contact center · published

Troubleshooting UCCE and CVP call failures

Verified 2026-10-02 · 58 sources · tier 2 · 1 disputed

Also known as ucce-cvp-call-failures.

Disputed: 1 source below carries conflicting evidence.

Cisco's technote on CVP SendToVRU and Run External Script failures requires a Type 10 network VRU with a label for every CVP PG, and it requires the VRU label length to match the Maximum Length of DNIS configured on CVP 5052. When a routing script cannot find a destination, UCCE uses the default label associated with the dialed number, and if no default label is defined, it returns an error to the routing client 4434.

How the VRU leg is built

In the comprehensive model, CVP needs a Type 10 network VRU with 1 label per Unified CVP Call Server or NIC routing client 27. See also Setting up a CVP comprehensive call flow in UCCE.

For the comprehensive model, CVP's Maximum Length of DNIS is set to the length of the network routing number 29. With a gateway pattern of 1800 followed by 6 digits, that length is 10 29.

CVP routes a returned label either through Local Static Routes in the Dialed Number Pattern configuration or through a SIP proxy server with static routes for the pattern 28.

UCCE selects a call type for a voice call from the dialed number, caller-entered digits (CED) and calling line ID (CLID) 1. When no call type matches, UCCE uses the routing client's default call type 5. If the routing client has no default call type, UCCE uses the general default call type 5.

If the routing client's timeout limit is reached before any on-time routing responses arrive, the NIC tells the routing client to stop sending routing requests 35. See also Troubleshooting UCCE calls that do not route to agents.

Diagnosing SendToVRU and Run External Script failures

WarmTransfer's reading of the sources is that a failed VRU leg can break at 3 separable points (the router finding a VRU label, CVP routing the label plus correlation ID, and the VXML gateway matching a dial-peer with the right application), and each point leaves evidence in a different log 53.

The router cannot find a VRU label

"Customer not found for dialed number ... using default": the dialed number has no customer instance assigned. Cisco's resolution is to define a customer instance for every dialed number rather than rely on the default 12.
"Customer (n) has no valid network vru defined - using default": the customer instance is not associated with a network VRU. The fix is to associate it with a valid network VRU 11.
"No label configured on network VRU ... for routing client ...": the Type 10 network VRU lacks a label for that routing client. The fix is to add a VRU label for each routing client 13.
"No NetworkVRU configured for peripheral for VRU capable call on dialed number": the CVP PG peripheral has no Type 10 network VRU assigned 14.

CVP cannot route the VRU label

A CVP log message that the B2BUA is not configured with a route for the label-plus-correlation-ID number means CVP has no static route and no proxy route that matches the VRU label toward the VXML gateway 10.

The VXML gateway does not run the VRU script

A VRU leg that reaches the VXML gateway but matches a default dial-peer instead of the VRU dial-peer fails to run the VRU script 58. Causes include no VRU dial-peer, a pattern that does not match the DNIS, or a missing signature digit 58.

If the VRU dial-peer has no application bound, the gateway hands the call to the default application 57. The IOS VXML gateway needs the TCL bootstrap application with the cvphost parameter set and header-passing enabled 57.

ECC variable errors

The Call Server log message "VRU APPLICATION ERROR: Assigning the tag 7 to the non-existing ECC variable" means the ECC variable is missing in ICM/NAM or is defined with different lengths 51. The fix is an identical ECC definition 51.

Traces and logs

Router tracing

For SendToVRU failures, Cisco's technote enables router tracing for Route Request, Script Selects, Network VRU Tracing and Call Queuing, and enables CVP Debug41 25.

In UCCE 15.0(1), Router trace levels are 0x00 (default/error), 0x0F (warning), 0xFF (informational) and 0xFFF (debug) 42. Router tracing above default captures Route Requests, Network VRU, Trans Route, VRU Bank, CIC Request, Script Select, Call Queuing, Agent changes and Call Type Real Time 41.

Router tracing can be set with RTRTRACE, RTTEST or the System CLI, and RTRTRACE settings take effect immediately 43. The router process must be running before router tracing can be set or router traces retrieved, whatever tool is used 40.

Reading UCCE logs with dumplog

UCCE process trace logs are stored under [Drive]:\icm\<instance>\<node>\logfiles as .ems files 24. Given only a process prefix such as rtr, dumplog dumps the current day's log for that process unless dates or times are specified 6.

Limit by time: /bd and /ed take dates (mm/dd/yyyy), /bt and /et take times (hh:mm:ss), and /hr takes a number of hours back 8.
Write to a file: /of 8.
Filter lines: /m keeps only lines matching a string, and /x excludes a string 7.
Add milliseconds: /ms 7.
Select a log file: /last selects the most recent log file, and /prev selects the previous one 7.

CVP Call Server and VXML Server logs

CVP Call Server logs are in C:\Cisco\CVP\logs by default, and the file to collect is the one covering the time of the failure 2.

To get SIP message detail, set com.dynamicsoft.DsLibs.DsUALibs to DEBUG on the CVP Call Server diag page (port 8000, /cvp/diag) 3. Restore the previous level after reproducing the failure 3.

