Webex Calling call quality troubleshooting in Control Hub

Verified 2026-09-05 · 62 claims · sources tier 2 · 5 disputed

Also known as Advanced Troubleshooting, PNC, call segment, hop, media session, troubleshooting view.

Control Hub provides visibility into Webex Calling media quality by evaluating endpoint telemetry and cloud media relay metrics across individual call hops 22​26. Cisco states that the troubleshooting view helps identify the problem area by providing relevant metrics and cannot necessarily provide the root cause of a poor call 58.

Troubleshooting Access and Retention

Opening the Webex Calling troubleshooting view requires the full administrator, read-only administrator, or support administrator role in Control Hub 56. WarmTransfer's reading of the sources is that no source read publishes a licence, subscription, or entitlement requirement for Advanced Troubleshooting, which is described only as a role a full administrator either has or lacks 39. The default troubleshooting view shows the last seven days, and the searchable window extends to twenty-one days 54, exposing calls from the past twenty-one days 2. The only statement in the sources tying data retention to licence level says Pro Pack customers generally have more historical data available, and it appears in a discussion of Analytics dashboards rather than of the troubleshooting view 47.

Webex Calling calls appear in the troubleshooting view only after the call ends, whereas meetings can be viewed while in progress 14. Sources disagree on how phone number search handles partial inputs: the troubleshooting article's search view section states that entering the first two digits of a phone number lists possible numbers and that entering digits then pressing enter attempts an exact match 52, while another section of the same article states that entering the first three to eight digits shows matching entries and that more than eight digits attempts an exact match 51. In contrast, the advanced diagnostics article describes phone number search as an exact string match and does not mention partial digit entry 37.

Exports from the troubleshooting view are capped at one hundred troubleshooting call records per comma-separated values (CSV) file 19. WarmTransfer's reading of the sources is that no application programming interface or programmatic export is stated for troubleshooting or media quality data, leaving per-call file downloads capped at one hundred records as the only documented mechanisms 38.

Media Quality Thresholds and Call Grading

A Webex Calling call is graded good when the endpoint reports packet loss below 5 percent, latency or round-trip time below 400 milliseconds, and jitter below 150 milliseconds 22. The Control Hub call quality drop alert defines good as both video and audio streams below 150 milliseconds jitter, 400 milliseconds latency, and 5 percent packet loss, and poor as both streams above all three 5. At the relay level, a hop is graded good when the media relay point reports packet loss below 2.5 percent, latency or round-trip time below 200 milliseconds, and jitter below 75 milliseconds 26. Cisco's November 2025 troubleshooting technical paper publishes the same hop quality thresholds of 2.5 percent packet loss, 200 milliseconds latency, and 75 milliseconds jitter as the September 2026 help article 43.

A call in the troubleshooting view carries one of three statuses: successful, refusal, or failure 13. Overall call quality takes one of three values: poor when either party had a poor end-to-end experience, good when neither exceeded the thresholds, and none or not available when no media quality data exists 41. A good quality call is one where every media session within the call was categorised as good quality 25. Quality grading applies to Webex Calling calls only and does not apply to Webex meetings or Call on Webex sessions in the same search interface 49.

Sources disagree regarding meeting quality evaluations across documents: Cisco's technical paper publishes a three-tier scheme with good below 300 milliseconds latency and 3 percent packet loss, whereas the advanced diagnostics article publishes a two-tier live map indicator where meeting quality is good at or below 5 percent packet loss and 400 milliseconds latency, with no jitter component 32​31. General metrics interpretation and estimation concepts from other environments are not covered by any claim in this article.

Call Hops, Flows, and Topology Analysis

A correlation identifier ties together multiple call segments or call legs belonging to the same call 18. A call segment represents one direct hop or connection between two participants and is the unit the troubleshooting view's hop model is built on 12. A call segment is typically composed of two call detail records, an originating and a terminating leg, but a public switched telephone network (PSTN) call may generate only the terminating leg from the Webex Calling perspective 53. Further architectural flow variations are not covered by any claim in this article.

Cisco names five supported call flow types for troubleshooting: on-net optimized, on-net unoptimized, on-net cloud hosted, and off-net through a cloud connected PSTN (CCPP) provider or through a local gateway 23. On an on-net optimized call where interactive connectivity establishment (ICE) succeeds, the cloud media relay is removed from the path and media flows directly between the devices 27. Cisco names three causes of media path optimization failure: unsuccessful access to the session traversal utilities server, a failed connectivity check, and the default path's round-trip time being similar to or better than any optimized path 57.

Visibility changes depending on whether traffic stays on VoIP or crosses a PSTN boundary:

• Cisco's troubleshooting technical paper advises preferring voice over internet protocol (VoIP) calls over PSTN calls for better diagnostic visibility and quality 46.
• For PSTN calls through a cloud connected provider or a local gateway, hop data is collected from the Webex Calling cloud and not from the public network endpoint 48.
• For an off-net call through CCPP, the end-to-end experience and receive stream metrics of the caller on the public network are not available because those metrics are not transmitted to the Webex Calling cloud 15.
• For an off-net call through a local gateway, the end-to-end experience and receive stream metrics of the callee are not available because those metrics are not transmitted to the Webex Calling cloud 30.
• WarmTransfer's reading of the sources is that the media server used to collect or deposit voicemail does not report metrics, and Cisco asserts that the quality of that segment is internal and always high quality and low latency 61.

In some call park and retrieve scenarios, only one call leg appears and the administrator must manually correlate the two legs using a common entity such as overlapping time, with Cisco naming a future enhancement to remove that need without giving a date 10. Trunk and route group fields appear only when a local gateway is involved in the hop, and the public network vendor name field appears only when the PSTN is involved 60.

