[asterisk-dev] should asterisk die for chan_zap misconfiguration?

Tzafrir Cohen tzafrir.cohen at xorcom.com
Tue May 8 17:18:11 MST 2007


On Mon, May 07, 2007 at 06:43:02PM +0300, Tzafrir Cohen wrote:
> Hi
> 
> I'd like to raise http://bugs.digium.com/view.php?id=7290 once again.
> 
> I've been using it for quite a while and found it very useful. Together
> with 'zap restart', it allows a sane handling in chan_zap config issues
> (e.g: when asterisk is not restarted every minute ;-)  ).
> 
> That patch is probbably dated, and may need some technical changes. But
> would such a change be acceptable?

Reopen the bug?

Some quick notes:

* I couldn't get trunk's autoconf to identify zaptel with an explicit 
  --with to either 1.4 or trunk svn zaptel and with include/ and zaptel/ 
  as symlinks to .)
  1.4 built just fine.

* The patch there is over-simplistic in that that it does not attempt to 
  leave the partially configured channels in a "clean state". What
  should be such a "clean state"? Destroy channels? What about spans?
  (zap_restart actually needs exactly the same "destroy all of zaptel"
  operation).

* In 1.4 and trunk when there is an error while generating channels from
  users.conf, the error from process_zap is ignored.

-- 
               Tzafrir Cohen       
icq#16849755                    jabber:tzafrir at jabber.org
+972-50-7952406           mailto:tzafrir.cohen at xorcom.com       
http://www.xorcom.com  iax:guest at local.xorcom.com/tzafrir


More information about the asterisk-dev mailing list