Server Fault is a question and answer site for system and network administrators. It's 100% free, no registration required.

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

Would it be right to compute it as the network time (2ms) + disk seek and block read (10ms) ~= 12ms?
Would that be the correct order of magnitude?

  • interested in the response time, not absolute rate.
  • Assume mid-range hardware with no caching on (access pattern is too random for caching to help).
share|improve this question
You may want to use a larger file size, 100KB will hardly provide anything worth measuring given 1Gb LAN being a standard these days. – SpacemanSpiff Feb 28 '13 at 23:35
if you have many files, avoid nfs protocol. – PersianGulf Mar 1 '13 at 0:45
@SpacemanSpiff I'm interested in the minimal response time for the operation, be it 1 bit of data, or 1MB. – Gili Nachum Mar 1 '13 at 7:16
@MohsenPahlevanzadeh can you add why? If I have 1 million files stored, made accessible via NFS, and I'm accessing a 10 random ones every second, wouldn't that work fine? – Gili Nachum Mar 1 '13 at 7:17
I think assuming that each side is not overloaded, and assuming the network is not congested, that is a reasonable assumption. It must be tested. – SpacemanSpiff Mar 1 '13 at 13:49

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Browse other questions tagged or ask your own question.