Bit of a strange one here.
We have a Avaya S8700, which has histoically been connected to two competely seperate PG Pairs on two ICM Platforms. Up to now this has been via one MAPD and one CVLAN Server. This configeration has never caused any issues.
We are currently in the process of upgrading this switch to a S8720 V 3.1.4 and as part of the upgrade we will change the CVLAN Server for 2 AES servers. For the time being the MAPD will stay in place.
After the upgrade all appears well, until we test a failover on the PGs connected to the new AES servers. The failover appears to go well, but then the call centre CTI never recovers and a Vector trace indicates that the call is failing the Adjuct route step. Here comes the bizarre bit - If the MAPD CTI link is now busied out this appears to fix the issue.
Avaya are continuing to investigate but are saying that this may be as a reault of the ICM that is using the MAPS taking over the 3rd party domain control on certain skills \ splits before the ICM using the AES has a chance to restablish.
Now we have admistered in configeration manager some skill groups on BOTH ICMS that are in fact belonging to agents on the OTHER ICM. We have done this as the ICM error log was reporting thousands of errors daily for every event ( such as agent login ) on skills that are not entered into configeration manager.
So basically my questions are:
1 - Does Avaya's statement sound correct and is anyone else have experiace on useing both a MAPD and AES pair on one switch.
2 - Is there anyway of telling ICM not to monitor a skill group but without haveing thousands of errors in the event viewer.
Please let me now if you have any questions!
Many Thanks
M
We have a Avaya S8700, which has histoically been connected to two competely seperate PG Pairs on two ICM Platforms. Up to now this has been via one MAPD and one CVLAN Server. This configeration has never caused any issues.
We are currently in the process of upgrading this switch to a S8720 V 3.1.4 and as part of the upgrade we will change the CVLAN Server for 2 AES servers. For the time being the MAPD will stay in place.
After the upgrade all appears well, until we test a failover on the PGs connected to the new AES servers. The failover appears to go well, but then the call centre CTI never recovers and a Vector trace indicates that the call is failing the Adjuct route step. Here comes the bizarre bit - If the MAPD CTI link is now busied out this appears to fix the issue.
Avaya are continuing to investigate but are saying that this may be as a reault of the ICM that is using the MAPS taking over the 3rd party domain control on certain skills \ splits before the ICM using the AES has a chance to restablish.
Now we have admistered in configeration manager some skill groups on BOTH ICMS that are in fact belonging to agents on the OTHER ICM. We have done this as the ICM error log was reporting thousands of errors daily for every event ( such as agent login ) on skills that are not entered into configeration manager.
So basically my questions are:
1 - Does Avaya's statement sound correct and is anyone else have experiace on useing both a MAPD and AES pair on one switch.
2 - Is there anyway of telling ICM not to monitor a skill group but without haveing thousands of errors in the event viewer.
Please let me now if you have any questions!
Many Thanks
M
