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.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Advin Servers - AMD Ryzen 9950X 10 Gbps VPS - Starting at $6/month for 3GB RAM
This discussion has been closed.
Comments
@advinservers Hi, can you send me the current yabs of the 2 packages Premium S and C2F in Johor, Malaysia? I am confused between these 2 packages, their prices are the same and I don't know which package to buy.
C2F
Premium S
@budi1413 @astleyne22 thank you both very much, in C2F malaysia package can you show me the network speed?
It would be the same between KVM Premium and Ryzen, because they are connected to the same network.
I bought the C2F package and here are the Yabs of the 3rd run
## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ##
Yet-Another-Bench-Script
v2025-04-20
https://github.com/masonr/yet-another-bench-script
## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ##
Mon Aug 18 17:12:16 UTC 2025
Basic System Information:
Uptime : 0 days, 0 hours, 5 minutes
Processor : AMD Ryzen 9 9950X 16-Core Processor
CPU cores : 2 @ 4291.932 MHz
AES-NI : ✔ Enabled
VM-x/AMD-V : ✔ Enabled
RAM : 7.5 GiB
Swap : 0.0 KiB
Disk : 237.9 GiB
Distro : AlmaLinux 9.6 (Sage Margay)
Kernel : 5.14.0-570.33.2.el9_6.x86_64
VM Type : KVM
IPv4/IPv6 : ✔ Online / ✔ Online
IPv6 Network Information:
ISP : Advin Services LLC
ASN : AS206216 Advin Services LLC
Host : Advin Services LLC
Location : Singapore, Central Singapore (01)
Country : Singapore
fio Disk Speed Tests (Mixed R/W 50/50) (Partition /dev/sda3):
iperf3 Network Speed Tests (IPv4):
iperf3 Network Speed Tests (IPv6):
Geekbench 6 Benchmark Test:
Test | Value
|
Single Core | 3019
Multi Core | 5284
Full Test | https://browser.geekbench.com/v6/cpu/13384806
YABS completed in 14 min 25 sec
Looks like the speed in Singapore is not 10gbps as I imagined
It is 10gbps. Network is shared unless you want to pay 200x more
It's difficult to get high speeds internationally through iperf3. A lot of the iperf3 servers in the benchmark are very far from the server and are in NA/EU.
You'll probably be able to come close to the maximum speed by running tests to servers in Asia, such as the result you saw with Leaseweb Singapore where upload speed was 8 Gbps and download speed was 6 Gbps.
Please note that we don't guarantee 10 Gbps speeds 24x7, it is a shared port (https://docs.advinservers.com/information/network#port-capacities).
So I opened a ticket about their Malaysia service having UDP being delivered out of order and they try to gaslight me that it is expected to be so while bluntly disregarding that other providers have no such issue.
yes, i'm talking about server speeds in asia, specifically in singapore. i don't understand why on VPS Benchmark https://www.vpsbenchmarks.com/compare/performances/yabs and also on yabs in the article, although in the same location in singapore, the Ryzen 9950x plans are on average 1gbps slower than the Epyc 9654 plans.
It's shared port, not dedicated. So it depends on your neighbors / server load / iperf3 servers load.
We're not gaslighting you, it is not unexpected for UDP packets to arrive out of order. TCP guarantees sequential packet delivery, UDP does not. It might be occurring due to the paths our primary upstream (Datapacket) takes.
You might get better/different results if you run it at a different time, since iperf3 servers can sometimes be loaded or certain networks can be congested.
Both of our EPYC 9654 and Ryzen hypervisors are connected to the same switches and same upstreams, so there should not be a network difference between them.
Wait 9950X arrived in Johor? Let me check my budget.
Even then TCP can also arrive out of order for a variety of reasons and sometimes those are out of the hands of the VPS provider and could happen somewhere upstream or downstream.
Networking is very very hard to troubleshoot and correctly pinpoint. Blaming the provider is just a cheap shot and it's a risk that you could be dead wrong about it.
