Cisco IPT Advanced: SIP Trunks, DTMF Relay, MVA, and Call Park – 50 Expert Q&As

Welcome to Article 4 of our Cisco IPT interview series. We have covered clustering, gatekeepers, ISDN, media resources, and SRST. Now, we tackle the protocols and features that dominate modern VoIP deployments – SIP trunks, DTMF relay, mobility features (MVA/SNR), and call park.

These SIP trunk DTMF relay and CUCM mobility features questions will prepare you for real-world enterprise scenarios involving SIP providers, remote workers, and advanced call handling.

Let’s dive in.

Part 1: SIP Methods and Message Responses (Q 151–160)

151. What are the six core SIP methods (requests)?

1. INVITE – Initiates a call session.

2. ACK – Confirms final response to INVITE.

3. BYE – Terminates an existing session.

4. CANCEL – Cancels a pending INVITE (before final response).

5. REGISTER – Registers a user agent’s location with a SIP server.

6. OPTIONS – Queries server/capabilities (no session established).

152. What is the purpose of the ACK method?

ACK confirms that the final response (e.g., 200 OK) to an INVITE was received. If the calling party includes a session description in the ACK, that description overrides the one in the INVITE.

153. What is the difference between BYE and CANCEL?

• BYE – Ends an established session (after 200 OK is sent).

• CANCEL – Cancels a pending INVITE (before final response). CANCEL does not affect completed requests.

154. What are the six classes of SIP responses?

155. What is a 100 Trying response?

A provisional response sent by the next-hop server to indicate that the INVITE was received and processing is in progress. It stops retransmissions of the INVITE.

156. What is a 180 Ringing response?

Sent by the called party’s user agent to indicate that the phone is ringing.

157. What is a 183 Session Progress response?

Used to send early media (e.g., ringback tones, announcements) before the call is answered. Can contain SDP for media establishment.

158. What does a 404 Not Found response indicate?

The requested user or domain does not exist at the server. Usually means a dial plan mismatch or unregistered endpoint.

159. What does a 486 Busy Here response mean?

The called party is currently busy and cannot accept the call. The endpoint is reachable but not available.

160. Scenario: You receive a 503 Service Unavailable from a SIP trunk provider. What does it mean?

Answer: The provider’s server is overloaded or undergoing maintenance. The call should be retried later or routed via an alternate trunk (PSTN backup).

Part 2: SIP UA and Voice Service Configuration (Q 161–170)

161. What is SIP UA (User Agent) configuration mode?

Entered via sip-ua command on a Cisco IOS gateway. It controls SIP client behavior including registration, timers, NAT traversal, and retry counts.

162. What does the registrar command do in SIP UA?

It allows a gateway to register E.164 numbers of non-SIP phones (analog FXS, etc.) with a SIP registrar or proxy server. Example:

registrar ipv4:10.30.25.250 tcp expires 3600

Secondary registrar can be configured for redundancy.

163. What does the sip-server command do?

Specifies the name or IP address of a SIP server (usually a proxy). When configured, dial peers can use session target sip-server instead of explicit IP addresses.

164. What does the max-forwards command do?

Limits the maximum number of hops a SIP request can traverse (default 70). Each proxy decrements the value. When it reaches 0, the request is discarded – prevents infinite loops.

165. What SIP UA timers can be adjusted?

• timers connect – Wait for 200 OK to ACK (default 500ms)

• timers disconnect – Wait for 200 OK to BYE (default 500ms)

• timers expires – Valid time for an INVITE (default 180 seconds)

• timers trying – Wait for 100 Trying response (default 500ms)

166. What is redirect ip2ip in voice service VoIP?

It allows SIP calls to be “hairpinned” – received on one VoIP dial peer and sent out another VoIP dial peer. Essential for gateway-to-gateway toll bypass.

167. What does bind control source-interface do?

Sets the source IP address for SIP control signaling (not media). Example:

bind control source-interface Loopback0

This ensures consistent source IP for SIP messages, critical for NAT and firewall traversal.

168. What does bind media source-interface do?

Sets the source IP address for RTP media streams. If not configured, the gateway uses the outgoing interface IP.

