too many idle processes
От | Soon-Son Kwon |
---|---|
Тема | too many idle processes |
Дата | |
Msg-id | 20020521165143.A15658@mail.kldp.org обсуждение исходный текст |
Ответы |
Re: too many idle processes
Re: too many idle processes |
Список | pgsql-general |
Hello: I am running pgsql on debian GNU/Linux woody box to run "sourceforge" program. (http://www.freesoftware.fsf.org/debian-sf/) Here is part of the /etc/postgresql/postgresql.conf: --- debug_level = 0 log_connections = on log_pid = on log_timestamp = on syslog = 2 # if syslog is 0, turn silent_mode off! silent_mode = off syslog_facility = LOCAL0 trace_notify = off max_connections = 64 # shared_buffers must be at least twice max_connections, and not less than 16 shared_buffers = 128 # TCP/IP access is allowed by default, but the default access given in # pg_hba.conf will permit it only from localhost, not other machines. tcpip_socket = 1 --- But when I run "ps aux | grep sourceforge" to see if how many pgsql processes are running, the numver reaches the maximum connections very quickly even if only few users accesses it. Access to the pgsql(7.2.1) is done by php(4.1.2) and web server is apache(1.3.24). The program uses "mysql_pconnect()" function but it seems something prevents reusing the existing pgsql socket connection, hence the pgsql always spawns a new process which then reaches the maximum connection possible. 1. Could anyone please give me a clue how to deal with this situation? Recalling my previous MySQL experience, I had similar problem but could cope with it by setting "wait_timeout" variable small which kills idle process if it has been idle for that specified time. 2. Is there such configuration variable in pgsql? Thanks very much.... -- -*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-* (o_ **WTFM** (o_ (o_ //\ (/)_ (/)_ V_/_ http://kldp.org -*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*
В списке pgsql-general по дате отправления: