FIX: phpBB forum cause high bot/spider server load and traffic

[s]

If all bots are not properly added to Spiders/Robots in phpBB Administration Control Panel (ACP) GENERAL/QUICK ACCESS settings, they will cause excessive load and traffic on the web server. This happens because bots will get every single URL with a different SID and therefore consider them as different pages. They keep crawling those pages and getting all the time new pages. When a bot is properly detected, phpBB will return it pages without SIDs as well as prevent access to pages like cron where there is absolutely nothing useful to find.

As a fix, you can a generic bot to your forum that will catch all current and new bots which are not already specified in Spiders/Robots. Simply add there a new bot using the following values:

Code: Select all

Bot name: Unspecified bot
Agent match: bot

This will make sure that all bots that have the string "bot" present in http User-Agent header. I proposed adding the configuration to phpBB by default, but they seem to have deleted my account and idea so it is not likely to be part of phpBB any time soon: https://www.phpbb.com/community/viewtop ... #p16011421

A quick experiment showed that catching all bots can reduce the load 90% and traffic 50%. This can be a viable alternative for blocking all bots from visiting the forum. The problem remains how to find and mark all current an future bots, because if any of them is not detected, they will start indexing SIDs.

Happy hacking,

Santeri

[s]

You need to add also spiders and crawlers to make sure all of them are covered. So add generic bots with these agent matches:

• bot - Unspecified bot
• crawler - Unspecified crawler
• spider - Unspecified spider

This will do the trick and you will be able to catch all unspecified spiders and crawlers, too. This should cover all the legitimate automated traffic.