AMD GENOA starting at (4c/16gb/200gb) @4,95€ VPS Deal at ph24 [...]

Khadeercasino

Brought one thank you gonna test it any thing can be get has a first customer? Bandwidth double ?

loay

Also kernel versions have different results in the test.

@maverick said:

@ProHosting24 said:

@wa44io4 said:
Picked up the Genoa L. Great performance, and I hope it remains consistent in the long run.

Consistency in performance is no issue as can be seen from previous offerings.

The Geekbench 6 Benchmark score is pretty much the same for Genoa M and L plan, expected more tbh.

Our 4 core multicore result was 5359
Yours with 6 cores resulted in 6640
6640/5359=1.23

Seems only like a 20% increase to me, that's odd.
But i don't know what the geekbench benchmarking results are based on.

Maybe someone else from LET can bring up some light onto this!

GB6 numbers for multicore (especially when you have plenty of cores) are totally useless and not to be trusted at all. Proven many times already... It's a shame that company that has benchmarking as its core business can't get that right, but it is what it is.

aaxaa

YABS good

# ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## #
#              Yet-Another-Bench-Script              #
#                     v2024-06-09                    #
# https://github.com/masonr/yet-another-bench-script #
# ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## #

Wed Sep  4 10:24:29 AM CEST 2024

Basic System Information:
---------------------------------
Uptime     : 0 days, 7 hours, 50 minutes
Processor  : AMD EPYC 9554 64-Core Processor
CPU cores  : 4 @ 3099.996 MHz
AES-NI     : ✔ Enabled
VM-x/AMD-V : ✔ Enabled
RAM        : 15.6 GiB
Swap       : 0.0 KiB
Disk       : 196.8 GiB
Distro     : Debian GNU/Linux 12 (bookworm)
Kernel     : 6.1.0-9-amd64
VM Type    : KVM
IPv4/IPv6  : ✔ Online / ✔ Online

IPv6 Network Information:
---------------------------------
ISP        : Oliver Horscht is trading as \"SYNLINQ\"
ASN        : AS44486 Oliver Horscht is trading as \"SYNLINQ\"
Host       : Prohosting24
Location   : Frankfurt am Main, Hesse (HE)
Country    : Germany

fio Disk Speed Tests (Mixed R/W 50/50) (Partition /dev/sda1):
---------------------------------
Block Size | 4k            (IOPS) | 64k           (IOPS)
  ------   | ---            ----  | ----           ----
Read       | 94.18 MB/s   (23.5k) | 938.37 MB/s  (14.6k)
Write      | 94.43 MB/s   (23.6k) | 943.31 MB/s  (14.7k)
Total      | 188.61 MB/s  (47.1k) | 1.88 GB/s    (29.4k)
           |                      |
Block Size | 512k          (IOPS) | 1m            (IOPS)
  ------   | ---            ----  | ----           ----
Read       | 1.75 GB/s     (3.4k) | 1.98 GB/s     (1.9k)
Write      | 1.84 GB/s     (3.6k) | 2.11 GB/s     (2.0k)
Total      | 3.60 GB/s     (7.0k) | 4.10 GB/s     (4.0k)

iperf3 Network Speed Tests (IPv4):
---------------------------------
Provider        | Location (Link)           | Send Speed      | Recv Speed      | Ping
-----           | -----                     | ----            | ----            | ----
Clouvider       | London, UK (10G)          | 1.90 Gbits/sec  | 1.97 Gbits/sec  | 11.5 ms
Eranium         | Amsterdam, NL (100G)      | 2.06 Gbits/sec  | 1.98 Gbits/sec  | 7.01 ms
Uztelecom       | Tashkent, UZ (10G)        | 319 Mbits/sec   | 1.79 Gbits/sec  | 93.5 ms
Leaseweb        | Singapore, SG (10G)       | 364 Mbits/sec   | 1.11 Gbits/sec  | 157 ms
Clouvider       | Los Angeles, CA, US (10G) | 373 Mbits/sec   | 1.18 Gbits/sec  | 139 ms
Leaseweb        | NYC, NY, US (10G)         | 376 Mbits/sec   | 1.87 Gbits/sec  | 83.9 ms
Edgoo           | Sao Paulo, BR (1G)        | 293 Mbits/sec   | 899 Mbits/sec   | 190 ms

