Apache memory release~

Question: Why does one Apache.exe in the process occupy hundreds of megabytes of memory and does not decrease after the server has been running continuously for many days or after the access peak?
Answer: Open apache2confhttpd.conf with Notepad, search for MaxRequestsPerChild, and change MaxRequestsPerChild 0 to MaxRequestsPerChild 50.

Usually you can see two apache.exe processes in "Windows Task Manager - Processes", one is the parent process and the other is the child process. After the parent process receives the access request, it hands the request to the child process for processing. MaxRequestsPerChild This directive sets the number of requests that an independent child process will be able to handle. After processing "MaxRequestsPerChild number" requests, the child process will be terminated by the parent process. At this time, the memory occupied by the child process will be released. If there is another access request, the parent process will re-spawn the child process for processing.
　If MaxRequestsPerChild is set to 0 (infinite) by default or a larger number (for example, more than 10,000), each sub-process can handle more requests, and access efficiency will not be reduced due to continuous termination and starting of sub-processes. However, MaxRequestsPerChild is set to 0. At this time, if 200~300M of memory is occupied, the memory occupied will not be reduced even when the load is reduced. Servers with larger memory can be set to 0 or a larger number. Servers with smaller memory may wish to set it to 30, 50, 100 to prevent memory overflow

================================ ======================================
apache memory usage

apache process is using memory When, it is "gradually growing". In other words, until the process dies, the amount of memory used will continue to increase without decreasing. In this case, the amount of memory used by the apache process depends on the maximum amount of memory used by your application.

keepalive parameter

KeepAliveTimeout This parameter determines how long an http process can wait before doing nothing? Imagine that if keepalive is set to on and keepalivetimeout is set to a relatively large number, the memory occupied by apache will increase very quickly. This is because, after an apache process completes a task (and reaches a certain memory usage, think of "progressive" mode), it will not exit immediately, but will wait for a keepalivetimeout. Assuming that user connection requests continue to arrive, there will be quite a lot of useless apache processes accumulated, and these processes will not be killed until timeout.

However, keepalive is indeed very effective for the transmission of static files, such as image files. Therefore, keepalive should be set to on, but keepalvietimeout should be set smaller, such as 5s

MaxRequestsPerChild

This parameter means that apache After the process has processed how many requests, it must exit and start again to avoid memory problems during processing.

For php scripts, it is beneficial to set this parameter smaller to avoid the pressure on apache caused by the continuous growth of memory used by the program: let this parameter release memory regularly, because php is after the script is executed. , automatically releasing only used resources (memory).

For example, set it to 50? If it is too small, regenerating an apache process will also consume resources. This is a balance issue.

The above is the content of Apache memory release~. For more related content, please pay attention to the PHP Chinese website (www.php.cn)!