1

I'm not fully sure it is appropriate to ask here - apologies if it is not.

I want to set up a webservice (REST API) using PHP on top of Ubuntu. I looked over what the hosting company I'm already using is currently offering, and have identified two options. As I'm not knowledgeable about hardware or system administration, I can't tell which one is a better fit for me.

I'm mainly wondering about the CPU difference.

The former has 2 vCores, whatever those are... I have a virtual server from 3 years back at this company, and the OS tells me its CPU is running at 3300 MHz. The later uses the ODROID-XU4 board, which has a CPU with 4 cores running 1.9 GHz and 4 cores running at 1.3 GHz (or so Wikipedia tells me).

If I run Ubuntu + Apache + PHP, then am I better of with the faster vCores, or the more plenty cores of the ODROID? And are these other differences that are likely relevant?

1 Answers1

2

Go with the vserver. Because:

  • You do not have to deal with low level hardware monitoring, like SMART for when will the disk fail nor with replacing hardware before it fails. Though I have no idea how well Hetzner does that for vservers.
  • Its cheaper.
  • Likely faster disk. SSD vs. SD card
  • Both have the same amount of RAM.
  • You don't list what resources one request will need (cpu, mem, disk i/o) so it is hard to tell if comparing those CPUs for your application makes even sense.

According to http://wiki.hetzner.de/index.php/VServer/en#How_high_is_the_performance_of_the_vCPUs.3F one vCPU of the vserver is comparable to an Intel Xeon E5 CPU running at 2.0 GHz. For request that don't need disk i/o and 2GB would fit 4 or more request in parallel the AX10 with the Exynos-5422 ARM CPU might be better. But it is only worth considering when you would need more than 1 server.

So if you actually want to know which is better first find out the resources required on any one system. How much RAM do you need as a base? How much RAM, CPU, disk write and read per request in the worst case? How many request fit into 2GB? Does your MySQL DB fit into memory?