Can a PCS support a SIP trunk when ComServer fails?

[sonoma76]

Hi I need your help and advice with SIP and Alcatel, a system I'm totally unfamiliar with.

We have two sites with an MPLS cloud between them.

Site A has an Alcatel OmniPCX running R8.0.1-g1.503-27-us-c80s1.

Site B has an Alcatel PCS server running R8.0.1-g1.503-27-us-c80s1

We ordered Site B's PCS server in December, intending to install it with SIP trunk, based on the advice of our vendor. VM would be hosted on Site A's PCX. But we have a huge problem.

MPLS provider has provisioned a SIP trunk with 20 DIDs pointing to Site A's OmniPCX because, our vendor says, Site B's PCS doesn't support a SIP trunk. So, mid-activation, we had to move the SIP trunk from Site B where we intended it to go, to Site A.

We went ahead with the activation. Either we misunderstood or our vendor isn't good with SIP but, at the PCS at Site B, we thought we were getting true remote shelf survivability. Maybe it can't host a SIP trunk full time, but if I have a secondary SIP IP from the provider that can failover in the case of a power outage at Site A, PCS Shelf at Site B should be able to host those 20 DIDs.

We went back to this with our vendor and they said it was worth a try. So we set up another activation with the provider pointing a secondary SIP IP at Site B's PCS, powered off Site A's router, and watched as the PCS took over. It failed. Site B's phone numbers rang busy. And our vendor is saying the equipment can't handle SIP survivability...that it's only designed for traditional PRI circuits.

Meanwhile I'm reading that in release 9 of the OmniPCX software, there was some SIP features released for PCS.Am I reading that right? Are we one version back for the PCS?

Anyway we're pretty desperate here having sunk nearly $20k into a PCS box that is largely pointless. Any help or guidance you can offer would be helpful. Thanks.

[alex]

Can a PCS support a SIP trunk when ComServer fails?
Yes, from release 9.1 external SIP gateway rescue is supported.
Basically IMHO it went wrong from the very beginning. It is a 100% misunderstanding in a network design.
Looks like spatial redundancy solution (with two identical CS, one active and another on hot-standby) much better fitted your requirements the\an use of PCS, because in case your site B rack looses IP-connectivity its CS could handle all features without any restriction.
If you decide to go to R9.1 then, of course, you can still stay with PCS.

[sonoma76]

Wow, thanks, that's good to hear. So this is a feature even though we're using a PCX 4400 with what I think is the Crystal Omni Exchange Enterprise?

Our vendor is saying both the PCX at Site A and the PCS at Site B have to be on the same software versions, is that true?

[alex]

So this is a feature even though we're using a PCX 4400 with what I think is the Crystal Omni Exchange Enterprise?

It is better to put it this way - PCS supports CS and AS architectures. If you use CPUs then your choice is spatial redundancy.

Our vendor is saying both the PCX at Site A and the PCS at Site B have to be on the same software versions, is that true?

That's true for any configuration.

[Konstantinos.E]

hi

if you wait a bit more release 10 is on the way.

but still you need an amount of money to upgrade...