Very Computer

Hi,
I'm a first timer with Apache so please bare with me.

I have been running Apache 1.3b3 for two weeks now on a Pentium II
w/dual 266
processors and 260 meg of RAM. Apache is configured to spawn 3 child
servers at start
up which it does about 15 to 20 seconds after the initial launch. At
times all 4 servers start up together and begin to use the CPU time all
at once and will not answer request until it's shut down and restarted.

The MaxRequestsPerChild is set to the suggested 10000  and a child will
only exit after the max requests about 10% of the time, leaving all 4
servers up and idle while the inital server holds on to the 50 or so
meg. Thus leaving the website unreachable.
Once this happens it takes 4 to 8 restarts of Apache to get it running
again, possibly because the memory has not been dumped and the children
can not allocate enough to kick in, so most of  the time I will have to
reboot the whole machine to clean out the RAM.  These outages appear to
happen about every two hours right about the time the max request is
reached.

The page's are served up lightning fast for the first 20 minutes of run
time. From there on it takes almost forever after the host is contacted
and waiting for reply.

Can anyone give me some insight on this misbehavior?  Any thoughts would
greatly be appreciated.

I would suggest you just wait for a 1.3b5 or later binary to be
available and try it with that.  There are known issues with 1.3b3
that could be causing this.

Quote:>Hi,
>I'm a first timer with Apache so please bare with me.
>I have been running Apache 1.3b3 for two weeks now on a Pentium II
>w/dual 266
>processors and 260 meg of RAM. Apache is configured to spawn 3 child
>servers at start
>up which it does about 15 to 20 seconds after the initial launch. At
>times all 4 servers start up together and begin to use the CPU time all
>at once and will not answer request until it's shut down and restarted.
>The MaxRequestsPerChild is set to the suggested 10000  and a child will
>only exit after the max requests about 10% of the time, leaving all 4
>servers up and idle while the inital server holds on to the 50 or so
>meg. Thus leaving the website unreachable.
>Once this happens it takes 4 to 8 restarts of Apache to get it running
>again, possibly because the memory has not been dumped and the children
>can not allocate enough to kick in, so most of  the time I will have to
>reboot the whole machine to clean out the RAM.  These outages appear to
>happen about every two hours right about the time the max request is
>reached.
>The page's are served up lightning fast for the first 20 minutes of run
>time. From there on it takes almost forever after the host is contacted
>and waiting for reply.
>Can anyone give me some insight on this misbehavior?  Any thoughts would
>greatly be appreciated.