Cisco Gatekeeper and ISDN interview questions and answers

Cisco Gatekeeper and ISDN interview questions – Welcome back. In our first article, we covered clustering, media resources, and basic dial plans. Now, we go deeper into the topics that separate a junior engineer from a senior architect – H.323 gatekeepers, ISDN signaling, numbering plans, and class of restriction (COR).

If you are serious about passing a CCIE Voice lab or a high-level Cisco IPT interview, these Cisco Gatekeeper and ISDN interview questions are essential. Let’s dive in.

Table of Contents

Part 1: H.323 Gatekeeper Functions (Q 51–60)

51. What is an H.323 Gatekeeper?

An H.323 gatekeeper is an optional but powerful entity that provides address translation (E.164 to IP), admission control, bandwidth management, and zone management for H.323 endpoints (terminals, gateways, MCUs).

52. What are the mandatory gatekeeper functions?

Per the H.323 standard, four functions are mandatory:

Address Translation – Translates H.323 IDs or E.164 numbers to IP addresses.
Admission Control – Uses RAS messages (ARQ, ACF, ARJ) to allow or deny endpoint admission.
Bandwidth Control – Manages bandwidth via BRQ, BCF, BRJ.
Zone Management – Controls registration of endpoints within its zone.

53. What are optional gatekeeper functions?

Call Authorization – Time-of-day or source-based restrictions.
Call Management – Maintains active call info to indicate busy endpoints.
Bandwidth Management – Rejects admission if bandwidth unavailable.
Call Control Signaling – Gatekeeper-routed call signaling (GKRCS). Note: Cisco IOS gatekeepers do NOT support GKRCS – they use direct endpoint signaling.

54. What is a Technology Prefix?

A technology prefix is a special prefix (e.g., 1#) that gateways register with the gatekeeper. It identifies the type of gateway or the technology it supports (e.g., H.323, voice, fax). The gatekeeper uses it to select the correct destination gateway.

55. Technology Prefix Matches vs. Default Technology Prefixes – what’s the difference?

Technology Prefix Matches: The gatekeeper uses the tech prefix appended in the called number to select the destination gateway or zone.
Default Technology Prefixes: The gatekeeper assigns a default gateway(s) for routing unresolved addresses based on the gateway’s registered tech prefix.

56. What are VIA zones (Voice Infrastructure and Application)?

VIA zones are software enhancements on Cisco gatekeepers that allow IP-to-IP gateway functionality. The gatekeeper recognizes two call legs on the same platform and load-balances across multiple IP-to-IP gateways within a predefined VIA zone. It acts as a VoIP transfer point (transit zone) for ITSP networks.

57. Scenario: How does a gatekeeper select an IP-to-IP gateway within a VIA zone?

Algorithm:

If a tech prefix is found (alias-based matching), look through gateways in the viazone that registered that tech prefix.
If no tech prefix, look through all gateways in the viazone.
Select the first IP-IP GW found with available resources.
If all are out of resources, select the first one anyway.
If none found, return failure (ARJ message).

58. What RAS messages are used for admission control?

ARQ – Admission Request (endpoint asks permission to call)
ACF – Admission Confirm (gatekeeper says yes, here is the destination IP)
ARJ – Admission Reject (gatekeeper says no)

59. What RAS messages are used for bandwidth control?

BRQ – Bandwidth Request
BCF – Bandwidth Confirm
BRJ – Bandwidth Reject

60. Scenario: A gateway registers with the gatekeeper, but calls fail with “ARJ” in debugs.

Answer: Use debug ras. If you see ARJ, check if the gatekeeper has the destination zone prefix configured (show gatekeeper zone prefix). Also verify that the calling gateway has enough bandwidth allocated in the zone.

Part 2: ISDN Q.921, Q.931, and Call Flows (Q 61–70)

61. What is ITU Q.921 (LAPD)?

Q.921 is the ISDN Data Link Layer (Layer 2) specification. Also called LAPD (Link Access Procedure on the D channel). It is based on HDLC and provides reliable transport for Layer 3 (Q.931) signaling messages over the D channel.

