(RADIATOR) Radiator issue

[Sergio Alejandro Gonzalez Z (S2010)]

Hello Hugh.


I would like to know how many request per second a radiator
instance would answer on a machine with the next
characteristics:

Sun V240 4 GB RAM 2 UltraSparc IIIi processors of 1.2 Ghz.
Quad Gigabit Ethernet Card

My question came up from a weird behavior I had yesterday
on a machine with those specs that is handling about 150
cisco RASes.

Once the service was stalled (because the admin of such
machine accidentaly filled up the / filesystem) and once it
was cleaned, the radiator process was killed and ran it
again. The machine was about 6.1 of uptime but the process
of radiator was eating only about 7% of the cpu, and also
mysqld (for accounting) was eating 40%. 

Many requests coming to the 1645 and/or 1646 ports was not
responded, but when running another instance of radiator
(using the same mysqld process) in another auth and acct
ports, it worked well.


Should be a reason the excesive amount of requests per
secound the origin of such behavior?


Once I guessed the non answering behavior was produced by
the excesive requests coming from the RASes I restarted the
server (init 6) but once it was up, once again the
processes on the 1645 1646 ports didn`t work, and after
killing the process and started again manually, everything
went ok.


What should I look for?. Any kernel tuning advice?.


Thanks a lot in advance.


Best regards




Sergio Gonzalez
IT Engineer.

--
Archive at http://www.open.com.au/archives/radiator/
Announcements on radiator-announce at open.com.au
To unsubscribe, email 'majordomo at open.com.au' with
'unsubscribe radiator' in the body of the message.