What is the real meaning and use of cache-size in ippool declaration [resend]

Florin hatespam00-mlist at yahoo.com
Fri Mar 23 20:06:05 CET 2007


Sorry guys, this is in fact a resend of a previous email, now using 
Thunderbird in an attempt to avoid sending HTML format message.
Hope it'll be OK this time.

~~~~~~~~~~~~~~

Hi everyone,

FreeRadius 1.0.1 here, the one which comes as standard with RHEL 4.

I want to use a whole class B subnet (172.16.0.0/16) in an ippool 
declaration to assign IP addresses from. Now it is said that the 
cache-size parameter should be equal to the number of IP addresses in 
the pool. But if I try to set it to 65536, the server won't start (it'll 
crash actually).

So what is the **real** meaning and use of this parameter ? You have to 
specify the start and end of the pool, so then why set another parameter 
with a value that can be trivially calculated by the server anyway ?

And more importantly, what are the right guidlines on setting the value 
for this parameter ? Making it equal to the number of IP addresses in 
the pool won't work and it just doesn't make any sense to me.

Finally, if I change the value of this parameter, beside restarting 
FreeRadius, do I have to do anything else ? Do I have to delete the 
related database and index files in order for the change to take place ?

Thank you in advance.

Best regards,
Florin




More information about the Freeradius-Users mailing list