In such a case you should give rightscale+ec2 a serious consideration.
It scales beautifully and is pretty easy to get going with and
maintain, plus monitoring is more-or-less built-in.
Sukrit
On 16-Mar-09, at 9:06 PM, John Campbell wrote:
I would like to thank everyone for their help and advice. I don't
think it is impossible to figure out what the underlying issue is, but
I made a bunch of changes based on the recommendations here.
1) I switched the hardware (pretty easy, since it's just a Xen
instance)
2) I upgraded Apache and PHP
3) I changed max requests per child to a sensible number rather than
unlimited. I think log rotate would achieve a similar effect by
restarting the webserver every day and I'll eventually go that route.
4) I set up monitoring scripts that will dump the output of `ps` and
`apache2ctl fullstatus` to a log in the event the apache craps out
again, and I'll be sure to get a gdb dump if it happens again.
5) For other reasons, I reduced the Keep-Alive time and moved most of
the static content to Amazon's CDN.
Thanks for the help.
Regards,
John Campbell
_______________________________________________
New York PHP User Group Community Talk Mailing List
http://lists.nyphp.org/mailman/listinfo/talk
http://www.nyphp.org/show_participation.php
_______________________________________________
New York PHP User Group Community Talk Mailing List
http://lists.nyphp.org/mailman/listinfo/talk
http://www.nyphp.org/show_participation.php
