Httpd – Apache httpd response time increase with concurrent connections

apache-2.2concurrencyhttpd

I know many questions have been asked on this subject, but I didn't find an answer to help me.

I'm trying to have Apache httpd answer 200 concurrent connections. I'm testing with jMeter, fetching a small static text file (about 100 bytes), using prefork; MaxClient & ServerLimit at 500, neither memory or CPU are limiting (both are under 50%), bandwidth is 1Gbps.

if I start only one connection, the response time is about 7ms
if I start a few concurrent connections (let's say 2 or 3), the response time goes to 70ms
if I start a little more (let's say 20), the reponse time is over 2s even for the first connection

Using keep-alive doesn't change much. Restarting httpd doesn't help neither.

Am I doing something wrong? how can I fix that?

Thanks in advance

greg

Best Answer

To solve the problem I had to permanently increase the number of running processes. Instead of:

StartServers          5
MinSpareServers       5
MaxSpareServers      10

I changed to:

StartServers        300
MinSpareServers     300
MaxSpareServers     300

Now the response times are acceptable. It seems creating/killing processus is very heavy even for fast computers...

Related Solutions

Httpd – Apache httpd Problem

There is not enough information to give any informed guess but here are a few things you can look into or try:

Try running ApacheBench on the server to see if the slow page loading issue can be duplicated. For example: ab -kc 10 -t 30 http://yourserver.com/somepage.php If the pages load fast on the server then the issue is network related.
Try loading different types of pages (images, JS, CSS, PHP, HTML, SHTML, etc...) and see which ones load fast and which don't. If static files all load fast then the issue is something related to the dynamic page setup/execution. If all files load slow it is a more likely a general server/configuration issue.
Do you have other Apache configuration files (check in '/etc/httpd/conf.d/' for example). Your virtual host entries reference proxy options but no ProxyPass or similar line is found. If you are proxying requests consider that the proxy source may be slow. Test it with ApacheBench to confirm.
Those entries from your error log are not actual errors but just debug log statements. Errors will have a [error] header string. You can also check other error/system logs for relevant messages (/var/log/messages for example).

Linux – NGINX timeout after +200 concurrent connections

You will need to dump your network connections during the test. While the server may have near zero load, your TCP/IP stack could be billing up. Look for TIME_WAIT connections in a netstat output.

If this is the case, then you will want to check into tuning tcp/ip kernel paramters relating to TCP Wait states, TCP recyling, and similar metrics.

Also, you have not described what is being tested.

I always test:

static content (image or text file)
simple php page (phpinfo for example)
application page

This may not apply in your case but is something I do when performance testing. Testing different types of files can help you pinpoint the bottlneck.

Even with static content, testing different size of files is important as well to get timeouts and other metrics dialed in.

We have some static content Nginx boxes handling 3000+ active connections. So it Nginx can certainly do it.

Update: Your netstat shows a lot of open connections. May want to try tuning your TCP/IP stack. Also, what file are you requesting? Nginx should quickly close the port.

Here is a suggestion for sysctl.conf:

net.ipv4.ip_local_port_range = 1024 65000
net.ipv4.tcp_rmem = 4096 87380 8388608
net.ipv4.tcp_fin_timeout = 30
net.ipv4.tcp_keepalive_time = 30
net.ipv4.tcp_tw_recycle = 1
net.ipv4.tcp_tw_reuse = 1

These values are very low but I have had success with them on high concurrency Nginx boxes.

Best Answer

Related Solutions

Httpd – Apache httpd Problem

Linux – NGINX timeout after +200 concurrent connections

Related Topic