Is there anyway to have a survivable remote site and not have the re registration down time?

[GregoryF]

We have two sites, the main center (MC) and the backup center (BC), each with two servers. Our BC needs to be able to act as an overflow to the MC during times of high call volume and each of the sites needs to be able to survive on it's own either due to a scheduled outage or the proverbial "smoking hole" scenario. Is there a way to set this up to avoid the re registration and the no service timer? We are a 911 center and cannot afford to have downtime for the phones, even if it is only for three minutes.
We are being upgraded to CM6 running in a virtual environment.

 

[montyzummer]

Phones need to register to something and if that one thing fails then they look for something else (the scenario you describe) , the only thing i could suggest is pay for more hardware on each site to cover larger call flows , or go down the bsm / off site session manager route and sip endpoints.





APSS (SME)
ACSS (SME)
ACIS (UC)

 

[GregoryF]

It looks likes BSM is the way to go, if I am understanding it correctly.
If I have a session manager running at each site the SIP phones will register to both and if the WAN goes down between sites then the phones will fail over to the session manager they see as still active with out interruption to service.

 

[GregoryF]

Apparently session manager only works with SIP and no H.323?

 

[montyzummer]

Correct and correct

APSS (SME)
ACSS (SME)
ACIS (UC)

 

[GregoryF]

Is there a similar solution for the H.323? One that would allow either local or dual registration so there won't be any interruption in phone service?

 

[montyzummer]

The one you have with duplex core servers is really quite hardy , the only way i could think off is to have SIP trunks so if everything fails the SIP provider at network level can point calls at whatever devices you want , cellular , analogue lines ...this is an expensive soloution but if you really need this type of redundancy duplex core servers paired off like you have with network level disaster recovery is probably the way to go.

APSS (SME)
ACSS (SME)
ACIS (UC)

 

[GregoryF]

Our biggest issue is scheduled outages. We will occupy both the main center and the backup center. Once our call takers are logged into the system they start answering calls. Our trunks, both PRI and E911 CAMA are distributed between both sites and will ring in both locations. When the WAN link is broken we will have a down time were we cannot receive / answer calls at our BC while the IP (H.323) phones are re registering to the servers at the BC. It is the period of time it takes to re register that is unacceptable since this is a 911 center.

 

[montyzummer]

So will the DR plan at network level not work , i know you would have to hands out a load of cheap pay and go cell phones but they would at least handle incomming calls until registration happens

APSS (SME)
ACSS (SME)
ACIS (UC)