169. What is the difference between session transport udp and tcp for SIP?

• UDP – Default. Lightweight but unreliable; no message fragmentation.

• TCP – Reliable; used when SIP messages exceed MTU (fragmentation avoidance). Can be globally set with session transport tcp or dynamically with transport switch udp tcp.

170. Scenario: Your SIP gateway sends large INVITE messages that get fragmented. How do you fix?

Answer: Configure transport switch udp tcp under voice service VoIP → SIP. This forces the gateway to switch from UDP to TCP when a SIP message exceeds the MTU (usually 1500 bytes).

Part 3: DTMF Relay Methods (Q 171–180)

171. Why is DTMF relay necessary?

Low-bandwidth codecs (G.729, G.723) compress voice and can distort or completely lose DTMF tones (touch tones). DTMF relay sends tone information out-of-band or using special packets.

172. What is RTP-NTE (RFC 2833) DTMF relay?

In-band method where DTMF tones are sent as special RTP Named Telephony Event packets instead of encoded voice. SDP negotiates the payload type. Most common and interoperable method.

173. What is SIP-NOTIFY DTMF relay?

Out-of-band method. DTMF digits are sent in SIP NOTIFY messages. Negotiated via Call-Info header in INVITE/200 OK. Preferred for SCCP to SIP interworking.

174. What is SIP-INFO DTMF relay?

Out-of-band method using SIP INFO messages. Always enabled on Cisco gateways. Good for interoperability but less common than NOTIFY.

175. What is KPML (KPML) DTMF relay?

Keg Pad Markup Language – XML-based method where the phone sends digit information in SUBSCRIBE/NOTIFY messages. More efficient for long digit strings (e.g., IVR menu traversal).

176. How do you configure DTMF relay on a SIP dial peer?

dial-peer voice 100 voip
 dtmf-relay sip-notify rtp-nte

This tells the gateway to prefer SIP-NOTIFY, but fall back to RTP-NTE if the remote side doesn’t support NOTIFY.

177. Scenario: SIP to SCCP calls have DTMF failures (IVR doesn’t recognize digits). Why?

Answer: SCCP uses out-of-band DTMF only. SIP trunk between CUCM and the gateway may be using in-band (RTP-NTE) by default. In CUCM 4.x, you must enable an MTP on the SIP trunk to translate between in-band and out-of-band. CUCM 5.x+ can negotiate without MTP if both sides support out-of-band.

178. What is the notify telephone-event max-duration command?

Sets the maximum duration (in milliseconds) between NOTIFY messages for a single DTMF event. Default 2000ms (2 seconds). The lower value between two SIP peers is used.

179. Scenario: DTMF digits are recognized but repeated multiple times. How to fix?

Answer: The DTMF relay method may be sending duplicate events. Check dtmf-relay configuration. If using RTP-NTE, ensure the gateway is not also sending in-band tones (disable dtmf-relay cisco-rtp). Adjust timers if using NOTIFY.

180. What is the default DTMF method if no dtmf-relay is configured?

In-band – DTMF tones are sent as normal voice audio through the codec. This fails with G.729 and is unreliable with G.711 in noisy environments.

Part 4: SIP Trunk Configuration – CUCM 4.x vs 5.x (Q 181–185)

181. How do you configure a SIP trunk in CUCM 4.x?

1. Device → Trunk → Add a New Trunk → SIP Trunk.

2. Device Protocol automatically set to SIP.

3. Enter destination address (IP of remote gateway/proxy).

4. MTP Required – Must be checked (to translate DTMF between SCCP and SIP).

5. Destination Port – default 5060.

6. Incoming Port – Must be unique for each signaling interface in CUCM 4.x.

182. How does SIP trunk configuration differ in CUCM 5.x+?

• MTP Required – Not automatically checked; may not be needed if endpoints negotiate out-of-band DTMF.

• Incoming Port – Multiple trunks can share the same incoming port (5060).

• New fields: Presence Group, SIP Trunk Security Profile, SIP Profile, SUBSCRIBE Calling Search Space.

