Howdy, Stranger!

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


HostCram KVM-8G benchmarks and first impressions :)
New on LowEndTalk? Please Register and read our Community Rules.

HostCram KVM-8G benchmarks and first impressions :)

vitobottavitobotta Member
edited November 24 in General

Hola, I ordered three virtual servers from HostCram yesterday (Dallas location) and I thought I'd share a few benchmarks. These are the KVM-8G model with 4 i9-11900K cores, 8 GB of RAM and 160 GB of NVME storage. I was especially keen to see GB5 scores.

First benchmark: curl -sL yabs.sh | bash -s -- -r

Server 1:
https://gist.githubusercontent.com/vitobotta/f066946833ebb4c34046770590591c87/raw/e5ade6666a9c0abdbd86e9b4674bfd2fa2fbfd47/gistfile1.txt

Server 2:
https://gist.githubusercontent.com/vitobotta/261ce4e9d95f765f760777294ddf588f/raw/165d258c532328fc744997b8bacbff9fcb1aac86/gistfile1.txt

Server 3:
https://gist.githubusercontent.com/vitobotta/967e806a3df07e9166e59d11adea6ed0/raw/485d5d1330db71f07541e3664f6e2929bf0954c5/gistfile1.txt

Second benchmark: curl -s wget.racing/nench.sh | bash; curl -s wget.racing/nench.sh | bash

Server 1:
https://gist.githubusercontent.com/vitobotta/be4cc376c2208f37001ed80e0fbd4216/raw/7537acc31c1ba3b55438660d7c6cd33f5e4f7e42/gistfile1.txt

Server 2:
https://gist.githubusercontent.com/vitobotta/0ab7a299731af798acaedcb9bab11839/raw/a49d424db84f990c9a92d683acc6d0cbe5950484/gistfile1.txt

Server 3:
https://gist.githubusercontent.com/vitobotta/dd90f094fabc20fc25decb02bdfbdace/raw/78c0836d8cc6190c7fc0b4d5ca28bf7282807e20/gistfile1.txt

As you can see the servers seem very zippy. I also tested the connection between the servers with iperf3 and the speed was around 870Mb/sec.

Overall the first impressions are pretty good. I wish there was a proper control panel though.

Later today I am going to set up a Kubernetes cluster with these nodes (actually I will order another, smaller server for the master so these three will act only as workers).

I am hoping it works out... the machines seem awesome but because of the lack of block storage I need to use Longhorn in Kubernetes to replicate storage volumes between the nodes for HA, using the main disks. So the network speed will be crucial in determining if this will work well or not.

The second thing is that the servers are in Dallas, so I am hoping that latency for European users is good enough.

But these are not issues with the machines of course. The machines look pretty fast so I am looking forward to setting things up later today. I can update this thread later if I have something more to share and if anyone is interested. :)

Comments

Sign In or Register to comment.