I'm currently in the process of reviewing some VPS providers to see if what I'm paying at slicehost is value for money or if I could get a better deal somewhere else. On the surface the pricing at slicehost is allot higher than some of the VPS providers around now. A year or two back I was more that happy to pay the prices slicehost is asking for, but now there are other options becoming available.

Moving VPS providers is a big deal and will take me a long time to migrate, but if the cost saving is significant and the servers, network infrastructure and customer support is up to the high standard that slicehost provide, it may be worth it. The factors I'm going to base my decision on are: server speed and redundancy (CPU, RAM, I/O), network speed and redundancy, management interfaces, customer support and of course cost.

The first one I'm reviewing is Burst.net. The prices they're offering seem too good to be true. i.e. USD $10/month vs $70/month at slicehost .. I'm wondering what is the catch. See pricing below.

http://www.slicehost.com/
https://www.burst.net/linvps.shtml

To test these two services I setup two VPS's details below:

=== Burst.net ===
Ram: 512mb
CPU: 1000mhz allocated
OS: ubuntu-9.10-x86 - 32 bit
HD: 20gb
Bandwidth per month: 1000gb

Note: I tried upgrading to 1Gb That was still pending after more than 1 day, so I ran the tests anyway.

=== Slicehost.com ===
Ram: 1024mb
CPU: 1 core dedicated (burstable to use all cores)
OS: ubuntu-9.10-x86 - 32 bit
HD: 40gb
Bandwidth per month: 600gb

Now for the tests ... I used a system called phoronix test suite for most of these performance tests and the tests were run on bare bones VM's without any additional daemons running.

http://www.phoronix-test-suite.com/

=== Server Performance ===

=== slicehost ===
FS-Mark (Disk I/O): Average: 14.70 Files/s
AIO-Stress (Disk I/O): Average: 10.82 MB/s
Apache Benchmark: Average: 3144.24 Requests Per Second
Timed Apache Compilation: Average: 152.03 Seconds
RAMspeed: Average: 1964.47 MB/s

(4x)
vendor_id : AuthenticAMD
cpu family : 15
model : 65
model name : Dual-Core AMD Opteron(tm) Processor 2212
stepping : 2
cpu MHz : 2010.300
cache size : 1024 KB
physical id : 0
siblings : 1
core id : 0
cpu cores : 1
fpu : yes
fpu_exception : yes
cpuid level : 1
wp : yes
flags : fpu tsc msr pae mce cx8 apic mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt lm 3dnowext 3dnow rep_good pni cx16 lahf_lm cmp_legacy svm extapic cr8_legacy
bogomips : 4064.43
TLB size : 1024 4K pages
clflush size : 64
cache_alignment : 64
address sizes : 40 bits physical, 48 bits virtual
power management: ts fid vid ttp tm stc

=== burst.net 512 ===
FS-Mark (Disk I/O): Average: 98.70 Files/s
AIO-Stress (Disk I/O): Average: 245.10 MB/s
Apache Benchmark: Average: 5024.04 Requests Per Second
Timed Apache Compilation: Average: 114.29 Seconds
RAMspeed: Average: 3208.64 MB/s

(1x)
vendor_id : GenuineIntel
cpu family : 6
model : 15
model name : Intel(R) Core(TM)2 Quad CPU Q6600 @ 2.40GHz
stepping : 11
cpu MHz : 2394.006
cache size : 4096 KB
physical id : 0
siblings : 4
core id : 0
cpu cores : 4
apicid : 0
fpu : yes
fpu_exception : yes
cpuid level : 10
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm syscall nx lm constant_tsc pni monitor ds_cpl vmx est tm2 cx16 xtpr lahf_lm
bogomips : 4791.29
clflush size : 64
cache_alignment : 64
address sizes : 36 bits physical, 48 bits virtual
power management:

Note the description of the apache load test: This test profile measures how many requests per second a given system can sustain when carrying out 500,000 requests with 100 requests being carried out concurrently.

In every area the actual server performance seems to be better on burst.net. And to really rub salt in the wound the burst.net tests were run on a lower spec server.

=== Network Speed ===
Both hosts are in the US and running ping tests on both show they are relatively similar in terms of network latency. Slicehost did come out on top by a small margin on the ping test. i.e. 240ms vs 290ms average

In terms of general responsiveness of the ssh console, slicehost was a long way ahead. The console on burst.net seemed to just hang there thinking for several seconds at a time on simple ls commands and the like. In general I found the network to seem a little flaky as burst.net. The verdict is still out on this, but I will continue testing it and post details here.

=== Customer Support / Management Tools ===

So far the customer support at burst.net has been quite responsive. The sign up and payment process was extremely convoluted and had me almost ready to give up right at the starting line. Their management systems are disjointed, each requiring different user account credentials. They have one system for the help desk, one for the account administration area and one for the VPS management. Also I found the management tools quite laggy. I'm not sure if this is because it's overloaded or if their network is a bit flaky.

Support, management tools, online documentation, etc, etc at slicehost are all A+, number 1, 10 out of 10 ... you can't beat it.

=== Conclusions ===

On server performance and pricing alone burst.net is the winner. On customer support and network infrastructure slicehost is ahead. If I had to decide right now I wouldn't know which way to go. It's a real dilemma. If anyone has any further info to add to this I'd really appreciate some feedback. I'll continue testing and add to this blog post as more information comes to light.