Drupal scaling and performance tuning - Part 1

As WSO2Con 2011 was a huge hit in IT sector, I had to face a problem with wso2.org (Oxygentank developer Portal). Which site couldn’t handle a large traffic constantly.

The old system we had was NGINX (Load Balancer) fronted 4 Drupal nodes which was running with Apache2 with master - master MySQL replication.

But during a high-traffic Our servers got more than 1 minute (average) time to respond to a request. As NGINX’s fail timeout for 45 seconds, users got 504 Gateway timeout most of the time. Same time Apache server load went above 30 - 70. First as a quick solution we spinned new 2 servers and routed the load across them. That helped us to cater for few hours but again there was a huge delay when site was loading. Then we found that MySQL server load also gone up and process-list has grown and MySQL Server take time process queries than it was before. Root cause for this new problem was 6 drupal nodes had started stress MySQL servers continually.

Then I started to ask some questioned by my self.

OK, Then why we didn’t see this new problem before we plug new nodes? Answer was simple, Drupal nodes couldn’t stress the DB as it got self killed (Frozen) during high traffics.

Before jump in the Apache, Our WSO2 Infrastructure team revisit our monitoring systems (Ganglia) and found that servers were running on swap most of the time

As a solution I stated to tune up the Apache to avoid memory swapping problem. Then I saw that there was some miss configurations which take more memories when Apache process start to handle the traffic. Due to that configuration server can easily excused.

MaxClients as was one of a major key parameter I had to change during the Apache memory optimization process.

####Apache Memory Optimization

Once I restart our servers with above setting, server could handle 150% traffic from Apache end and server load was always below 2 (in a burst server-load was 1-2)

This solved my Apache hanging problem. Then I ran a another load test to verify my setting. I was amazed with the improvement but that haven’t solved my problem.

server is taking bit long to response but I didn’t see a connection drop.

Now, What is wrong with my settings, Why server response is slow ? MySQL!!

I will discuss in part 2, how I addressed this problem .

comments powered by Disqus