We have new Nagios monitoring in place to track issues with yankeemagazine.com, and this new monitoring happened to coincide with the release of the latest Great Yankee Giveaway yesterday, with an email to 17,000 readers around 2:15 p.m.
Almost immediately we began to see issues with Apache saturation – almost every Apache process on two running servers was consumed with HTTP POSTs to /giveaway, and these processes appeared to hang on for quite a long time, meaning that access to Yankeemagazine.com quickly because very sluggish as there were no available processes. This wasn't a server load issue, so no additional servers were spawned.
Can you give me some background on the system you put in place to process giveaway entries? Are there any known issues that might describe the behaviour that we saw yesterday? We are running a simple gravity form that is embeded in a post that is displaying in a feed at this URL: http://www.yankeemagazine.com/giveaway