Seeking Recommendations

[Phil Davis]

We are planning to deploy 3-4 KVM VPS on below Node
========================================
Processor : 3.4GHz Hexa-Core E-2236 Coffee Lake Xeon
Memory : 64 GB
Hard Drive : 1TB NVMe SSD
Operating Sys : AlmaLinux 8.x
Bandwidth : Transfer per month/ 20 TB
Port speed : 1GB
=========================================

Your expert opinion is invited to know whether above configuration is suitable / sufficient for numbered KVM vpses or does it need more higher end node.

 

[RackWorks]

That'll run 3-4 KVM guests without breaking a sweat, but it really comes down to what's going on inside them. Idle web and mail boxes? You could put a lot more than 4 on there. Databases or game servers? 12 threads goes quick.

Only thing I'd change is the OS. Alma 8 is supported to 2029 so there's no rush, but for a fresh hypervisor build I'd go 9 or 10.2 just for the newer kernel and QEMU. No reason to start a new node two major versions back.

 

[Phil Davis]

Thanks for your valuable input.Well not sure , what would would be there inside those atm.Those would be for clients ordering kvm vpses.With fresh build , we simply dont want to rush many on to it, so planning to have max 5 machines / vpses.

 

[fairyhostingcom]

Yes, this configuration should be absolutely fine for 3-5 and even more KVM VPS.

As RackWorks mentioned, it mainly depends on what customers will actually run inside them. For normal VPS usage I don’t see any problem with this hardware, especially if you plan to start with only 5 guests and monitor the load.

 

[Mivocloud]

Yes, absolutely. The node configuration is more than capable of running 3–4 KVM VPS with excellent performance and plenty of headroom for each instance.

 

[Zen_Viju]

That should be enough for 3–4 small or medium KVM VPSs, depending on their workloads and resource allocations.

For example, four VPSs with 2 vCPUs, 8 GB RAM and 150 GB storage each would leave capacity for the host. Shared vCPUs are fine, but avoid promising dedicated CPU resources beyond the node’s capacity.

The main concern is the single NVMe drive: its failure would affect every VPS. Consider mirrored drives for availability and separate off-server backups.

What will the VPSs run, and how much CPU, RAM and storage will each need? Those details matter more than the number of VPSs alone.