Resiliency Basic Setup

[mitelmitel]

Hi, can you just assure me this is all I need to get a basic resiliency situation setup on MCD 4.0 on a cluster of 5 3300's [1 x remote 3300, 1 x user gateway, 3 x trunk gateways] with Ent & Ops Manager. All linked by IP Trunking.

I want some hotdesk, and acd hotdesk users on my User Gateway to fail over to my Remote site.

I currently do not have resiliency setup.

I have 40 spare device, user, and ACD licenses on the remote site. Do I Need To Purchase Anything Else From Mitel To Enable Resiliency?

Is this OK?:-

1] In Ops Manager program my ACD and Standard Hotdesks to have a secondary controller [Remote Site].
2] Dir Sync

3] Program through ESM my Agent Skills Groups to have a secondary controller [Remote site].
4] Dir Sync


Thanks for your help.

PS. I am waiting for MCD 5.0 before upgrading my cluster.

 

[LoopyLou]

What's on the trunking gateways? If you have your PRI's split then the best ACD resiliency is to have a path on the gateway with ACD hotdesk agents on the user 3300. Make those ID's and the sets resilient to the secondary. Duplicate the same path on another gateway with the same reporting number and add the same agents and groups to both paths. If calls come to the ACD queue via both gateways then if one gateway fails, calls can still reach the queue via the other gateway. If the user 3300 fails the agents and sets fail to the secondary with no loss of calls. If you want the paths to have a name then create path 3000 on gateway one and 3001 on gateway 2 ( with the same reporting number, agent groups and ID )'s and then create a speed dial on gatway 2 pointing 3000 to path 3001. That, unless I am missing something, is a fully resilient ACD setup.

For licenses you need enough device licenses to support the failing over phones and ACD licenses to support the failed over agents.

There may be no i in team but there are three f's in fudge off.

 

[mitelmitel]

Our network is not as simple as that!

We have 3 trunk gateways. 1,2,3 all used for ISDN30s. Inbound calls hit 3,then overflow when busy to 2, then 1.

We have a system speed call for every external number we own [around 800].

Trunk gateway 3 has all our paths and also IQ attached.

All calls have to hit IQ.

In order for IQ to pick up on the correct DNIS I have had to setup the routing as follows:-

Call gets overflowed on PSTN to HT1:
HT1 accepts that last 6 digits of DDI [888555] which is a system speedcall that routes it to the node ID of trunk gateway 3 = 888555 => 6663888555

This then passes the call to trunk gw 3 and strips the node ID off so that 888555 is referenced in its own sys sp call form and send it to IQ HG:
888555 => 5400

IQ picks up on 888555 as the DNIS and references the call flow.

The call then gets sent to the relevant path on HT3.

For full resiliencey I need another IQ server with load sharing between gw2 and gw3.

Path resiliency will be difficult with the above and also IQ administration of 2 servers without some form of syncing the databases!

I could use ARS to pickup on a 3300 failing and modify the path DN and route it back in on itself to send it to a cloned emergency path.

If I use the same P number on two different paths I take it CCM will report on the two paths as if they were one?

Was my method of programming resiliency through Ops+ESM OK [in my first post]?

 

[LoopyLou]

Confused as to why you are routing with code ID. Are the gateways not within the cluster?

The phones the agents log into would also have to be made resilient.

There may be no i in team but there are three f's in fudge off.

 

[mitelmitel]

To pass the DNIS through the cluster to the IQ when calls were sent to overflow gateways. Believe me, I developed the programming because Mitel couldnt figure it out!

Anyway, I have redesigned my cluster now and have developed a method of load sharing IQ ports through the TGs and virtualising the IQ server, whilst also setting up ARS and IP Loopback trunks for GW3 failure to resilient paths, etc.

Thanks

 

[Supernova99]

To pass the DNIS across you need to have a redirect for each DNIS in your call flow then if the call hits the secondary controller it will then match the redirect number and complete correctly.
I know it is a pain to program but for every DNIS you will need to add a redirect. All DNIS numbers on the secondary controller will need to be call rerouted always to the IQ hunt group (so that redirect is detected)

Share what you know - Learn what you don't

 

[mitelmitel]

I have around 1100 DNIS numbers, maybe a lot more! But thanks for that info!

I cant wait until WFM is released!

 

[Supernova99]

Me either!

Share what you know - Learn what you don't

 

[LoopyLou]

WFM?

The single biggest problem with communications is the illusion that it has taken place.

 

[Supernova99]

work forcr management which will be in rel 6.0

Share what you know - Learn what you don't