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

I have a managed semi-dedicated server running Centos and a LAMP setup with PHP 5.2.9. I've noticed over the past couple of months that HTTP requests are taking much longer and page loading speeds across a number of sites have slowed down considerably. There doesn't appear to be any significant load on the server and I've checked this with the sysadmin support, who say there isn't any obvious issue. However, there has been a definite deterioration of performance, consistently, for at least two months now - so I really need to try and figure out what's going on and am hoping someone might be able to suggest possible causes. The server is at around 80% of its disk usage capacity, so maybe that could be impacting on things.

Any help or pointers greatly appreciated!

EDIT: I have eAccelerator enabled, Apache KeepAlives enabled, mysql config as below. I should also point out that the bulk of the loading time for web pages is making the connection to the server, not downloading the page content.

max_connections = 400
key_buffer = 16M
myisam_sort_buffer_size = 32M
join_buffer_size = 1M
read_buffer_size = 1M
sort_buffer_size = 2M
table_cache = 1024
thread_cache_size = 286
interactive_timeout = 25
wait_timeout = 1000
connect_timeout = 10
max_allowed_packet = 16M
max_connect_errors = 10
query_cache_limit = 1M
query_cache_size = 16M
query_cache_type = 1
tmp_table_size = 16M
open_files_limit = 8192
max_allowed_packet = 16M
key_buffer = 32M
sort_buffer = 32M
read_buffer = 16M
write_buffer = 16M
share|improve this question

There are a few things you can try to do to narrow down the source of the performance issue (assuming it is mainly one thing anyways):

  • Create a small HTML file and test accessing it remotely. If there is a performance issue here it is either somewhere on the network link between the servers or the Apache configuration.
  • Create a small PHP file that does something basic (like output numbers 1-100) to see if the issue may be on the PHP side.
  • Create a test file that accesses the database in a trivial/simple query to check the basic database performance.
  • Run some typical queries locally on the server in the mysql client. If they run slow locally they are going be to slow remotely too. Enabling the slow query log is also a good idea if you suspect database performance issues.
  • If you suspect the increase in database size to be the issue try testing queries on several copies of the database/tables (small, medium and large). If your queries or indexing is poor then you'll see query times increase rapidly as the number of records grow.
  • Check the basic server status like memory (make sure you are not hitting the swap regularly), hard drive errors (smartctl), server load, CPU idle time, free space on all partitions (like /tmp), other processes eating up the CPU/memory (top), and the system logs for relevant errors.
  • If none of the above reveals anything you can start with a page/query that you know performs poorly and work backwards removing things until the issue is isolated.
share|improve this answer

If you're doing a lot of database lookups in your php apps you might review your indexed tables as well as running an optimization on the database(s). Also consider increasing the thread_concurrency, buffers and table_cache in mysql's configuration.

You might stand to gain some speed increases by optimizing apache and php as well. For instance, make sure you have KeepAlives enabled in Apache and consider installing and activating mod_gzip if applicable. For php you may want to look into the eaccelerator module.

share|improve this answer
Thanks for the reply @meanasspenguin - I already have eaccelerator enabled and running, while KeepAlive is enabled and mysql config I added to my original post. – franko75 Jan 20 '11 at 15:19

It's propably the disk subsystem. It's either shared with another host which slows down your system, or worse, it's some kind of big over-the-network disk subsystem. In both cases the responsetime of the disk is seriously affected.

It's a good idea to use real dedicated hardware in production environment.

If you want to test this, move documents and mysql databases to tmpfs and see what happens. It's not everything as system still needs to load stuff from the root, but it's a start.

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.