load and threads
Velikanov
alex at asu.farlep.net
Fri Feb 17 12:24:39 CET 2006
Good day.
I use FreeRadius 1.1.0 with Oracle DB on
Red Hat Enterprise Linux ES release 3 (Taroon Update 3)
Kernel 2.4.21-20.ELsmp on an i686
I made a test load on server and everything is good when I start Freeradius
in
debug mode
# /usr/local/sbin/radiusd -X
top shows me
13:04:10 up 70 days, 1:06, 5 users, load average: 1,77, 1,57, 1,88
114 processes: 111 sleeping, 3 running, 0 zombie, 0 stopped
CPU states: cpu user nice system irq softirq iowait idle
total 66,7% 0,0% 3,5% 0,0% 0,0% 0,3% 29,2%
Mem: 2307652k av, 2283924k used, 23728k free, 0k shrd, 116652k
buff
1578192k actv, 300500k in_d, 36092k in_c
Swap: 2048152k av, 219780k used, 1828372k free 1694328k
cached
But when I start Freeradius in work mode
# /usr/local/sbin/radiusd under the same test load
top shows me in 10-15 minutes:
11:50:27 up 69 days, 23:52, 3 users, load average: 37,25, 30,91, 31,58
114 processes: 92 sleeping, 22 running, 0 zombie, 0 stopped
CPU states: cpu user nice system irq softirq iowait idle
total 91,1% 0,0% 8,7% 0,0% 0,0% 0,0% 0,0%
Mem: 2307652k av, 2250580k used, 57072k free, 0k shrd, 151768k
buff
1541388k actv, 299248k in_d, 35108k in_c
Swap: 2048152k av, 178648k used, 1869504k free 1671720k
cached
I see radiusd threads when do the command
#ps -efm | grep radiusd
in radius.log there are very many such string:
Fri Feb 17 12:08:40 2006 : Error: WARNING: Unresponsive child (id
1347701680) fo
r request 2937
and such string:
Fri Feb 17 12:06:16 2006 : Info: rlm_sql (sql): There are no DB handles to
use!
skipped 0, tried to connect 0
My questions are:
1. What is the reason of such FreeRadius behavior? What must I do?
2. Why load of server during #radiusd -X is lover then during #radiusd ?
Seems to be vice-versa?
3.How must I start FreeRadius to make many processes (not threads), maybe it
help me to decrease the load of server and to increase the speed of packets
handling
Thanks.
More information about the Freeradius-Users
mailing list