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.
Comments
Thats faulty node/cpu then isnt it and case for support/complain isnt it? Basically its shit
Its just oversold and overpopulated, hence why its giving you these low results. It is to be expected from certain providers.
When you complain - they will likely move you over to a less crowded node, which will run fine/better for a few weeks, until that gets overpopulated as well.
It is not faulty one, it just crowded & busy neighbour.
What location is that ?
IC so kind of normal? Can happen anytime anywhere?
Opened support ticket, will see what they will say.
Hopefully @dustinc can look in to it.
this, explained by @NobodyInteresting
Bench.sh
Is that okay? Guess not the best test script?
CPU Model : AMD Ryzen 9 3900X 12-Core Processor
CPU Cores : 2
CPU Frequency : 3792.874 MHz
CPU Cache : 512 KB
Total Disk : 53.5 GB (8.0 GB Used)
Total Mem : 2992 MB (332 MB Used)
Total Swap : 3071 MB (7 MB Used)
System uptime : 0 days, 18 hour 46 min
Load average : 0.00, 0.00, 0.00
OS : Ubuntu 20.04.3 LTS
Arch : x86_64 (64 Bit)
Kernel : 5.4.0-84-generic
TCP CC : cubic
Virtualization : KVM
Organization : AS36352 ColoCrossing
Location : Buffalo / US
Region : New York
I/O Speed(1st run) : 1.2 GB/s
I/O Speed(2nd run) : 1.2 GB/s
I/O Speed(3rd run) : 1.2 GB/s
Average I/O speed : 1228.8 MB/s
Node Name Upload Speed Download Speed Latency
Speedtest.net 822.83 Mbps 809.21 Mbps 41.82 ms
Shanghai CT 193.69 Mbps 552.48 Mbps 212.74 ms
Shanghai CU 79.50 Mbps 469.35 Mbps 243.19 ms
Guangzhou CT 192.32 Mbps 293.77 Mbps 235.69 ms
Guangzhou CU 155.57 Mbps 598.63 Mbps 263.82 ms
Hongkong CN 102.26 Mbps 883.33 Mbps 226.43 ms
Singapore SG 190.56 Mbps 387.94 Mbps 257.59 ms
Tokyo JP 56.65 Mbps 176.04 Mbps 151.04 ms
I believe it's Dallas
Okay or not depend on your usage case. You can install a simple lemp stack on it and do some bench with PHP & load testing. Compare the waterfall of each provider & configuration.
## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ##
Yet-Another-Bench-Script
v2021-06-05
https://github.com/masonr/yet-another-bench-script
## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ##
Sat 18 Sep 2021 01:20:58 PM BST
Basic System Information:
Processor : AMD Ryzen 9 3900X 12-Core Processor
CPU cores : 2 @ 3792.874 MHz
AES-NI : ✔ Enabled
VM-x/AMD-V : ✔ Enabled
RAM : 2.9 GiB
Swap : 3.0 GiB
Disk : 52.6 GiB
fio Disk Speed Tests (Mixed R/W 50/50):
iperf3 Network Speed Tests (IPv4):
Provider | Location (Link) | Send Speed | Recv Speed
| | |
Clouvider | London, UK (10G) | 371 Mbits/sec | 306 Mbits/sec
Online.net | Paris, FR (10G) | 176 Mbits/sec | 207 Mbits/sec
WorldStream | The Netherlands (10G) | 392 Mbits/sec | 92.9 Mbits/sec
Biznet | Jakarta, Indonesia (1G) | busy | busy
Clouvider | NYC, NY, US (10G) | 835 Mbits/sec | 648 Mbits/sec
Velocity Online | Tallahassee, FL, US (10G) | 736 Mbits/sec | 320 Mbits/sec
Clouvider | Los Angeles, CA, US (10G) | 582 Mbits/sec | 239 Mbits/sec
Iveloz Telecom | Sao Paulo, BR (2G) | 133 Mbits/sec | 125 Mbits/sec
Geekbench 5 Benchmark Test:
Test | Value
|
Single Core | 943
Multi Core | 1670
Full Test | https://browser.geekbench.com/v5/cpu/9906991
i know someone who will have alot of words for this.
definitely contact support and complain.
in any case, it shoule be superior to your last vps, but it still isnt great for a ryzen.
a> @Shakib said:
Is this the same 11900K from the previous thread? Also why push this hard on rocket lake? nothing impressive about it if anything a small 5-10% perf gain at the cost of extra power and less cores to go with it. It makes no sense pal unless you're really trying to fill out some nodes.
Funny thing, i am kinda not pushing it. All our nodes automatically boosting themselves upto 5.3 GHz and not even overheating. My guess would be it's the better cooling in Carrier-1 DC that's helping our nodes to boost up when it can.
