3CX Failover Dialing Out

[Rhys Haskell]

I have a failover server setup and everything fails over as it should. I have 2 sip trunks setup with the IP's of the servers. When a failover occurs it takes 10-15 seconds to dial out as it tried the first sip trunk and then after about 10 seconds tries the 2nd sip trunk. Is there any way to lower the amount of time it takes to try the 2nd trunk?

 

[YiannisH_3CX]

Hello @Rhys Haskell

The time it takes to try the second route depends on the reply from the provider. If no reply is received then it will take 32 seconds to switch to the second trunk as per RFC.

 

[leejor]

What is the actual failure on the first trunk? Is it an ISP issue, or do they simply fail to respond at times? Does the first trunk register? If so, and that fails, then calls should not be attempted on that trunk at all. Can the re-registration time be lowered and still be acceptable to that provider?

 

[Rhys Haskell]

What is the actual failure on the first trunk? Is it an ISP issue, or do they simply fail to respond at times? Does the first trunk register? If so, and that fails, then calls should not be attempted on that trunk at all. Can the re-registration time be lowered and still be acceptable to that provider?

The SIP trunk is IP based registration so we use 2 trunks with the primary server IP and the backup server IP.

 

[cobaltit]

@Rhys Haskell

Is this really a problem? I think 15 seconds for recovery is pretty good. If your instances are failing often enough for this to make a difference you have a different problem .

 

[Rhys Haskell]

@Rhys Haskell

Is this really a problem? I think 15 seconds for recovery is pretty good. If your instances are failing often enough for this to make a difference you have a different problem .

This is just a request that has been made by a customer during DR testing. The instance hasn't failed yet and hopefully never will.

 

[YiannisH_3CX]

Unfortunately this is not something that can be controlled as it is done automatically by the PBX. You could however use a different outbound rule with a prefix and inform the users to use that in case of a failover occurs.
Alternatively i would recommend posting this as a feature request to our ideas section of the forum https://www.3cx.com/community/forums/3cx-ideas/ . All posts are reviewed by their corresponding Departments and, if feasible and voted for by other users as well, may be implemented in an upcoming service pack or update.