This is a discussion on Inconsistency of number of httpd processes within the Linux Web Servers forums, part of the Web Server and Related Forums category; $ pwd /punopo/oracle/product/ias_902/Apache/Apache/logs $ ../bin/httpd -v Server version: Oracle HTTP Server Powered by Apache/1....
|
|||||||
| FAQ | Members List | Calendar | Search | Today's Posts | Mark Forums Read |
|
|||
|
$ pwd
/punopo/oracle/product/ias_902/Apache/Apache/logs $ ../bin/httpd -v Server version: Oracle HTTP Server Powered by Apache/1.3.19 (Unix) Server built: Jan 6 2003 10:46:06 $ ps auwx | grep -c [h]ttpd 274 $ fuser error_log error_log: 14346 14347 14400 14401 14402 14403 14404 14405 14406 ....[snipped] $ fuser error_log | perl -nle '$_=s/ +//g;print' #how many spaces in output 77 $ ps auwx | grep -c [h]ttpd 234 $ grep MaxClients ../conf/httpd.conf #MaxClients 150 MaxClients 512 Only approximately 77 of 274 open error_log. Why? In fact, the reason I post this message is that before last Friday, our MaxClients was set to 150. And we still saw about 300 httpd processes. That's a more interesting puzzle. Apache hung and we restarted it with a higher MaxClients, although that may not be the cause of the hang (grep MaxClients error_log didn't show anything). So two questions. Is less than 1/3 of httpd's opening error_log normal? How do we explain MaxClients was 150 and there were 300 httpd's? Thanks. Yong Huang |