Hello people,
I've been implementing the DNS SRV feature of the SPA122. It nicely works and I am really happy about it but I would like to tweak it or at least understand how Cisco implements it.
I have two servers; one being the primary and the other one being used only if the primary is down (or the SPA can't reach out for some reason). Some SPA do switch at times onto the second one.
Here's a snippet of the config
<Proxy_1_>mydomain.com</Proxy_1_>
<Register_Expires_1_>600</Register_Expires_1_>
<Use_DNS_SRV_1_>Yes</Use_DNS_SRV_1_>
<DNS_SRV_Auto_Prefix_1_>Yes</DNS_SRV_Auto_Prefix_1_>
<Proxy_Fallback_Intvl_1_>600</Proxy_Fallback_Intvl_1_>
<Proxy_Redundancy_Method_1_>Normal</Proxy_Redundancy_Method_1_>
The SPA will return to the primary as expected after 600 seconds.
Now what I need to better understand is what makes the SPA to use the second server in the list and how I can make it a little more permissive; meaning "try a little more with the first server before going to the other" ...
I'm thinking it is the SIP timers that need to be tweaked but I don't really want to "try values and see what it does".
Also, I was thinking of adding multiple entries that would point to the primary server so if SPA has to switch, it would hit the primary server again ... do that one or two times and then go to the secondary. What I'm not sure is if the SPA will accept to have multiple entries that resolves to the same IP.
If a Cisco Engineer or anybody that has a deep understanding of how SPA deals with DNS SRV could provide an indepth technical explanation, that would be awesome.
Thanks,
Benoit Raymond
TelSip Communication