62. What is ITU Q.931?

Q.931 is the ISDN Network Layer (Layer 3) specification. It is responsible for call control, setup, teardown, and supplementary services. It uses a message-based system on the D channel (common channel signaling – CCS).

63. Describe the basic ISDN call setup flow (en-bloc).

Calling party sends Setup message (contains called number, bearer capability, etc.).
Local ISDN switch responds with Call Proceeding.
Remote switch sends Setup to called party.
Called party responds with Call Proceeding and then Alerting.
Alerting travels back to calling party (ringing).
Called party answers → Connect message travels back.
Calling party responds with Connect Acknowledge.
Voice path established.

64. What is the difference between en-bloc and overlap signaling?

En-bloc: All digits are sent together in the Setup message (common for PRI, DID).
Overlap: Digits are sent one by one; a Setup Acknowledge is used between digits. The switch waits for more digits until a timer expires.

65. Describe the ISDN call disconnect procedure (calling party hangs up first).

Calling party sends Disconnect.
Local switch releases B channels and sends Release to calling party.
Calling party responds with Release Complete.
Remote switch sends Disconnect to called party.
Called party sends Release back, receives Release Complete.
Call cleared on both sides.

66. What are Information Elements (IEs) in Q.931?

IEs are fields within a Q.931 message that carry specific data. A single Setup message can contain 15+ IEs, including:

Called party number
Calling party number
Bearer capability (e.g., voice 64k)
Transit network selection
Date and time
Low-layer compatibility

67. What is the difference between single octet and multiple octet IEs?

Single octet: One byte, used for simple flags (e.g., shift, repeat indicator).
Multiple octet: Variable length, includes a length field and actual data (e.g., called party number).

68. Scenario: `show isdn status` shows Layer 1 ACTIVE but Layer 2 “TEI_ASSIGNED” – what does this mean?

Answer: TEI_ASSIGNED means Layer 2 is not fully up. The router has a TEI (Terminal Endpoint Identifier) but is not exchanging SABME/RR frames. Common causes: wrong isdn switch-type, D-channel down, or telco side issue. Desired state is MULTIPLE_FRAME_ESTABLISHED.

69. What do SABME messages indicate in `debug isdn q921`?

SABME (Set Asynchronous Balanced Mode Extended) messages indicate Layer 2 is trying to reinitialize. If they appear repeatedly, the link is unstable – check clocking, cabling, or switch-type mismatch.

70. What is the purpose of a mid-span repeater on an ISDN BRI loop?

It regenerates the digital signal. BRI has a maximum loop length of ~18,000 feet. Beyond that, a repeater “dusts off” and retransmits the signal for another 18,000 feet. Also deployed if dB loss is between –38 and –42 dB.

Check – IP Address Information Tool

Part 3: Numbering Plans – NANP, ETNS, Integration (Q 71–80)

71. What is the North American Numbering Plan (NANP)?

NANP is an integrated telephone numbering plan serving 19 North American countries (US, Canada, Caribbean, etc.). Numbers are 10 digits: NXX-NXX-XXXX, where N = 2–9, X = 0–9. First three digits = area code (NPA), next three = central office code, last four = station number.

72. What are N11 codes? Give examples.

N11 codes are three-digit service codes not used as area codes. Examples:

211 – Community information (US)
311 – Non-emergency government services
411 – Local directory assistance
511 – Traffic info
611 – Repair service
711 – Telecommunications relay services
811 – Business office
911 – Emergency

73. What are ERC (Easily Recognizable Codes)?

ERCs are area codes where the second and third digits are the same (e.g., 800, 888, 877, 866). They are reserved for toll-free services.

74. What are ANI II digits?

ANI II digits are two-digit pairs sent with the originating telephone number during call setup. They identify the type of originating station (e.g., payphone, prison, regular subscriber).

75. What is the European Telephony Numbering Space (ETNS)?

ETNS is a pan-European numbering space parallel to national plans, using country code 388. It allows European international companies to get a single European number. Four services exist: Public Service Application, Customer Service, Corporate Networks, and Personal Numbering. Each has a European Service Identification (ESI) code after 3883.

