cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
Announcements
Walkthrough Wednesdays
242
Views
5
Helpful
8
Replies
rnarayana
Contributor

ICM 4.6.1 & CCM 3.13a.

I have ICM 4.6.1 and ccm 3.13a ( PUB & SUB).

When I shut down the CTI Server in the PUB the iCM system never gets registred to the SUB.

In the PIM configuration I have specified IP address of both the Call Mangers as follows.

10.254.254.77;10.254.254.67.

Whats is the best method to provide redundency for this setup.

8 REPLIES 8
adignan
Collaborator

Verify CTI Manager is running on your Subscriber as well as DC Directory. We had a similar problem.

CTI is running on the SUB and also the DC directory.

Do you have duplexed PG's? If so, try having one PIM process on each PG, one PIM process communicating with CCM PUB, and the PIM process on the other PG communicating with CCM SUB.

Regards, Gary

Hey gary,

I am setting this up in the 4.6.2 verison in the duplexed mode.

I have configured one PIM for PUB in side A and one more PIM in Side B for SUB.

what happens in a situation when my PUb and PG2 goes down.

Do I need to configure two more PIM's. One more in Side A pointing now to SUB and one more in side B pointing to PUB.

Can you be little more detailed.

Riccardo Bua
Contributor

Hi,

please check the included diagram here to illustrate the configuration in ICM and CallManager required to ensure a successful fail over in the case of either PG or CallManager failure.

The black dotted lines indicate the connections between CTI Manager1 and all CallManager nodes in the cluster. The red dotted lines indicate the connections between CTI Manager2 and all CallManager nodes in the cluster.

There are two layers architecture for CallManager since CallManager 3.1 release. CTI Manager Service node is JTAPI application broker, which managers all devices within a CallManager cluster, while CallManager Service node is responsible for media processing.

Once you configure a group of CallManagers in a cluster by doing CallManager Groups and Device pool configuration in CallManager Admin page, IP address in PG configuration is real address for a CTI Manager. Do not use the comma separated IP addresses.

There will be peripheral offline event generated in the event of either PG1A or CallManager failure. CallManager failure here means CTI Manager failure , the whole CallManager box failure, or network outage between PG1A and CallManager box. The peripheral offline event would drive IPCC fault tolerance system to fail over to PG1B

If there is one CallManager service node failure, there will be no periheral offline event generated ( no PG failover occurred). However, CTI Manager will re-register all devices from failed CallManager node with the secondary CallManager service node configured in device pool.

PG1A PG1B

| |

CTI Manager1 CTI Manager2

/ | \ / / |

/ | \ / / |

/ | / \ / |

/ | / / \ |

/ / | / \ |

/ / | / \ |

/ / | / \ |

CallManager1 CallManager2 ..... CallManager6

The diagram got a bit messed up, it should show the PG1A pointing to the primary CTI Manager1 and PG1b pointing to the secondary CTI Manager, both of them interconnected and connected to the other nodes.

You mean I need to just give th IP address of the Active call manger.

I do not have PG1B. I just have this set up with 1 PG for the CCM and One pG for the IVR.

Hi,

what happens in the evnt of Publisher failure and also PG1B Failure.

Raj.

Content for Community-Ad

Spotlight Awards 2021