I have been reading many articles to understand how to configure apache server for a high traffic website. the article here tries to calculate Thrashing point-where swapping occurs-and then calculates an initial estimate for the number of ServerLimit parameter for mpm_event based on that:
Thrashing Point=(buff/cache - reserved)/(Avg. Apache)
in the Buff/Cache Memory section of the article, the writer says:
Don’t be fooled by the column labeled “available.” We are solely looking at the memory we can reappropriate, which is the buff/cache column
this is what confuses me. in the article example, there is about 700 MB of buffer/cache and about 700 MB of available memory, but only the 700 MB of buffere/cache is taken into account for calculating the maximum number of apache children (ServerLimit). my question is that why the remained 700 MB of available memory is not considered while it is large and unused, and if taken into account can increase ServerLimit significantly?!