iperf3 Network Speed Tests (IPv6):
---------------------------------
Provider        | Location (Link)           | Send Speed      | Recv Speed      | Ping
-----           | -----                     | ----            | ----            | ----
Clouvider       | London, UK (10G)          | 1.90 Gbits/sec  | 1.94 Gbits/sec  | 11.5 ms
Eranium         | Amsterdam, NL (100G)      | 2.04 Gbits/sec  | 1.96 Gbits/sec  | 7.01 ms
Uztelecom       | Tashkent, UZ (10G)        | 203 Mbits/sec   | 1.81 Gbits/sec  | 91.7 ms
Leaseweb        | Singapore, SG (10G)       | 321 Mbits/sec   | 1.13 Gbits/sec  | 156 ms
Clouvider       | Los Angeles, CA, US (10G) | 383 Mbits/sec   | 1.24 Gbits/sec  | 139 ms
Leaseweb        | NYC, NY, US (10G)         | 352 Mbits/sec   | 1.86 Gbits/sec  | 82.1 ms
Edgoo           | Sao Paulo, BR (1G)        | 318 Mbits/sec   | 901 Mbits/sec   | 189 ms

Geekbench 6 Benchmark Test:
---------------------------------
Test            | Value
                |
Single Core     | 1873
Multi Core      | 5775
Full Test       | https://browser.geekbench.com/v6/cpu/7624708

YABS completed in 12 min 37 sec

Reiki

Hi, im interested using it for our Encoder Farm, do you offer Crypto payment options?

ProHosting24

@senos said:

@ProHosting24 said:

@harris1111 said:

@asthen said:

@harris1111 said:

@asthen said:

@lirrr said:
what is cpu fair-use policy

@ProHosting24 here bruh, I also want to know cpu fair use policy

I thought that with LET user, FUP is the same between low-end provider. For example, you buy 4c CPU and run 100% x 4 core at 3-4 hour/day it's fine. but running it at 100% 24/7/30 is banned

Yes, nobody expect to run 24/7 , we don't know if it's 1 or 2 or 3 hours , just need some clarification, it's not funny being suspended suddenly.

I think that means, just use your VPS as your daily workload (some heavy job when building binary, or encode some data) and then idle, don't let it run 24/7 100% and you will be fine.

Most customers are very happy with our way of dealing with high CPU utilization.
However, I am surprised that there are not so many reports from former satisfied users in this post regarding such things ^^

Is it possible that they are sound asleep in different time zones? Laughing

Well, we are based in germany 🇩🇪 👋

lowendlurker

I did a reboot of one older machine, it came back up on the new node.
For "some" reason the ssh fingerprint changed, why did it change?
It's a VM, what did I miss?
Just wondering on a technical side, if anyone has quick answer for me to understand

harris1111

@Khadeercasino said:
Brought one thank you gonna test it any thing can be get has a first customer? Bandwidth double ?

it FUP, just don't abuse the bandwidth

ProHosting24

@lowendlurker said:
I did a reboot of one older machine, it came back up on the new node.
For "some" reason the ssh fingerprint changed, why did it change?
It's a VM, what did I miss?
Just wondering on a technical side, if anyone has quick answer for me to understand

This is an issue based on our implementation of cloud-init with proxmox.
But we are going to migrate to openstack anyway soon.

Staging is up since Oct 23 now and Prod should be released somewhere around Nov/Dec 24

ProHosting24

@Khadeercasino said:
Brought one thank you gonna test it any thing can be get has a first customer? Bandwidth double ?

Hi!

Sorry, there are no first customer benefits.

mw

are you guys prepaid-host

faleddo

@mw said:
are you guys prepaid-host

prepaid-host use epyc milan while prepaidhost24 use epyc genoa

ProHosting24

@mw said:
are you guys prepaid-host

Hey, technically and legally we are not.
I proposed to prepaid-host to merge our ceph infrastructure as i was tired of fixing stuff in that cluster with my hands cuffed most times.

Now when deploying a VM with us you will ALWAYS be deployed on ph24 hypervisors with ph24 quality assurance and monitoring. We only further make sure that prepaid-host hosts are capable enough of also processing our ceph workload, which is distributed over our whole rack at the Maincubes rn.

That, for example, is the reason why we offer Genoa nodes, although prepaid-host does not.

Here a picture of our rack attached:

mw

@ProHosting24 said:

@mw said:
are you guys prepaid-host

Hey, technically and legally we are not.
I proposed to prepaid-host to merge our ceph infrastructure as i was tired of fixing stuff in that cluster with my hands cuffed most times.

