Thx Alan I coded a very simple script to check the mysl server latency: #!/bin/sh ExecQuery () { echo "select count(*) from radacct;" | mysql -h mysqlserver -u radiususer --password=radiuspassword radaccoutingdb } while [ -n 1 ] do echo -n "starting query....." date time ExecQuery echo -n "query finished......" date echo "" echo "" sleep 300 done Executing this script during last 24h (with nohup), from the radius machine, I got always latencies below 1 second........ How can the script get always latencies under 1sec, and the radius threads on teh other hand get latencies over max-request-time? I'm a little bit lost.... :( thx again On Jueves 26 Febrero 2009 11:00:31 Alan DeKok wrote:
magicboiz wrote:
Do you know how can I modify the source code, in order to print in the error message the exact time when the child was created/forked? If I want to find errors in my sql backend server, I need to know the exact time when the requests are lost...... checking the code, It looks like this error message is printed only when the server detects threads that are stalled.....................
That message is printed "max_request_time" seconds after the packet is received. The stall happens at some point during that time period.
Alan DeKok. - List info/subscribe/unsubscribe? See http://www.freeradius.org/list/users.html