There's an Apache 2 server choking under multiple demands. Quantity of processes reaches number set by Maxclient, and new connections stack up in queue. I have to identify which demands take a lot of time x quantity of demands. Apache logs may show either:

%...D Time come to serve the request, in microseconds.

or

%...T Time come to serve the request, within minutes.

And this time around come to serve the demands make it's difficult to evaluate. Because at that time when process table overflows it's time around which I have to measure. There serve time equals time put in connection queue plus time processed.

So b .perl script which required 2 seconds can have 2 seconds. And .cgi script which spent minute (real data )) in queue can have 60.5 seconds.

It is possible to method to log time come to process demands, ideally not including engaging with hooks?

You should use mod_status.
This module permit you to monitor do you know the demands presently being offered,
as well as their process status,
for example

  • "_" Awaiting Connection,
  • "S" Setting up,
  • "R" Reading through Request,
  • "W" Delivering Reply
  • "K" Keepalive (read)
  • "D" DNS Research,
  • "C" Closing connection,
  • "L" Logging
  • "G" Beautifully finishing,
  • "I" Idle cleanup of worker
  • "." Open slot without any current process