Now when deploying a VM with us you will ALWAYS be deployed on ph24 hypervisors with ph24 quality assurance and monitoring. We only further make sure that prepaid-host hosts are capable enough of also processing our ceph workload, which is distributed over our whole rack at the Maincubes rn.

That, for example, is the reason why we offer Genoa nodes, although prepaid-host does not.

Here a picture of our rack attached:

i see, because i saw the same 24GB offer, which i prepaid annually with prepaid-host, and i got the 16GB offer you posted from you guys

Dylhost

@remy said:

@ProHosting24 said:

@Dylhost said:
According to past threads 'fair use' is '2TB'.

Hey, we have increased our commitment.

So is 3TB okay for L plan ?

2TB was the assumed for small plan, so I assume 4TB back then for L?

Considering uplink has doubled I would assume 4TB is definitely fine for L now

Note this isn't off shore, but not to assume your use case....

ProHosting24

@Dylhost said:

@remy said:

@ProHosting24 said:

@Dylhost said:
According to past threads 'fair use' is '2TB'.

Hey, we have increased our commitment.

So is 3TB okay for L plan ?

2TB was the assumed for small plan, so I assume 4TB back then for L?

Considering uplink has doubled I would assume 4TB is definitely fine for L now

Note this isn't off shore, but not to assume your use case....

Yes thats fine

Khadeercasino

@ProHosting24 can I buy more vps quantity or it is limited with 1 per customer ?

emgh

@maverick said:

@ProHosting24 said:

@wa44io4 said:
Picked up the Genoa L. Great performance, and I hope it remains consistent in the long run.

Consistency in performance is no issue as can be seen from previous offerings.

The Geekbench 6 Benchmark score is pretty much the same for Genoa M and L plan, expected more tbh.

Our 4 core multicore result was 5359
Yours with 6 cores resulted in 6640
6640/5359=1.23

Seems only like a 20% increase to me, that's odd.
But i don't know what the geekbench benchmarking results are based on.

Maybe someone else from LET can bring up some light onto this!

GB6 numbers for multicore (especially when you have plenty of cores) are totally useless and not to be trusted at all. Proven many times already... It's a shame that company that has benchmarking as its core business can't get that right, but it is what it is.

Links?

ProHosting24

@Khadeercasino said:
@ProHosting24 can I buy more vps quantity or it is limited with 1 per customer ?

You can buy more.
At least as long as stock is available :'D

Khadeercasino

@ProHosting24 said:

@Khadeercasino said:
@ProHosting24 can I buy more vps quantity or it is limited with 1 per customer ?

You can buy more.
At least as long as stock is available :'D

Thats good way to go thank you!

maverick

@emgh said:

@maverick said:

@ProHosting24 said:

@wa44io4 said:
Picked up the Genoa L. Great performance, and I hope it remains consistent in the long run.

Consistency in performance is no issue as can be seen from previous offerings.

The Geekbench 6 Benchmark score is pretty much the same for Genoa M and L plan, expected more tbh.

Our 4 core multicore result was 5359
Yours with 6 cores resulted in 6640
6640/5359=1.23

Seems only like a 20% increase to me, that's odd.
But i don't know what the geekbench benchmarking results are based on.

Maybe someone else from LET can bring up some light onto this!

GB6 numbers for multicore (especially when you have plenty of cores) are totally useless and not to be trusted at all. Proven many times already... It's a shame that company that has benchmarking as its core business can't get that right, but it is what it is.

Links?

https://chatgpt.com/share/71b6c117-f4f4-43b4-a551-13cdb635c6ea

dev_vps

@maverick said:

@emgh said:

@maverick said:

@ProHosting24 said:

@wa44io4 said:
Picked up the Genoa L. Great performance, and I hope it remains consistent in the long run.

Consistency in performance is no issue as can be seen from previous offerings.

The Geekbench 6 Benchmark score is pretty much the same for Genoa M and L plan, expected more tbh.

Our 4 core multicore result was 5359
Yours with 6 cores resulted in 6640
6640/5359=1.23

Seems only like a 20% increase to me, that's odd.
But i don't know what the geekbench benchmarking results are based on.

Maybe someone else from LET can bring up some light onto this!

