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.
Dedicated server and VDS comparison. Help needed.
I was trying to compare two instances that I have. Both are very similar in benchmarks and don't know which one to keep.
To give some context, I am running a small image optimization API that averages at 50000-60000 optimizations per day and 30000-40000 requests (one request can have multiple images). Currently, the API is hosted on the Dedicated Box and is up for at least 2 years without any problem. The same box is running a website, database, and API.
As the service grows I want to run a separate server for the database and the control panel (small Hetzner instance CPX11) and then a failover server that will be used if the main server goes down. This will allow me to scale up and based on the growth add more processing servers.
1.) Dedicated Box (Hetzner) - 24 EUR
- Intel i7 4770
- 32GB (4x8GB) DDR3-1600 RAM
- 2x2TB HDD in software raid
# ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## #
# Yet-Another-Bench-Script #
# v2020-12-29 #
# https://github.com/masonr/yet-another-bench-script #
# ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## #
Thu 06 May 2021 11:40:01 PM CEST
Basic System Information:
---------------------------------
Processor : Intel(R) Core(TM) i7-4770 CPU @ 3.40GHz
CPU cores : 8 @ 1706.044 MHz
AES-NI : ✔ Enabled
VM-x/AMD-V : ✔ Enabled
RAM : 31.2 GiB
Swap : 16.0 GiB
Disk : 1.8 TiB
fio Disk Speed Tests (Mixed R/W 50/50):
---------------------------------
Block Size | 4k (IOPS) | 64k (IOPS)
------ | --- ---- | ---- ----
Read | 643.00 KB/s (160) | 8.92 MB/s (139)
Write | 674.00 KB/s (168) | 9.43 MB/s (147)
Total | 1.31 MB/s (328) | 18.35 MB/s (286)
| |
Block Size | 512k (IOPS) | 1m (IOPS)
------ | --- ---- | ---- ----
Read | 34.06 MB/s (66) | 42.69 MB/s (41)
Write | 36.29 MB/s (70) | 45.93 MB/s (44)
Total | 70.36 MB/s (136) | 88.63 MB/s (85)
iperf3 Network Speed Tests (IPv4):
---------------------------------
Provider | Location (Link) | Send Speed | Recv Speed
| | |
Clouvider | London, UK (10G) | 923 Mbits/sec | 923 Mbits/sec
Online.net | Paris, FR (10G) | 915 Mbits/sec | 937 Mbits/sec
WorldStream | The Netherlands (10G) | 916 Mbits/sec | 938 Mbits/sec
Biznet | Jakarta, Indonesia (1G) | 751 Mbits/sec | 275 Mbits/sec
Clouvider | NYC, NY, US (10G) | 870 Mbits/sec | 562 Mbits/sec
Velocity Online | Tallahassee, FL, US (10G) | 840 Mbits/sec | 453 Mbits/sec
Clouvider | Los Angeles, CA, US (10G) | 819 Mbits/sec | 358 Mbits/sec
Iveloz Telecom | Sao Paulo, BR (2G) | busy | busy
iperf3 Network Speed Tests (IPv6):
---------------------------------
Provider | Location (Link) | Send Speed | Recv Speed
| | |
Clouvider | London, UK (10G) | 914 Mbits/sec | 920 Mbits/sec
Online.net | Paris, FR (10G) | 910 Mbits/sec | 924 Mbits/sec
WorldStream | The Netherlands (10G) | 910 Mbits/sec | 924 Mbits/sec
Clouvider | NYC, NY, US (10G) | 866 Mbits/sec | 439 Mbits/sec
Clouvider | Los Angeles, CA, US (10G) | 811 Mbits/sec | 311 Mbits/sec
Geekbench 5 Benchmark Test:
---------------------------------
Test | Value
|
Single Core | 1047
Multi Core | 3776
Full Test | https://browser.geekbench.com/v5/cpu/7762259
2.) Rootserver (Netcup) 16 EUR
- AMD EPYC 7702
- 16GB DDR4 ECC
- 320GB SSD
# ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## #
# Yet-Another-Bench-Script #
# v2020-12-29 #
# https://github.com/masonr/yet-another-bench-script #
# ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## #
Thu 06 May 2021 11:54:19 PM CEST
Basic System Information:
---------------------------------
Processor : AMD EPYC 7702P 64-Core Processor
CPU cores : 4 @ 1996.248 MHz
AES-NI : ✔ Enabled
VM-x/AMD-V : ❌ Disabled
RAM : 15.6 GiB
Swap : 0.0 KiB
Disk : 314.9 GiB
fio Disk Speed Tests (Mixed R/W 50/50):
---------------------------------
Block Size | 4k (IOPS) | 64k (IOPS)
------ | --- ---- | ---- ----
Read | 174.21 MB/s (43.5k) | 1.68 GB/s (26.3k)
Write | 174.67 MB/s (43.6k) | 1.69 GB/s (26.5k)
Total | 348.88 MB/s (87.2k) | 3.38 GB/s (52.9k)
| |
Block Size | 512k (IOPS) | 1m (IOPS)
------ | --- ---- | ---- ----
Read | 1.61 GB/s (3.1k) | 1.59 GB/s (1.5k)
Write | 1.69 GB/s (3.3k) | 1.69 GB/s (1.6k)
Total | 3.31 GB/s (6.4k) | 3.28 GB/s (3.2k)
iperf3 Network Speed Tests (IPv4):
---------------------------------
Provider | Location (Link) | Send Speed | Recv Speed
| | |
Clouvider | London, UK (10G) | 1.93 Gbits/sec | 2.37 Gbits/sec
Online.net | Paris, FR (10G) | 2.02 Gbits/sec | 2.37 Gbits/sec
WorldStream | The Netherlands (10G) | busy | busy
Biznet | Jakarta, Indonesia (1G) | 512 Mbits/sec | 209 Mbits/sec
Clouvider | NYC, NY, US (10G) | 817 Mbits/sec | 1.70 Gbits/sec
Velocity Online | Tallahassee, FL, US (10G) | 597 Mbits/sec | 1.23 Gbits/sec
Clouvider | Los Angeles, CA, US (10G) | 483 Mbits/sec | 1.06 Gbits/sec
Iveloz Telecom | Sao Paulo, BR (2G) | busy | busy
iperf3 Network Speed Tests (IPv6):
---------------------------------
Provider | Location (Link) | Send Speed | Recv Speed
| | |
Clouvider | London, UK (10G) | 2.03 Gbits/sec | 2.34 Gbits/sec
Online.net | Paris, FR (10G) | 1.37 Gbits/sec | 2.21 Gbits/sec
WorldStream | The Netherlands (10G) | 1.65 Gbits/sec | busy
Clouvider | NYC, NY, US (10G) | 792 Mbits/sec | 1.71 Gbits/sec
Clouvider | Los Angeles, CA, US (10G) | 541 Mbits/sec | 943 Mbits/sec
Geekbench 5 Benchmark Test:
---------------------------------
Test | Value
|
Single Core | 1048
Multi Core | 3967
Full Test | https://browser.geekbench.com/v5/cpu/7762387
In my opinion, the most important part of the server specs for this service is surely CPU, and maybe networking. The least important thing is RAM as the processing itself doesn't consume much. Which server you will keep and why?
Comments
Well, is your app ram intensive? If not, I’d go with the cheaper one
Instead of asking us I would suggest to try analyze it yourself.
I can recommend "Netdata" for this purpose - you can check cpu, i/o, networking, ram usage. Try to collect data for at least 2 days and then view historical information about usage - you will get very good indication what is important for your app and then you will make good choice
YABS is definitely not good indication for what is better now, its just some peak values done in like 2 minutes. This is helpful only after you already know what you need exactly.
If you want my personal opinion - I would pick Netcup because it has A LOT faster drive and networking.
well a vds is never fully dedicated i'd always go for the actual dedi
Netcup will cripple your server after one hour of intensive usage. Despite dedicated CPU, HDD and Network are shared. If your usecase involves a lot of downloading, then go with hetzner. If it involves hosting or plex, go with netcup.