Our operating cost still stays the same even if our nodes uses a bit more power wattage. I don't see why i should disable it from the Dell BIOS.
But not if the Ryzen is suffering from overheating because there is no enterprise-grade chassis that can adequately cool them (that I know of).
So much inexperience..
yeh but would expect it to be properly cooled
its https://www.racknerd.com/new-york-datacenter location
Server chassis manufacturers do not design chassis for desktop CPU's as such with the limited options available for the high-end Ryzen CPU boards that even can be retrofitted into server chassis, they run hot.
I have to say, this is just my personal opinion but Ryzen in a server is a marketing gimmick at best, the exception would be with BuyVM where Fran had a MASSIVE amount of experience and knowledge so is able to do a very high end job of putting things together.
Server chassis have very limited space, intel gets away with it as they use the same socket for the i9's as some of the server CPU's, AMD do not.
How does it limit it, and how do you monitor it?
Simple and best explanation.
👍🏼👍🏼
After support migrated me to new node, as well they migrated to same one at first by mistake so got free month for that, then migrated again to new one.
## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ##
Yet-Another-Bench-Script
v2021-06-05
https://github.com/masonr/yet-another-bench-script
## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ##
Sat Sep 18 18:14:51 BST 2021
Basic System Information:
Processor : AMD Ryzen 9 3900X 12-Core Processor
CPU cores : 2 @ 3792.872 MHz
AES-NI : ✔ Enabled
VM-x/AMD-V : ✔ Enabled
RAM : 2.9 GiB
Swap : 3.0 GiB
Disk : 51.2 GiB
fio Disk Speed Tests (Mixed R/W 50/50):
iperf3 Network Speed Tests (IPv4):
Provider | Location (Link) | Send Speed | Recv Speed
| | |
Clouvider | London, UK (10G) | 903 Mbits/sec | 170 Mbits/sec
Online.net | Paris, FR (10G) | 902 Mbits/sec | 180 Mbits/sec
WorldStream | The Netherlands (10G) | 884 Mbits/sec | 204 Mbits/sec
Biznet | Jakarta, Indonesia (1G) | busy | busy
Clouvider | NYC, NY, US (10G) | 948 Mbits/sec | 937 Mbits/sec
Velocity Online | Tallahassee, FL, US (10G) | 903 Mbits/sec | 517 Mbits/sec
Clouvider | Los Angeles, CA, US (10G) | 912 Mbits/sec | 390 Mbits/sec
Iveloz Telecom | Sao Paulo, BR (2G) | 838 Mbits/sec | 120 Mbits/sec
Geekbench 5 Benchmark Test:
Test | Value
|
Single Core | 1269
Multi Core | 2469
Full Test | https://browser.geekbench.com/v5/cpu/9913106
looks better, and a bonus free month is nice
hopefully performance remains solid
@xaos Thank you for choosing RackNerd and for your business! I'm happy to see that we got you sorted out in the end.
I took a look at the original benchmarks you posted pre-migration in this thread, and while post-migration results are certainly an improvement, the benchmarks even pre-migration were not too shabby (and even at the older benchmark, in my opinion, was already more than suitable for real world use cases that are in line with a VPS environment). Generally speaking (and this is a very generic rule of thumb, for reasons I'll explain below), when a Ryzen VPS host node is at capacity (not overloaded, but rather - simply at capacity), you'll notice a filled 3900X VPS host node to be stabilized at around the 900 to 1000 GB5 Single Core range, depending on how busy the host node is at the time of the benchmark test (and if any other neighbors happen to also be running benchmarks at the same time as you, among other various factors). This can be higher or lower at any given time and the benchmark results are relative to different factors, after all, this is a VPS environment whereby CPU and Disk I/O resources are shared with neighboring VM's, and not a bare metal environment where such resources are fully dedicated. All in all, we monitor and maintain our VPS infrastructure very well and ensure that our customers are able to obtain the performance that they need in order to be successful in their real world environments.
In this particular instance, we didn't mind migrating you over to a newer host node as that is what you had requested in your support ticket, and I'm happy it turned out to be favorable for you.
If we ever fall short of your expectations or if you ever need anything in the future - feel free to reach out to me at [email protected] - I'm always happy to help
@dustinc
The customer support at RackNerd is simply amazing. One of the most premium support at LET.
👍🏼👍🏼
Having said that … the fastest single core vps I have is from HostCram
Happy to be a customer from RackNerd as well as Hostcram
Thank You @dev_vps -- appreciate you recognizing that, we try our best!
There's always a possibility for a hiccup, but we'll be here to help in the event there is. We're committed to providing stable services, competitive pricing with awesome support.
I agree 💯 percent
The stability and support “forced” me to upgrade the VPS.