GB6 numbers for multicore (especially when you have plenty of cores) are totally useless and not to be trusted at all. Proven many times already... It's a shame that company that has benchmarking as its core business can't get that right, but it is what it is.

Links?

https://chatgpt.com/share/71b6c117-f4f4-43b4-a551-13cdb635c6ea

I am working on a multi core benchmarking program.

GB5 is much more accurate, especially for cases vCores >= 6, as compared to GB6

dev_vps

That's why I post both GB5 and GB6 for benchmark tests in the YABS thread.

example
4x vCore Ryzen 7950x / 2gb ram

Geekbench 5 Benchmark Test:
---------------------------------
Test            | Value
                |
Single Core     | 1609
Multi Core      | 4515
Full Test       | https://browser.geekbench.com/v5/cpu/22756351

Geekbench 6 Benchmark Test:
---------------------------------
Test            | Value
                |
Single Core     | 2243
Multi Core      | 6178
Full Test       | https://browser.geekbench.com/v6/cpu/7245377

dev_vps

@ProHosting24
Here is benchmark comparison for my VPS between previous node and current AMD Genoa Node

-- before 
 CPU Model          : AMD EPYC 7543 32-Core Processor
 CPU Cores          : 4 @ 2800.000 MHz
 CPU Cache          : 512 KB
 AES-NI             : ✔ Enabled
 VM-x/AMD-V         : ✔ Enabled
 Total Disk         : 196.8 GB (3.2 GB Used)
 Total RAM          : 15.6 GB (94.2 MB Used)

-- after
 CPU Model          : AMD EPYC 9554 64-Core Processor
 CPU Cores          : 4 @ 3099.996 MHz
 CPU Cache          : 512 KB
 AES-NI             : ✔ Enabled
 VM-x/AMD-V         : ✔ Enabled
 Total Disk         : 196.8 GB (3.0 GB Used)
 Total RAM          : 15.6 GB (94.8 MB Used)

-- before
Geekbench 5 Benchmark Test:
---------------------------------
Test            | Value
                |
Single Core     | 1019
Multi Core      | 3345
Full Test       | https://browser.geekbench.com/v5/cpu/22836195

-- after
Geekbench 5 Benchmark Test:
---------------------------------
Test            | Value
                |
Single Core     | 1344
Multi Core      | 4778
Full Test       | https://browser.geekbench.com/v5/cpu/22837535

-- before
Geekbench 6 Benchmark Test:
---------------------------------
Test            | Value
                |
Single Core     | 1303
Multi Core      | 3979
Full Test       | https://browser.geekbench.com/v6/cpu/7620880

-- after
Geekbench 6 Benchmark Test:
---------------------------------
Test            | Value
                |
Single Core     | 1779
Multi Core      | 5865
Full Test       | https://browser.geekbench.com/v6/cpu/7626652

dev_vps

@ProHosting24
one picture == 999 words
Windows 11 Enterprise OS

ProHosting24

@dev_vps said:
@ProHosting24
one picture == 999 words
Windows 11 Enterprise OS

These windows hosts run very smooth.
I have used them to do advanced support on the way working from phone. I can recommend 100%

dev_vps

@ProHosting24 said:

@dev_vps said:
@ProHosting24
one picture == 999 words
Windows 11 Enterprise OS

These windows hosts run very smooth.
I have used them to do advanced support on the way working from phone. I can recommend 100%

I have used only Windows Server OS (+ Android Emulator) since I got @ProHosting24 VPS last year.

themantra

How is the server uptime quality? Does it experience frequent downtime?

lowendlurker

Any docs or config about network settings for Proxmox?
IPv4 NAT and IPv6 passthrough or something like that?
I tried couple things but couldn't get network in Proxmox VMs

ProHosting24

@themantra said:
How is the server uptime quality? Does it experience frequent downtime?

No, we utilize only enterprise hardware and monitor our environment 24/7.
The hypervisor makes use of a storage area network. Therefore VM Data is present on all nodes anytime with inluded 3x replication of vm data, which makes our service high available.

https://lowendtalk.com/discussion/comment/4010506/#Comment_4010506
But you could also go through that older discussions to see some reviews.

tra10000

@ProHosting24 said:

@themantra said:
How is the server uptime quality? Does it experience frequent downtime?

The hypervisor makes use of a storage area network. Therefore VM Data is present on all nodes anytime with inluded 3x replication of vm data, which makes our service high available.

So in case of storage failure all nodes will be inaccessible