Thanks I will check this. I have another concern related to performance. on RHEL machine, running ab for 20000 requests with 100 concurrent requests takes almost 46 seconds, while this was taking just 2 seconds in 2.2 Please can you tell if this is the case and how can this be improved?
Thanks On Sat, Sep 24, 2016 at 1:02 AM, Stormy <storm...@stormy.ca> wrote: > At 12:18 AM 9/24/2016 +0530, Rashmi Srinivasan wrote: > >> Hi, >> >> This is the output of ulimit -a >> > [snip] > >> max user processes        (-u) 1024 >> > > Pure guesswork, but this looks low -- on a 32bit system, try 2^15; on > 64bits, 2^16 ... or more ... > > P. > > virtual memory      (kbytes, -v) unlimited >> file locks            (-x) unlimited >> >> Thanks, >> >> On Fri, Sep 23, 2016 at 8:25 PM, Rubén Toribio Aldeguer < >> rtori...@riu.com> wrote: >> I think your problem is not Apache, but your OS. >> I think Apache 2.2 and 4.4 uses diferent mpm mode (my memory is poor). >> I think you OS is limiting the number of Opens Files that can Manage. >> (safety reasons) >> Please, can you print you ulimit output, as I advise on a previous mail? >> >> Thanks. >> >> 2016-09-23 11:55 GMT+02:00 Rashmi Srinivasan < >> rashmisrinivasan2...@gmail.com>: >> Hi, >> Do you feel this is the issue with some configuration or would this be an >> issue with apache itself. >> Has anybody encountered a performance issue of this kind. >> Please comment >> >> Thanks, >> RS >> >> On Fri, Sep 23, 2016 at 1:43 PM, Rashmi Srinivasan < >> rashmisrinivasan2...@gmail.com> wrote: >> Is there a minimum requirement for apache 2.4 to run these tests? >> I have tried on different machines with different configurations with the >> same outcome. >> Please advise. >> >> Thanks, >> RS >> >> On Fri, Sep 23, 2016 at 1:14 AM, Rashmi Srinivasan < >> rashmisrinivasan2...@gmail.com> wrote: >> Hi, >> Thanks for checking this, >> >> I wanted to check the below:' >>     Stress testing of apache 2.2 works very well with the default >> mpm worker onfiguration >>     However, apache 2.4 doesnt work at all with the default >> configuration, gives the error: >>      2000 concurrent requests >>    Benchmarking x.x.x.x (be patient) >>    socket: Too many open files (24) >> >> After changing the configuration, as described in the previous email, it >> starts processing them: >> Lots of timeouts are occurring. >> Error in the error_lo is: >>               "AH00046: child process still >> did not exit, sending a SIGKILL" >>               (2)No such file or directory: >> [client x.x.x.c] AH01256: connect #7 to cgi daemon failed, sleeping before >> retry >>                11)Resource temporarily >> unavailable: AH01252: couldn't create child process: 11: >> I also noticed that the CPU utilzation reached 99.9% >> Which is when apache terminates >> >> What changes are to be done in this case. >> Please advise >> >> Thanks, >> RS >>  >> >> On Thu, Sep 22, 2016 at 9:54 PM, Yann Ylavic <ylavic....@gmail.com> >> wrote: >> On Thu, Sep 22, 2016 at 5:20 PM, Rashmi Srinivasan >> <rashmisrinivasan2...@gmail.com> wrote: >> > >> > Access log has the following status: >> > [22/Sep/2016:20:17:21 +0530] "GET /cgi-bin/script.cgi HTTP/1.0" 200 16 >> >> What's in the *error* log (path from the ErrorLog directive)? >> What about resources (CPU, load average, ...) on the httpd machine >> during the stress? >> >> Regards, >> Yann. >> >> --------------------------------------------------------------------- >> To unsubscribe, e-mail: users-unsubscr...@httpd.apache.org >> For additional commands, e-mail: users-h...@httpd.apache.org >> >> >> >> >> >> >> >> -- >> >> Rubén Toribio Aldeguer >> Técnico Sistemas DataCenter >> Informática à rea Sistemas >> (+34) 971743030 >> www.riu.com / www.riuplaza.com >> >>            >>  >> >> This e-mail and its attachments, if any, are confidential and may be >> legally privileged. If you have received it in error, you are on notice of >> this status. Please do not copy or use it for any other purpose or disclose >> its contents to any other person: to do so could be a breach of confidence. >> You may contact us at +34 971 74 30 30 or at sender's e-mail address. >>  Please, consider the environment before printing this email. >> >> > > --------------------------------------------------------------------- > To unsubscribe, e-mail: users-unsubscr...@httpd.apache.org > For additional commands, e-mail: users-h...@httpd.apache.org > >