76. Scenario: Private vs. Public numbering plan integration – what are the challenges?

Challenges include:

Varying number lengths (local, long-distance, international)
Need to strip or add area codes/prefixes
Auto-attendant for non-DID numbers
PSTN backup for WAN failure (must manipulate numbers dynamically)

77. What are the three basic functions of private/public numbering plan integration?

Reachability to external PSTN destinations via gateway.
Auto-attendant (IVR) when DID range doesn’t cover all extensions.
PSTN backup path when IP WAN fails or is congested.

78. Scenario: Overlapping numbering plans – how do you solve it?

Answer: Implement site-code dialing. Example:

Intersite prefix: 8
Site A code: 81
Site B code: 82
User at Site A dials 8-82-2001 to reach extension 2001 at Site B.
Internal extensions start with 1, 2, 3, etc., never with 8 to avoid ambiguity.

79. What is the disadvantage of non-overlapping numbering plans?

They require centralized numbering assignment and careful design from the beginning. Very difficult to implement after growth. Overlapping with site-code dialing is more scalable and decentralized.

80. Scenario: A company has DID range 200-555-3xxx but internal extensions are 1xxx. How do you map them?

Answer: On the voice gateway, configure translation patterns or digit manipulation. Incoming call: called number 200-555-3001 → strip prefix → extension 1001. Outgoing call: extension 1001 → prefix 200-555-3 → send to PSTN. Avoid complex mapping (e.g., 1xxx + 50) – keep it one-to-one.

Part 4: Dial Peer Matching, DID vs. Non-DID, COR (Q 81–90)

81. What is the inbound dial peer matching sequence on a Cisco IOS gateway?

Called number (DNIS) – match incoming called-number.
Calling number (ANI) – match answer-address.
Calling number (ANI) – match destination-pattern (rare, but possible).
Voice-port – match port command (POTS only).
Default dial peer 0 – if nothing matches.

82. Scenario: Why does my incoming PRI call always match dial peer 0?

Answer: You did not configure incoming called-number . on any POTS dial peer. Without it, the router falls back to default peer 0, which often treats the call as a modem call and drops it. Fix: add a POTS dial-peer with incoming called-number . and direct-inward-dial.

83. What is the difference between DID and non-DID (two-stage dialing)?

DID (Direct Inward Dial): The Setup message contains all digits. The router does NOT collect more digits. Outbound matching uses the entire dial string. Use direct-inward-dial on the inbound POTS peer.
Non-DID (two-stage): The router collects digits one by one (inband). Outbound matching is digit-by-digit. Call routes as soon as a partial match is found.

84. Scenario: You have dial peers with destination patterns `816` and `81690`. A call comes in with digits `81690` in non-DID mode. Which peer matches?

Answer: Because it is non-DID, the router matches after receiving 816 (first peer) and routes the call immediately, only sending 816 out. That’s wrong. To fix, either use DID on the inbound peer or add a wildcard (e.g., 816. with a dot) to force longest-match.

85. What is Class of Restriction (COR)?

COR is a Cisco IOS feature that provides Class of Service (calling privileges) on voice gateways, commonly used with SRST and CME. It uses a lock and key mechanism:

Outgoing COR list = lock (on outbound dial peer)
Incoming COR list = key (on inbound dial peer or ephone-dn)
Call succeeds only if the key list is a superset or equal to the lock list.

86. Scenario: Give a COR configuration example.

Requirement: Only managers can dial 1900 numbers.
Configuration:

dial-peer cor custom
 name local
 name longdistance
 name premium
!
dial-peer cor list calllocal
 member local
!
dial-peer cor list callpremium
 member premium
!
dial-peer voice 10 pots
 destination-pattern 1900...
 corlist outgoing callpremium
 port 1/0/0
!
ephone-dn 1
 number 2001
 corlist incoming callpremium   ! Manager phone has premium key
!
ephone-dn 2
 number 2002
 corlist incoming calllocal     ! Employee only has local key

87. What happens if either incoming or outgoing dial peer has no COR list?

The call succeeds. COR only blocks when both sides have lists and the incoming list does NOT contain all members of the outgoing list.

