cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
518
Views
10
Helpful
4
Replies

911 Compliance

DerikArnie4760
Level 1
Level 1

I’ve inherited a CUCM with endpoints spread out over approximately 30 locations. Most of our phones are still 7900 series running on SCCP with the vast majority of our phone numbers on 2 SIP trunks. I have an Avaya admin background so I am familiar with theories and methods use in VoIP systems.

 

What I’m looking for some advice on how to make our locations 911 compliant. The situation I have is multiple locations using the same route patterns dumping 911 calls on SIP trunks but giving out the wrong location information due to digit manipulation. I plan on creating new device pools and calling search spaces for each location which will then use the new 911 route patterns using the correct local phone number when dialing 911.

 

What I’m wondering is if there is an easy way to export or organize all the phone, only about 300, in a spreadsheet so I can sort via IP address to identify which phones are in which location. Also is my plan to create separate calling search spaces is for each location sound like a good idea?

 

Sorry if I am missing something easy but I’m am just starting to learn CUCM.

 

Thanks for any assistance!

 

4 Replies 4

Chris Deren
Hall of Fame
Hall of Fame

What is your CUCM version?

Do you need compliance beyond just providing the address, i.e. detailed location such as "13th Floor NW", etc? If so you will need an E911 application such as Cisco Emergency Responder, 3rd party such as RedSky or if you are on CUCM 11+ it has built-in limited e911 capabilities.  

If you only need to provide the address of where the call is made from, you'll need to ensure your SIP provider can send the 911 call to local PSAP based on caller ID presented and then make sure that proper local caller ID is presented when 911 call is made.  Since you state you have centralized (2) SIP trunk updating device pools, CSSes will have no bearing on this, it would if you had decentralized trunks local to each site.

Also, not sure if you are aware of Kari's law which is US Federal law active since February 2020 that among the requirements to allow direct 911 calling (no prefix) you need to provide some kind of onsite alert when emergency call is made, so consider that in your implementation plan as well. 

Thanks for the response Chris!

 

We are on version 12.5, nice to know there is limited E911 native in CUCM but at this time we don't need that.

 

I was thinking of using the CSS/Device Pools as a starting point to get the "main number" sent to the PSAP since I have found many of the DN in the system do not have an External Phone Number Mask set up on the endpoints and we have a lot of non-DID numbers. That way I could cover groups of phones rather than having to drill down to each individual phone right away. Thankfully most of our locations are small and most every phone has an instance of the "main number" DN.

 

My testing has shown without having the digit manipulation on the 911 route pattern Caller ID only gets the 4 digit extension (plus the #1 for the US) and not the correct 10 digit phone number. Is there a way to correct this without hitting each device and adding the correct outgoing number to the DN instance?

 

Kari's law is what triggered me looking into this initially, I haven't gotten to the point of looking at how to set up the alerting yet. I want to first make sure if someone calls 911 they actually get the local PSAP and the correct address is sent, then I will tackle the internal alerting (insert standard sob story of being the understaffed/under funded IT department).

 

Another fun fact I found is that we have locations that a local POTs line is forwarded to a DID that is part of our CUCM. Any suggestions on how to tackle 911 for these?

 

Thanks again


I was thinking of using the CSS/Device Pools as a starting point to get the "main number" sent to the PSAP since I have found many of the DN in the system do not have an External Phone Number Mask set up on the endpoints and we have a lot of non-DID numbers. That way I could cover groups of phones rather than having to drill down to each individual phone right away. Thankfully most of our locations are small and most every phone has an instance of the "main number" DN.


You need to be careful with presenting main number to PSAP as that will prevent callbacks from reaching the actual phone that dialed 911 in case the call disconnects. This is where application such as CER can be helpful as it can route callbacks to non-DIDs, etc.

 

 

My testing has shown without having the digit manipulation on the 911 route pattern Caller ID only gets the 4 digit extension (plus the #1 for the US) and not the correct 10 digit phone number. Is there a way to correct this without hitting each device and adding the correct outgoing number to the DN instance?

 Yes, it all comes down to how your CUCM and GW are configured.  Is the route pattern configured not to use external phone number mask and the "calling party transformation" was set with the 10 digit number?  How about the Route List/Route group, is external mask disabled there too?  

Are there any transformations taking place on the GW?

What does the debug show (debug ccsip messages)?

 


Another fun fact I found is that we have locations that a local POTs line is forwarded to a DID that is part of our CUCM. Any suggestions on how to tackle 911 for these?

 


Not sure I follow this, you have POTS lines (FXO ports) used by CUCM or inbound or outbound or both? Are there 911 patterns pointing to it, if not how is this related to 911?

I will be careful in how I mask extensions to the main number. Like I said most of our locations have 2-5 phones with the main number going to a single DN that is on each phone. So it this situation I should be OK.

 

As for the route pattern's "calling party transformation" is set for 911 (in most cases) with a 10 digit #.

The route list/group part I am not sure (sorry still learning CUCM).

Debug, unknown. Still learning so haven't done that yet but I'll look into it today.

 

The POTs line situation is the previous person "managing" this system apparently could not port some #s over to the SIP trunks so to eliminate the local gateway they or the local telco manually forwarded the number to a DID on the SIP trunk. Does that make sense?