Buying vs renting
The one question you will face is whether to rent or buy one. I faced the same dilemma when I was about to setup my own dedicated server.
After careful analysis and reading a lot of material off the internet I came to the conclusion that renting has more advantages compared to buying. Here is a summary of my reasoning:
1) Cost – A decent sized server will cost $2K to buy, monthly hosting will be another $50 per month (10TB bandwidth and 1U space). Assuming a life of 3 years for the server, it works out to $105 per month. A similar sized server on rent can be obtained for $100 per month (I recommend iweb.com).
2) Upgrading server – for smaller upgrades like RAM, HDD, etc, own servers work out cheaper. For CPU upgrades renting works out cheaper and easier to do. Your hosting provider can do most of the work for you in you are renting it from them. If you own the server you will end up doing all the work.
3) Availability & Maintenance – if the server goes down, you have to pretty much buy another buy and wait for the time for it to be delivered and ready. If you rent the server your hosting provider can fix the problem quickly (most decent hosting providers will keep spares at the datacenter).
4) Cost spread – buying involves spending money upfront whereas renting is easier on your cashflow.
Keeping all this in mind I chose to rent a server than buy one.
With cloud computing coming in, things have changed as far as scaling flexibility and availability is concerned, although cost is a major concern.
I used AWS (Amazon cloud) for a project and found it pretty good for my work. It was short-term in nature and required creating 20 instances quickly. Flexibility offered is tremendous.
- My experience running PHP on IIS 7.5
- Speed testing a website
- To RAID or not to RAID
- Color Conversions
- Rotating or Resizing an Image in OpenCV
- Do all versions of OpenCV run at the same speed?
- RAID 1 vs RAID 5
- How to detect the color of a person’s shirt
- Hardware Platforms for Embedded Computer Vision, Image Processing and Deep Learning