88. Scenario: A user can dial local and long distance but not international. How to implement?

Create three COR lists: local, national, international. Outbound dial-peer for international has corlist outgoing international. The user’s incoming COR list includes only local and national – missing international → call blocked.

89. What is the default behavior if no COR is configured on an ephone-dn?

By default, an incoming call leg has the highest priority – it can go out any dial peer regardless of the outgoing COR list. So if you want restrictions, you must explicitly assign an incoming COR list to the ephone-dn or inbound dial peer.

90. What is the difference between `corlist incoming` and `corlist outgoing`?

corlist incoming – Assigned to the inbound leg (the key). Placed on ephone-dn or inbound POTS dial peer.
corlist outgoing – Assigned to the outbound dial peer (the lock).

Part 5: Translation Patterns, Transformation Masks, Non-Urgent (Q 91–100)

91. What is a Translation Pattern in CUCM?

A translation pattern matches dialed digits like a route pattern but does not route the call to a gateway. Instead, it manipulates the digits (calling or called party) and then reroutes the call for a second digit analysis. It is used for mapping PSTN DID numbers to internal extensions.

92. What is the difference between a Translation Pattern and a Transformation Mask?

Translation Pattern: A configuration object in CUCM that matches digits, changes them, and reroutes.
Transformation Mask: A field inside a route pattern, translation pattern, or gateway configuration that modifies either the called party number or calling party number (Caller ID).

93. What is a Calling Party Transformation Mask?

It manipulates the Caller ID (ANI) for outgoing calls. Example: A user with extension 1001 calls externally. The mask 2005551XXX changes the outgoing Caller ID to 2005551001.

94. What is a Called Party Transformation Mask?

It manipulates the dialed digits (DNIS). Example: Incoming PSTN call to 4085551234 – mask 4XXX changes it to internal extension 1234.

95. Scenario: DID range 408-555-1XXX maps to internal extensions 4XXX. Configure translation pattern.

CUCM Configuration:

Translation Pattern: 4085551XXX
Called Party Transformation Mask: 4XXX
Partition: PSTN_Inbound
CSS: Internal_Phones
When a call arrives to 4085551234, CUCM matches the pattern, strips the prefix, and reroutes to extension 4234.

96. What is a Non-Urgent Translation Pattern? Why use it?

In CUCM 7.0 and later, translation patterns can be non-urgent (uncheck “Urgent Priority”). Without this, the pattern matches as soon as possible and routes immediately, breaking variable-length dialing (e.g., international 9011….). Non-urgent allows the system to wait for more digits.

97. Scenario: International dialing fails with `9011.` pattern. Why?

Answer: The pattern 9011. was marked Urgent (default in older CUCM). As soon as the user dials 9 then 0 then 1 then 1, the pattern matches and CUCM routes immediately without waiting for the country code and number. Make it Non-Urgent to fix.

98. What is the maximum number of translation pattern iterations?

CUCM allows up to 10 iterations of digit analysis through translation patterns. This prevents call routing loops.

99. Can a translation pattern block a call? How?

Yes. Select the Block This Pattern radio button and choose a cause code:

Unallocated Number
Call Rejected
Number Changed
Invalid Number Format
Precedence Level Exceeded

100. Scenario: An old extension 4333 no longer exists. How to send callers to the operator (4111)?

Create a translation pattern XXXX (where XXXX matches any 4-digit extension not otherwise routed). Apply Called Party Transformation Mask 4111 and set to non-urgent. When CUCM cannot find DN 4333, it matches the generic pattern and sends the call to the operator.

Conclusion

These 50 Cisco Gatekeeper and ISDN interview questions cover the advanced topics that many engineers neglect – gatekeeper RAS messaging, ISDN Layer 2/3 troubleshooting, numbering plan integration, and Class of Restriction. Master these, and you will confidently handle CCIE-level scenarios.

Also Check – Cisco IPT Media, SRST, DSP farm and transcoder configuration Interview QnA