183. Why did CUCM 4.x require an MTP on SIP trunks?

SCCP phones use out-of-band DTMF. SIP trunks default to in-band DTMF (RTP-NTE). The MTP translated between the two. CUCM 5.x+ improved SIP stack to negotiate out-of-band without MTP.

184. Scenario: Your SIP trunk works for voice but fails for fax (G.711 pass-through). What is wrong?

Answer: SIP trunks often need dtmf-relay configured on the gateway side. For fax, ensure fax protocol pass-through g711ulaw and disable VAD (no vad). Also verify that the SIP trunk is not forcing an MTP that breaks T.38 or pass-through.

185. What is toll bypass using SIP?

Toll bypass sends calls between two sites over the IP network (SIP between routers) instead of the PSTN, avoiding long-distance charges. Configure VoIP dial peers on both routers pointing to each other’s IP addresses with session protocol sipv2.

Check – AWS Policy Generator Tool

Part 5: Mobile Voice Access (MVA) and SNR (Q 186–195)

186. What is Mobile Voice Access (MVA)?

MVA is a CUCM feature that allows a PSTN user (e.g., mobile phone) to dial into the office, authenticate with a PIN, and receive an internal dial tone. The mobile phone then acts like an internal IP phone – can dial extensions, transfer calls, place calls that appear from the desk phone.

187. How does MVA hairpinning work?

The PSTN call comes into the voice gateway, is sent to CUCM, authenticated via VXML, and then CUCM sends a new call leg back out the same gateway to the final destination. The carrier may charge for both legs (hairpinning fee). Negotiate this with your telco.

188. What is the VXML URL used for MVA?

CUCM serves an IVR script (VXML) that prompts the user for PIN. Example configuration on the gateway:

application
 service frog http://142.2.64.11:8080/ccmivr/pages/IVRMainpage.vxml

189. How do you configure a dial peer for MVA on an H.323 gateway?

dial-peer voice 20 voip
 service frog
 session target ipv4:142.2.64.11
 incoming called-number 99992222   ! MVA DID number
 codec g711ulaw
 no vad

The service frog command links the dial peer to the VXML application.

190. What is Single Number Reach (SNR) / Mobile Connect?

SNR allows incoming calls to a desk phone to simultaneously ring up to four remote destinations (mobile phone, home phone). The user can answer from any device. Call appears to come from the desk phone (Caller ID preservation).

191. What is the difference between MVA and SNR?

192. What is the BHCA formula for SNR remote destinations?

Each remote destination adds 0.5 call legs per call. Formula:

Remote Destination BHCA = 0.5 × (Number of users) × (User BHCA)

Example: 300 users, each 5 BHCA, each with 1 remote destination:

• Remote BHCA = 0.5 × (300 × 5) = 750

• Total system BHCA = (300 × 5) + 750 = 2250

193. Scenario: A user reports that MVA authentication fails (PIN not recognized).

Answer: Check:

• User has a PIN configured in CUCM (not just a password).

• User has “Enable Mobile Voice Access” checked in Mobility settings.

• The MVA DID number in clusterwide parameters matches the inbound dial peer.

• CSS on the MVA route pattern allows access to internal partitions.

194. What DTMF digits are available during an MVA call for call control?

195. Scenario: MVA call connects but there is no dial tone after PIN entry.

Answer: Check:

• The route pattern for MVA DID has “Provide Outside Dial Tone” checked.

• The gateway has voice service voip with allow-connections h323 to h323 (for H.323 hairpinning).

• The CSS after authentication includes access to route patterns.

Part 6: Call Park, Meet-Me Conference, Custom Ringtones (Q 196–200)

196. What is Call Park?

A feature that allows a user to “park” an active call into a virtual slot (extension). The phone displays the slot number. Any other phone can dial that slot number to retrieve the call.

197. What is Directed Call Park? How is it different?

Retrieval prefix is configured separately (e.g., *51). User parks call to slot “01”. Retrieving user dials *5101.

