Server Fault is a question and answer site for system and network administrators. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

After searching serverfault for similar questions without success, these are my numbers for one magento instance, running on multiple servers:

  • After varnish about 4 requests per second hit the webservers
  • The magento cache is configured to use one separate memcache server where I'm measuring about 210 Mbit/s bandwith usage.

Compared to other projects, magento and non-magento, this number seems way off (as in extremely high).

I'd like to get some data to compare to, or even better, if you have any idea what exactly causes this/how to find it and how to improve the situation.

share|improve this question
up vote 1 down vote accepted

I just found the answer: TinyBrick_Lightspeed. It is a full page cache which didn't make any sense as varnish would cache anything this module could have cached. Nevertheless, it writes full html output to the cache for each url hit. Sounds still okay for its original purpose.

Not okay is their "creative" way of tagging support for a cache backend that doesn't support tags per se (memcached).

If you write a key X to the cache and want to attach tag Y to it, it loads the content of an additional cache key TAG_Y, adds the new cache key to it, and saves it back to TAG_Y.

The more products/pages/categories/... you have, the bigger those tag entries get, the more data has to be received and written to the cache for every new request.

Some Tags as "MAGE" grew to 10MB+ ...

Before and after the fix:

Before and after the fix

share|improve this answer
HMPF. Whow. Epic design failure in that system. Nice you found it. – TomTom Feb 11 '14 at 8:17

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.