Are Spiders Eating Your Servers? The Impact Of Their Unexpected Load and How To Counter It

Jul 21, 2017
2:30 PM - 3:30 PM
Holeman Room

For years I’ve watched people try to tame “server problems” with a focus on their code, their SQL, the jvm, and so on. Yet often it turns out that the root cause is actually unexpected load. And that load may be from things you never expected (automated), at volumes you never expected. I’ve found folks with as much as 80% of their web traffic to be such unexpected automated traffic! Worse, there are characteristics of such automated visits that may actually have MORE IMPACT than “real users”: for instance, did you know they create a new session–and run session startup code–for each page they visit?!

The good news is there are solutions to better manage (or simply block) such automated requests which may already exist in your environment, and tools you may consider (some free, some commercial) which can be easily implemented. There are even SAAS solutions that could help alleviate such problems with just a single tiny change in your environment! You may also want to consider some admin configuration options related to sessions and/or client variables, as well as reconsider some coding choices in your session startup code.

In this session, veteran server troubleshooter Charlie Arehart will guide a more detailed review of the issues above, including how to identify such traffic, more on these specific impacts, and most important identifying the solutions along with their pros and cons. He has helped shops achieve dramatic reductions in impact from such automated requests, resulting in greater server stability and performance.