seq=3 ttl=253 time=0.230 ms
    64 bytes from 1.1.1.1: icmp_seq=4 ttl=253 time=0.226 ms

    --- 1.1.1.1 ping statistics ---
    4 packets transmitted, 4 received, 0% packet loss, time 2ms
    rtt min/avg/max/mdev = 0.226/0.268/0.380/0.064 ms, ipg/ewma 0.710/0.330 ms

    could u please ping 172.65.175.107? thanks

  • I'm with everyone else; I don't see how you will achieve this. Peering with Cloudflare at an exchange point isn't going to accomplish it. Here's a trace doing exactly that:

    traceroute to 1.1.1.1 (1.1.1.1), 30 hops max, 60 byte packets
     1  172.96.185.28  0.155 ms  0.142 ms  0.131 ms
     2  64.120.118.93  0.092 ms  0.120 ms  0.142 ms
     3  64.120.119.8  0.541 ms 64.120.119.4  0.219 ms 64.120.119.8  0.562 ms
     4  31.31.38.194  0.214 ms  0.223 ms 31.31.38.36  0.190 ms
     5  36.255.56.48  16.259 ms  16.255 ms  15.642 ms
     6  103.22.203.75  7.832 ms 103.22.203.79  0.747 ms  0.709 ms
     7  103.22.203.71  2.665 ms 103.22.203.75  7.698 ms 103.22.203.23  0.886 ms
     8  1.1.1.1  0.644 ms  0.502 ms  0.469 ms
    

    Cloudflare has a private network interconnect service but I can't really comment on it: https://developers.cloudflare.com/network-interconnect/ or if anyone is using that in HK.

  • Nah, I don't think that's related to high-frequency trading. HFT is all about chasing ultra-low latency to the stock exchange, so it wouldn't really involve Cloudflare.

    Thanked by 1mandala
  • davidobrik567davidobrik567 Member
    edited September 2025

    cloudflare + datapacket is in equinix HK2, so 0.5ms is possible.

    GSL (customers such as HH) is in equinix HK1 + mega-I IDC, so latency will be higher.

    go look for providers in equinix HK2

    Thanked by 2Hitori0221 mandala
Sign In or Register to comment.