[Serdev] Re: was [Serusers] Timers

[Richard]

Hi Jiri,

More thoughts on the timer issue.

The main issue to have multiple timer is the overhead
to insert the element into the timer list. If only
fixed timer is used, the event can be added at the end
of the list and incur little cost. If we maintain a
fixed timer list and use a flexible timeout value for
an element, we can still add the element to the end of
the list. When the element is called, the timeout
value is reduced by fixed time amount and put back
into the end of the list. The repeats until the
timeout value becomes 0 or negative. Then the real
processing is called.

For example, the fixed timer is 20 second and the
invite timeout for a call is 60 second. The
t_on_failure is inserted to the timer list with a
timeout value 60 seconds. First time it is called
after 20 seconds, the timeout value becomes 40 seconds
and the function is re-inserted back to the list. Then
second time, it becomes 20 seconds. In the third time,
it reduces to 0 and the callback function is finally
executed.

So the granularity of this is the fixed timer value,
not the real timeout value specified. Most people
would accept this. If someone wants really accurate
value, they can use 10 seconds or even 5 seconds.

Any comment?

Thanks,
Richard

--- Jiri Kuthan <jiri at iptel.org> wrote:
> There is only single-timer value in the current SER
> code.
> If all of your SER code uses stateful forwarding, it
> will
> use a single timer. Use two SERs, use stateful
> forwarding
> selectively, or wait till someone develops a
> different
> timer system.
> 
> -jiri



	
		
__________________________________
Do you Yahoo!?
Friends.  Fun.  Try the all-new Yahoo! Messenger.
http://messenger.yahoo.com/