[Users] Best practice for DNS failover using OpenSER?
jiri at iptel.org
Wed Jan 17 18:19:10 CET 2007
At 16:46 16/01/2007, Bogdan-Andrei Iancu wrote:
>the TM functions do that, but again, in most of the case it, as the relaying is done at the end of processing, it can be a wast of computation and time - just to apply all changes and perform ops to figure out at the end that the request was a retransmission.
That's a trade-off but I guess that for most uses the threat caused by blocked
processes is more compelling that the penalty caused by extra transaction lookup.
>So, you can use the t_check_trans() function from the beginning:
Is it safe against race conditions? (I mean after the function was left,
can its result be obsoleted by another process running in parallel?)
>the great part is that the function does not affect the further processing and has no computation penalties.
>Users mailing list
>Users at openser.org
Jiri Kuthan http://iptel.org/~jiri/
More information about the sr-users