status_check and response_timeouts in proxy.conf documentation

[Alan DeKok]

On Jan 25, 2019, at 2:33 AM, Fekete Tamás <fektom at gmail.com> wrote:
> 
> Dear  Colleagues,
> 
> I would like to ask help with interpreting proxy.conf file documentation as
> I found contradiction between two sentences on one place and I have another
> question regarding response_timeouts variable, also.
> 
> The contradiction comes because of these sentences:
>        #  If "status_check" below is something other than "none", then
>        #  the server will start sending status checks at the start of
>        #  the zombie period.

  The start of zombie_period means that FreeRADIUS thinks the home server is dead.  But since UDP is lossy, it can't *know* that the home server is dead.

> And documentation says just few lines below:
>        #  These status packets are sent ONLY if the proxying server
>        #  believes that the home server is dead.

  Yes, that's true.

> So my question is that when the status packets start to go out by default?
> At the starting of dead status or at the starting of zombie status?

  Zombie *is* dead.  See zombie movies for further explanation.

> My second question is in relationship with response_timeouts variable. The
> documentation says about response_timeouts:
>        #  Start "zombie_period" after this many responses have
>        #  timed out.
> 
> However it raises a question, namely, how long can be the gap between the
> two timed out response? Can the first one happen at 1pm and the second one
> at 2pm and Freeradius counts these two together with many successful
> requests in between, or the two responses must to follow each other in row,
> or which factors are taken into account?

  It means that FreeRADIUS has missed that many responses *in a row*.  i.e. all together.

  When FreeRADIUS is receiving replies from a home server, then the home server isn't dead.  And FreeRADIUS doesn't mark it as dead.

  Alan DeKok.