Since CVP 8.0(1), VXML Server receives callid (the call GUID), _dnis and _ani by default, and these values appear in the VXML Server log 55. The callid value links Call Server logs to VXML Server logs 55.

VXML Server global call and error logs are written daily under %CVP_HOME%\VXMLServer\logs\GlobalCallLogger and GlobalErrorLogger 56. The VXML Server application error logger records Java exceptions (optionally with stack traces), application timeouts and errors from other loggers in that application 54.

IOS gateway debugs

For VRU-leg failures, Cisco lists these IOS gateway debugs: debug voip ccapi inout, debug ccsip message, debug voip application vxml, debug voip application tcl commands and debug http msg 4.

Script error checking

CVP reports micro-application problems in the user.microapp.error_code ECC variable, which is usually set when control leaves the X (failure) branch 9. Cisco advises testing this variable after every X branch 9.

Value	Meaning
0 to 3	No error, caller hangup, network error and system error, in that order 30
5	Unknown micro-application 33
16	Caller reached the maximum number of invalid tries 31
31	ASR error 32
32	TTS error 32

SIP server group heartbeats

A CVP SIP server group lets the originating endpoint know the destination's status before it sends the INVITE, so failed endpoints are skipped faster 45. CVP server groups are not added automatically and must be configured explicitly 46. At least 1 Call Server must exist before the server group settings can be saved 46. A server group can send heartbeats only to the endpoints defined in it 47.

CVP SIP server group heartbeats use the SIP OPTIONS method as a ping to learn endpoint status without placing a call 23. See also SIP OPTIONS pings and trunk keepalives.

The CVP 12.5(1) configuration guide gives the following heartbeat defaults 192116:

Heartbeat Local Listen Port: 5067 19.
Heartbeat SIP Method: OPTIONS, with PING as the alternative 19.
Heartbeat Transport Type: UDP, with TCP as the alternative 19.
Heartbeat Timeout: 500 ms, with a range of 100 through 3000 ms 21.
Up and Down Endpoint Heartbeat Interval: both default to 5000 ms, with a range of 5000 through 3600000 ms 16.
Overloaded Response Codes: 503, 480 and 600 18.

This is disputed: the CVP 12.5(1) configuration guide gives a default of 1 for Number of failed Heartbeats for unreachable status, with a range of 1 through 5 15.

CVP keeps failed destinations in an unreachable destination table, and the DsUnreachableDestinationTableTimer default is 180 seconds 22. Heartbeat state changes appear in CVP logs as UnreachableDestinationTable add and remove entries and as endpointUnreachableClear events 17.

Failover and survivability

IVR Service failure: in CVP 15.0(1), calls in progress are default-routed to an alternate location on the originating gateway 26.
Ingress gateway cannot reach CVP: survivability TCL scripts on the ingress gateway can reroute calls to another site or dialed number, or play a locally stored .wav file 49.
SIP Service failure after transfer: the call usually continues, but CVP cannot transfer that call again 48.
Router failure: in-progress calls that have already reached a VRU or an agent are not affected 39.

Ring-no-answer and requery

CVP's RNA timeout is set per pattern in OAMP under System > Dialed Number Pattern 38. The RNA timeout must be longer than the Agent Desk Settings RONA timer 38.

When the RNA timeout expires, CVP sends the router an EventReport of No Answer 37. That EventReport triggers a requery if target requery is enabled on the Queue to Skill Group node 37. In a requery script, the RequeryStatus variable is tested, and a value greater than zero indicates the call is a requery 36.

Applicability

Applies to: Cisco Unified CVP and Cisco Unified Contact Center Enterprise. Deployments: on-premises. Sources checked 2026-10-02. The SendToVRU and Run External Script material comes from a Cisco technote 25. The VXML gateway guidance in that technote applies to the IOS VXML gateway 57.

We infer the 3-point breakdown of VRU-leg failures from that technote, and its applicability to current releases is assumed 53.

The heartbeat defaults are from the Unified CVP 12.5(1) configuration guide 21.

Server group configuration and IVR Service failover are documented for Unified CVP 15.0(1) 4626. Routing, call type and Router trace behaviour are documented for UCCE 15.0(1) 142.

The dumplog source does not pin a release 6.

What remains uncertain

How VRU-leg failures present when Cisco VVB is used instead of an IOS VXML gateway is not covered by the sources below. See also Cisco VVB.

A CVP 15.0(1) table of SIP server group heartbeat defaults is not covered by the sources below. The defaults listed above come from the 12.5(1) guide 21.

Whether any release uses a Heartbeat Timeout default other than the 500 ms in the 12.5(1) guide is not covered by the sources below.

Sources disagree on the default for Number of failed Heartbeats for unreachable status, which the 12.5(1) guide gives as 1 15. How that setting relates to a separate heartbeat retry count is not covered by the sources below.

Which SIP failure responses make CVP requery the router or invoke survivability is not covered by the sources below.

Dialed Number Pattern behaviour for VRU labels beyond Local Static Routes, including any send-to-originator setting, is not covered by the sources below.

The meanings of user.microapp.error_code values other than 0 to 3, 5, 16, 31 and 32 are not covered by the sources below.

CVP 15.0 Call Server log file names and any changes to the diag page are not covered by the sources below.