Server Fault is a question and answer site for system and network administrators. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Processor Info: enter image description here

Memory: enter image description here

OS: Cent OS

I have a third party API I will be using, that will be feeding millions of possible queries every 30 seconds - 1 minute to my VPS. The datasets to be sent will be small, as in, the numbers '88' and '95' for example. Or '100', '75', and 'by John Jones'

Can my VPS handle this?

share|improve this question

closed as not constructive by EEAA, Zoredache, Sven, Zypher Feb 9 '12 at 22:39

As it currently stands, this question is not a good fit for our Q&A format. We expect answers to be supported by facts, references, or expertise, but this question will likely solicit debate, arguments, polling, or extended discussion. If you feel that this question can be improved and possibly reopened, visit the help center for guidance.If this question can be reworded to fit the rules in the help center, please edit the question.

Millions of API requests every 30 seconds? At one million, that's ~33.3k API requests per second. I don't care how well you optimize your code, there's no possible way you can do this with a single VPS. As an aside, scaling questions are for all intents and purposes unanswerable without vast amounts of data and testing, which only you can perform. – EEAA Feb 9 '12 at 21:55
It sounds quite unlikely. but you haven't told us anything about what is done with the data? is it thrown away? do you add 1 to the number and return it? write it to disk? Do millions of calculations before returning a result? Depending these answers there would be wildly different requirements. – stew Feb 9 '12 at 22:06
millions of API requests every minute on a cpanel box... good luck – anthonysomerset Feb 9 '12 at 22:13
Once the data is fed to the system, the system just outputs the value... doesnt do any calculation or anything. – John Deer Feb 9 '12 at 22:14
@JohnDeer - for this load, you're going to need multiple servers somewhere behind a load balancer, and perhaps multiple load balancers. Where those servers live is up to you. EC2, while very nice to use, is not magic. You still need to have expertise on developing and supporting scalable, distributed applications. – EEAA Feb 9 '12 at 22:23
up vote 1 down vote accepted

To Answer your question NO is probably the answer you will find, not at anywhere near the number your thinking of.

if the memory screen shot is correct you only have 100MB of ram (i suspect its cut off) EDIT it shows 1GB ram

The screenshot looks awfully like its from the cpanel interface which is more than definitely not tailored and optimized for your possible use case so performance would be far less than optimal

as the comments suggest, how many requests you can receive and process depends totally on what your doing with them. but with hardware that spec i doubt its even possible for the box to even receive that number of requests in the time frame you want

companies that do API processing on that level tend to have multiple (large) servers dedicated to the task of processing that data and be load balanced between them for HA also

nothing beats doing proper research and testing, why not test a smaller amount of queries and monitor performance, test and different levels and graph average performance metrics and you should be able to work out what you actually need or what the box is actually capable of

share|improve this answer
Thanks for this. Much needed anthony. – John Deer Feb 10 '12 at 2:00

Not the answer you're looking for? Browse other questions tagged or ask your own question.