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

We have a number of servers, all in the same datacenter, all similar spec, all run the same application (LAMP ecommerce app), all have similar traffic levels and all have similar average load. However, one of these servers regularly scores much worse than all the other servers in response time reports from Pingdom. All our servers bar this one have response time averages between 190ms - 210ms however one server avaerages between 280ms - 300ms. One of the only differences between these servers is the version of CentOS. On the slower site the version is 4.8 but on the faster servers they are all running 5.5. Is it likely that the OS version is enough to slow down response time on the server or is it likely to be something else?


share|improve this question
up vote 1 down vote accepted

Try running free -m to see if it's using swap space from RAM. Very common on similar builds but different OS.


On the last line look for anything above zero in the middle column Swap entries:

total used free shared buffers cached

Mem: 3145 2882 262 0 919 1131

-/+ buffers/cache: 831 2313

Swap: 254 0 254

My machine is saying the "262"MB of RAM is free in the Mem line and no swap file has been used with the "0" in the last line which would slow it right down.

share|improve this answer
what am i looking for specifically? – seengee Mar 28 '11 at 20:03
although this wasnt actually the issue this caused me to realise that one of the RAM sticks in the server had actually died so the points are yours :) – seengee Apr 5 '11 at 12:48
Glad to help :-) – Jonathan Ross Apr 5 '11 at 12:49

When you have eliminated the impossible, whatever remains, however improbable, must be the truth. -- Sherlock Holmes

If your servers all the same hardware, then the reason must lie within software. Older version of CentOS means older kernel and older LAMP stack (unless you compile your own). CentOS 4.8 runs kernel 2.6.9. CentOS 5.5 is 2.6.18. 2nd factor would be tuning. If you fine-tuned 5.5s, but run an out-of-the-box 4.8, this will factor in as well.

Developers strive to improve speed both of kernel and applications, so yes, OS version may play important role in the speed delta. 30% improvement on identical hardware? Not impossible.

OTOH if your servers don't run on identical or very similar hardware, or if the 4.8 one is the one connected to the rest of the net by that ancient over-heating half-duplex 10 Mb hub locked in the broom shed, then hardware may be the dominating factor explaining the speed difference. CentOS 5.5 on quad Nehalem with 32 GB RAM should be significantly faster than CentOS 4.8 on a P IV.

share|improve this answer
I believe all have a similar up to date Stack - we use cPanel/WHM and they have all been updated recently to allow for PHP 5.3 – seengee Mar 28 '11 at 20:26

Perhaps the slower server is sharing his connectivity or other resources, e.g. storage, with high load servers?

share|improve this answer

Your Answer


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

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