Shouldn't you take that up with datapacket then? 3ms to Singapore and out of order for that short distance is not a red flag that need to be fixed?
Since when customer can directly contact hosting provider's network carrier to fix something? Pretty sure many network carriers ignores such contact. I think only gsl entertain such direct contact, I guess that is also why their network has less issue.
Also, I did provide them in the tickets various test results from different users across the globe and also controlled test results using public iperf3 test to prove these users have no issues on their end. Only for their staff to say it is expected aka they don't intent to do anything about it.
Now, another staff took over the ticket and say will look into it.
Out of order isn't really a big problem, especially because it's not abnormal. But if you believe the latency is abnormal, please send a bidirectional MTR.
You should send those bidirectional mtrs and those pcaps, including trying from different end points.
You need to do as much as the provider will because they will have flexibility.
Don't know where in my message I say contact the upstream provider but that's not the process.
Slow down, you're in good hands and you will get support, but you do need to put in the work.
Takes two to tango right?
You made that implication while also made assumptions. including assumptions that I didn't provide them sufficient informations. In case it isn't obvious, @SagnikS is the staff that dismissed out-of-order packets as an issue.
If one really think about it, it must be silly for other hosting providers like hosthatch, speedypage, kuroit, serverpoint etc. to put effort in ensuring their network reliability. Really must be foolish for the different network carriers to do the same.
If I didn't know better that even with TCP built-in packet reordering, packet reordering task alone incur extra latency and extra processing power. And due to TCP mechanism, it can incur extra bandwidth if too much out-of-order packets. It is intuitively not something considered as "not a problem". It definitely not something acceptable for latency sensitive applications which I have also kindly indicated in my ticket I am running a game server. Despite that, that said staff still nonchalantly mentioned TCP as if I have control over how the game server is programmed, it is udp game server if it isn't obvious which I also highlighted so in the said ticket.
Anyway, the owner aka @advinservers took over the ticket and seemingly agreed that is a problem and is looking into it.
@advinservers I think you should update all possible software for your server system to the latest version, this can bring new features, improve performance and network speed, such as: tools, applications, emulators, drivers, bios, firmware, operating system,...
We're currently looking into your ticket and will get back to you when we have a response.
I still don't know how you got an implication where...
...I said nothing about contacting the upstream provider, to contact the upstream provider I said blaming the provider (your hosting company, aka the company you raise tickets to) could be completely wrong as there could be an issue upstream and this may or may not be within their control.
You also haven't really stated what the fault is that your users are seeing/saying. You're running a game server, great! What are your players seeing/feeling
I really think you need to re-read my posts because I made no implications or assumptions. I said TCP
Which is completely valid and is the case with a lot of issues.
I used to work in telco by the way
I know a thing or two.
See, you're in great hands! Sagnik and Anush have got you covered!
We already use the latest versions of everything when possible.
When provisioning new hypervisors, we already update the BIOS and firmware to the latest stable versions. Moreover, we already use the latest stable versions of any software and operating systems that we use.
We also schedule occasional maintenance to bring our nodes up to date.
@MaxTakeba @Lu5ck
Hey everyone, it's alright, there's no need to continue this conversation. I think that this is just a misunderstanding.
We're already assisting the client in the ticket and we're looking into possible resolutions for their problem. If there is any information we need from them, we'll ask them through a support ticket.
We do put in a significant amount of money and time into optimizing and improving our network in Johor. Our network is a bit newer, so there can be issues from time to time. If we receive reports, then we do usually thoroughly look into it and try to act on it.
Thanks for your support!
Great, but now I see you are still using noVNC 1.4.0 for the server, which is over 2 years out of date. Please update to noVNC version 1.6.0
We'll look into updating it, but there hasn't been many improvements in the past 2 noVNC versions.
Also, please update Convoy Panel to the latest version 4.5.0
We use a custom fork of Convoy. It already has the latest changes from version 4.5.0.
I already mentioned above that we use the latest stable versions of software whenever possible. We do audits and update when we find it necessary.