IPCC 6.0 DR Test Fails - IPCC Expert needed

Unanswered Question

Hi all,

I am running a duplexed IPCC Enterprise 6.0 environment with duplexed Rogger, duplexed PG and duplexed IPIVR. And sideA and sideB are located at different state connected via a full DS3. I just did a DR test and found that side B will only work if either Rogger A or PG1a is online. If both RoggerA and PG1a are offline then the call center is dead agents can't login, calls can't go through.

Has anyone seen this type of issue before ? Can someone point me to the right direction on where I should start digging ?

Thanks very much ! and please let me know if you need more info or clarifications !!!

Danny

I have this problem too.
0 votes
  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 0 (0 ratings)
Loading.
david.macias Wed, 06/25/2008 - 02:45

Do the pims go active fairly quickly on side B after shutdown in side A? Are you using CAD or CTIOS desktop? Can you verify the IPs for both CTI servers are in the agent desktop registry?

david

Hello David,

Yes, The PIMs on PG1b go active right away as soon as the PG1a is offline but it seemed the PIMs on PG1b can only go active with RoggerA is online. If both RoggerA and PG1a are offline then the PIMs on PG1b are dead.

I am using CAD 6.0 and I've checked the registry and the IP of the both CTI servers are there.

Thank you very much David !!!

Danny

By type - just looking to confirm that there is only 1 PG in enterprise and by your description it sounds like it's a generic PG with 2 Pims on each side - normally Pim1 going to CCM and Pim2 going to IPIVR.

Next few things to add clarity is get a bit more detailed in testing.

1. Stop RoggerA - just for simplicity both Logger and Router processes. Validate that CCAG on RouterB still shows an active connection. If routing stops, both your PG's aren't configured right to talk to the B side router (hostname type issues)

2. Stop PG1A - validate again that CCAG shows 1 PG connected. If routing stops you're b side PG isn't configured correctly to talk to the Router (LID in PG1B, hostname, etc type issue)

3. Run RTTest if you can and capture that both peripherals and PG1 are all online - the same PG/Peripherals should be seen online as before you started testing. (if this is all online and active then need to keep moving out from the core of the system - CTI Server, CAD server, application, etc)

Few other ways to test this but without knowing more these should cover it...

Each of these is testing/validating another layer of the redundancy as with the information so far there's a few different points which are really being tested and need to isolate if and where there is an issue. Obviously something is at issue but need to narrow down the variables :)

Additionally - if this is a lab - then having both Public and Private networks riding over the same DS3 is fine - but if this is production, the system will not make it through A2Q nor should it ever be deployed in that scenario because you are implementing a single point of failure for the backbone of the ICM (and even with HW redundancy for the DS3 - you'll still have issues). It's a bit of a side bar from your question but it's one of those core issues waiting to happen.

Hope it helps!

Skippy

rbua Wed, 06/25/2008 - 10:01

Just to add to all this notes, have a dummy PG on side B, this should allow side B to stay up being connected to a majority of devices.

As said previously just the PG with no activated PIMs.

Regards,

Riccardo

rbua Wed, 06/25/2008 - 10:57

You just need to add a PG on the Roggers, when it comes to define the peripheral you could configure whichever one you want but you simply do not enable it.

Nothing really fancy with it.

Actions

This Discussion