198. What is the difference between Ad-Hoc and Meet-Me Conference?

199. How do you add a custom ringtone to a Cisco IP Phone?

Step 1: Convert MP3/WAV to CCITT u-Law, 8,000 kHz, 8-bit, no header (use Sound Recorder or SOX).
Step 2: Save as .raw file (max 1.4 MB).
Step 3: Copy to CUCM TFTP server path (C:\Program Files\Cisco\TFTPPath).
Step 4: Edit Ringlist.xml (add <DisplayName> and <FileName> entries).
Step 5: Save and restart TFTP service. Phones download new ringlist on next reboot.

200. Final Scenario: Call Park works but Directed Call Park fails. What do you check?

Answer: Verify:

• Directed Call Park slots are configured (Call Routing → Directed Call Park).

• Retrieval prefix is set (e.g., *51) and does not overlap with other patterns.

• The user’s phone CSS includes the partition containing the Directed Call Park slots.

• The user dials the retrieval prefix before the slot number.

Part 7: Advanced Topics – SAF, WNIC2, card type, ISDN troubleshooting (Q 201–210)

201. What is SAF (Service Advertisement Framework)?

SAF is a network protocol modeled after EIGRP (Cisco proprietary). It enables sharing of call routing data between call agents (CUCM, CUCME, CUBE, SRST, IOS Voice Gateways). It advertises and requests routing data to/from the SAF network.

202. What is WNIC2?

WNIC2 is a voice interface card (port adapter) that can be configured for E1 or T1 connectivity. There is no default card type – you must specify using the card type command. Without it, the port adapter is non-functional and does not appear in show commands.

203. How do you set the card type for E1?

card type e1 0

204. Scenario: show isdn status shows Layer 2 as TEI_ASSIGNED (not MULTIPLE_FRAME_ESTABLISHED). How do you fix?

Step 1: Verify isdn switch-type matches telco (e.g., primary-5ess, primary-ni).
Step 2: Verify pri-group timeslots is correct.
Step 3: Ensure D-channel interface (serial x:23) is not shutdown.
Step 4: Check cabling and clocking.
Step 5: Contact telco to verify their side is up.

205. What do SABME messages in debug isdn q921 indicate?

SABME (Set Asynchronous Balanced Mode Extended) indicates Layer 2 is trying to reinitialize. If seen repeatedly, the link is unstable. Causes:

• Wrong switch-type

• Clocking mismatch

• Physical layer issues (cabling, distance)

206. What is the desired state for ISDN Layer 2?

MULTIPLE_FRAME_ESTABLISHED – Means Layer 2 frames are being exchanged and initialization is complete.

207. What are common causes of clocking problems on T1/E1 PRI?

• Incorrect DSU/CSU configuration

• Cables longer than 50 feet (unshielded)

• Poor patch panel connections

• Multiple cables connected in series

• Loopback left enabled on the interface

208. How do you isolate a clocking problem?

1. Perform local loopback test (router looped to itself – errors stop? Problem is external).

2. Perform remote loopback test (telco loops the line – errors continue? Problem is telco or CSU/DSU).

3. Use show controllers t1 to check for alarms (LOS, LOF, RAI, AIS).

209. What is the debug isdn q921 command used for?

Displays data link layer (Layer 2) access procedures on the D-channel. Shows SABME, RR (Receiver Ready), RNR (Receiver Not Ready) messages. Essential for troubleshooting PRI/BRI bring-up.

210. Final advanced scenario: T1 PRI Layer 1 is ACTIVE, Layer 2 is MULTIPLE_FRAME_ESTABLISHED, but calls fail with “Unallocated Number”. What is wrong?

Answer: Layer 1 and 2 are fine. Layer 3 is the issue. Possible causes:

• The called number is not in the allowed range (telco side).

• The gateway is not sending the correct number of digits.

• The isdn incoming-called-number or dial peer matching is stripping digits incorrectly.

• The telco has not activated the DID range for that PRI.
  Use debug isdn q931 to see the exact Setup message and called number being sent.

Conclusion

These 60 questions (151–210) on SIP trunk DTMF relay and CUCM mobility features cover SIP signaling, DTMF relay methods, MVA/SNR configuration, call park types, and advanced ISDN troubleshooting. You now have the knowledge to configure SIP trunks, deploy mobility features, and diagnose Layer 2/3 ISDN issues.

Also Check – Top 50 Amazon Connect Interview Questions and Answers for 2026