Problem Averted-
Michael <[email protected]>
| Newsgroups | gmane.comp.web.fastcgi.devel |
|---|---|
| Message-ID | <[email protected]> |
Boy do I feel stupid! So my problem was just that FastCGI was doing its job. I set the growth and decay values to be low in hopes of getting a quickly dynamic pool of processes. What I wasn't realizing, is that what I was doing really didn't need more than one process. Despite a concurrency of 20-40 requests at a time from ApacheBench, PHP was processing the requests just as fast as they came in, not creating a large enough input queue to spawn more processes. Adding a sleep(5) [simulating a longer running script] into my perl and php scripts caused me to get 20-45 interpreters running. So it is spawning multiple copies of perl and PHP as I want now, managed entirely from the fastCGI process manager rather than from php itself (which would be rather frustrating to have since I have some accounts which receive enormous numbers of request and some that may get one a day- making it silly to have 4-8 interpreters for that user). And the SUEXEC problem just means that I need to add variables I"m trying to pass to SUEXEC's allow list so they're not stripped. Thank you for your help (now if only the question was actually worthwhile). -M ___________________________________ fastcgi-developers mailing list http://fastcgi.com/fastcgi-developers/