Tranlation patterns vs. CSS: loads on a server

[Tommer Catlin]

We as in deployment engineers, see all kinds of crazy things in the field. Deployment strategies change between clients or certain environments. My question is this:

What will load up a server more with 80 sites:

Site1_CSS_Internal

Site1_PT_Phones

Site2_PT_Phones

Site3_PT_Phones

Site4_PT_Phones

Site5_PT_Phones

(all the way to 80)

Granted, CCM is going down the list to get to the correct PT where that number resides. Each site may have 30 phones max at it. If it was 200 at each site, i would be a little more worried utlizing processing power.

This is taking into consideration that all DID lines have unique extensions.

Scenario 2:

Site1_CSS_Internal

Site1_PT_Phones

Translations to/from each site. This would be a total of 80 translations per site. So 80*80 = 6,400 translations.

Granted, if a customer got to 80 sites even, at what level does the customer consider a Gatekeeper to handle all call routing?

Any comments on this?

Thanks!

[mchandak]

One thing to look at here is how many CallManager cluster's do you have. If all these devices are a part of the same cluster, then I dont think you can use a Gatekeeper. Gatekeepers should be considered depending on number of CallManager clusters along with dial-plan pattern.

Hope this helps

[Tommer Catlin]

It's just one cluster right now. If we every get to 80 sites or the client, they will need more subscribers or to break it out farther I believe. I was just wonder what everyone thought about the scenarios. Is one better than the other?

[mchandak]

With one cluster in picture, CallManager can natively handle the routing. If we have multiple CallManager clusters then using a Gatekeeper could be a possible solution. In this setup, I dont see a reason to use a Gatekeeper.