Hop Details and Diagnostic Indicators

Local and public internet protocol (IP) addresses in the hop details view are partially masked to preserve users' personal identity 29. Cisco states the geolocation derived from a public address is not accurate when the endpoint connects over private network connect (PNC), or when a Webex App user reaches the enterprise through a virtual private network (VPN) 24.

Network infrastructure details are collected only when an administrator explicitly turns on the setting allowing Cisco to collect network details under organisation settings, which is off by default 35. RoomOS network infrastructure coverage splits by release: switch chassis identifier, switch port identifier, and wireless basic service set identifier (BSSID) arrive in RoomOS 26.9.1 of September 2026, while switch name and wireless network name arrive in RoomOS 26.10.1 of October 2026 36. Cisco instructs administrators not to use a missing network field as evidence of a network fault and to look for a repeatable pattern across affected calls instead 34.

The minute-by-minute view is available only for calls handled through Webex clients and RoomOS endpoints, is not available for multiplatform phones (MPP) or Cisco-managed third-party endpoints, and is stated as coming for PhoneOS clients in the future 33. Webex Calling captures and processes media quality metrics from three named third-party device vendors: Polycom, Yealink, and AudioCodes 55. Media quality metrics and originating and terminating network-side flows are not available for analog phones or for endpoints using IPv6 7.

A minimum version cell in the advanced diagnostics device support matrix contains the single unseparated string ROOMOS 2019-08-16 3DAC3CFA60C, which reads as two identifiers merged into one cell 21.

Diagnostic Workflows and External Tooling

Cisco's troubleshooting technical paper offers a nine-step linear checklist rather than a branching decision tree: identify the issue, gather details, search in Control Hub, review indicators, check connection type, analyse hardware and environment, inspect the network path, advise the user, and set alerts 42. WarmTransfer's reading of the sources is that in one worked example in Cisco's technical paper, correlated end-to-end and receive packet loss spikes are deduced to point at the local participant's own connection, recommending an investigation into that participant's local network internet connection and computer 45. Cisco also maps no-media conditions to the party responsible for fixing them: media not received from a local gateway is for the customer administrator to investigate, while media not received from cloud public network or not sent to the receiver is addressed by Webex engineering 40. For network assessment beyond what Control Hub itself shows, Cisco's technical paper directs the reader to the ThousandEyes and Meraki integrations, detailed in ThousandEyes and Meraki for voice network assessment 44.

Control Hub Alerts and Notifications

Cisco renamed Alert Center to Notifications in a July 2026 change that also removed the Manage tab, moved the all rules and my rules sub-tabs under the Alerts tab, and renamed Requests to Approvals 3. Additional alerting rules are not covered by any claim in this article. Call quality drop alerts are sent when a call ends, and an organisation may create at most ten of them 11. Two Cisco documents give different illustrative alert thresholds for the same kind of rule: the notifications article uses more than 300 milliseconds latency or 8 percent packet loss, whereas the technical paper uses more than 800 milliseconds latency or 8 percent packet loss 4.

Call quality drop, trunk status change, and trunk concurrent call limit alerts are not triggered on Dedicated Instance based trunks 6. Trunk certificate expiry alerts are sent sixty, thirty, fifteen, and seven days before expiry and once immediately after the certificate expires 59. A separate certificate expiry alert for Dedicated Instance is sent sixty-seven days before expiration and then every twelve days until the certificate is renewed 20. When concurrent calls exceed the route list calls licence limit, up to twenty percent additional capacity is allowed, and off-net calls beyond that bursting limit are blocked 50.

Analytics Evolution and Changelog Notes

In April 2026, four call-level media quality indicators—total calls, call segments, total calls with media, and number of media sessions—were added to troubleshooting search by user or email address 9. In the same month, the Webex Calling media quality dashboard in Analytics renamed call legs to media sessions and split its data into Webex Calling and Calls on Webex categories 8. The Control Hub changelog carries the same entry titled Group repeating alerts in Alert Center under both July 2026 and February 2026, with no explanation of the duplication 16. WarmTransfer's reading of the sources is that the Control Hub changelog does not state its own update cadence or refresh commitment anywhere in its text 17.

Applicability

Across the claims in this article, the evidence covers Webex Calling multi-tenant deployments, verified as of 2026-09-05. Specific Dedicated Instance behaviors for certificate schedule and alert restrictions apply as noted 6​20. Media quality metrics are not currently supported for customers based in India, with support planned for a later date 28. Device-specific network infrastructure fields apply to RoomOS 26.9.1 (September 2026) and RoomOS 26.10.1 (October 2026) 36.

What remains uncertain

The exact release schedule for the planned support of media quality metrics in India is not covered by any claim in this article. The target delivery date for automatic leg correlation on call park and retrieve scenarios is not covered by any claim in this article. The availability schedule for minute-by-minute diagnostic views on PhoneOS clients is not covered by any claim in this article. Why the Control Hub changelog duplicated the repeating alerts entry across both July 2026 and February 2026 is not covered by any claim in this article.

WarmTransfer. (2026, September 5). Webex Calling call quality troubleshooting in Control Hub. WarmTransfer. https://warmtransfer.net/knowledge/webex-calling-troubleshooting

@misc{warmtransfer-webex-calling-troubleshooting,
  title  = {Webex Calling call quality troubleshooting in Control Hub},
  author = {{WarmTransfer}},
  year   = {2026},
  url    = {https://warmtransfer.net/knowledge/webex-calling-troubleshooting},
  note   = {Verified 2026-09-05}
}