Reviewing Genesys Cloud Log Files for Network Issues

Analyze Genesys Cloud log captures and use admin-side tooling to identify network-related indicators that may be causing call quality issues, connectivity failures, or softphone problems.

How to Obtain the Logs

1. In Genesys Cloud, navigate to Admin → IT and Integrations → Log Capture
2. Search for the affected user and enable log capture
3. Wait up to 24 hours for the capture period to complete

Note: Logs are not available until the capture period ends. Enabling log capture does not retroactively capture past events.

---

Log Structure Overview

Levels:

• Debug
• Info
• Log
• Warn
• Error

Each log line follows this format: MESSAGE [TIMESTAMP]

• Message — Prefixes indicate the subsystem: [webrtc-sdk], [streaming-client], CP:, ACD:, etc.
• Timestamp — Locale-style, non-ISO format: MM/dd/yyyy HH:mm:ss.fff

---

Step 1 — Search for WebRTC / Softphone Connection Issues

Search terms:

• [webrtc-sdk]
• softphone
• station

What to look for:

Log Entry	Healthy	Problem Indicator
streaming client connected	Present	Missing = connection failure
Fetched user station — type: 'inin_webrtc_softphone'	Present	Missing or wrong type = station misconfiguration
webRtcPersistentEnabled: true	true	false may cause repeated re-registration
webRtcForceTurn: false	false	true = TURN relay forced, indicates direct media path is blocked (network issue)
webRtcCallAppearances	1 or more	0 = no call appearances configured
concurrentSoftphoneSessionsEnabled	Informational	Check if it matches expected config

Understanding webRtcForceTurn:

When webRtcForceTurn: true appears, all media is routed through Genesys TURN servers rather than peer-to-peer. This is automatically set when the platform detects that the direct UDP media path is blocked. It almost always indicates a firewall or NAT configuration issue on the user’s network:

• Corporate firewalls blocking UDP ports 16384–65535
• Symmetric NAT preventing ICE hole-punching
• SSL inspection proxies interfering with DTLS/SRTP media

---

Step 2 — Check Streaming Client Connectivity

Search terms:

• [streaming-client]
• stanza instance
• channelId

What to look for:

Log Entry	Meaning
new stanza instance connected	WebSocket signaling channel established — good
attempting to connect streaming client on channel	A reconnection attempt — possible if it repeats frequently
Streaming Service: Realtime connection finished	Realtime WS connected
streaming client connected / CP: Streaming service -> CP Connected	Full signaling path is up

Problem indicators:

• Repeated attempting to connect entries = unstable WebSocket connection (network drops, proxy interference)
• Long gap between attempting to connect and connected = slow network or DNS resolution issue
• Multiple channelId values in quick succession = client is being disconnected and reconnecting (network instability)

WebSocket requirements:

Genesys Cloud signaling relies on a persistent WebSocket connection over TCP port 443 to *.mypurecloud.com (or the regional equivalent). Issues arise when:

• Proxies or load balancers terminate WebSocket connections after an idle timeout
• SSL/TLS inspection proxies break the UPGRADE request
• Intermediate network devices do not support WebSocket keep-alives

---

Step 3 — Check Media Device Enumeration

Search terms:

• webrtcDevices enumerated
• getUserMedia
• availableDevices

What to look for:

Log Entry	Healthy	Problem
webrtcDevices enumerated — devices: [{}, {}, ...]	Multiple devices listed	Empty array [] = no audio devices found
requesting getUserMedia — audio: true	Media request initiated	Never appears = browser permission issue
returning media from getUserMedia	Media acquired	Missing = mic blocked or device error
finished requesting media permissions. destroying media	Normal init flow	If followed by errors = device problem
availableDevices listing	Shows named devices like Jack Mic (Realtek(R) Audio)	Devices with displayLabel: '' = permission not yet granted or device issue

Key check — Healthy audio device path sequence:

A healthy softphone initialization will show these entries in order:

webrtcDevices enumerated
requesting getUserMedia — audio: true
returning media from getUserMedia
finished requesting media permissions. destroying media
availableDevices — [<named devices>]

If this sequence is incomplete or out of order, the issue is likely a browser permission or audio device problem — not a network issue. This means the problem is outside of Genesys and should be directed to the agent to check their browser permissions and audio device setup.

---

Step 4 — Look for Errors and Warnings

Categorize each error:

Error Pattern	Category	Action
TransitionAborted	UI/Routing — typically benign, Ember.js route transition interrupted	Ignore unless frequent
Failed to authenticate gcba / gcbaId is not provided	Non-critical — GCBA (Genesys Cloud Background Assistant) not configured	Ignore; will retry automatically
missing.permissions / status: 403	Permissions — not network related	Inform admin to check user role/permissions
Failed to retrieve export expiration summary	Permissions — user lacks analytics:dataExport:view	Not network; role config issue
[gux-button-slot] errors	UI Component — cosmetic rendering issue	Ignore
Connection timeout / socket errors	NETWORK — investigate further	Check firewall, proxy, DNS
WebSocket close/error events	NETWORK — signaling path interrupted	Check TCP 443, proxy/SSL inspection
ICE candidate failures	NETWORK — NAT traversal failing	Check STUN/TURN ports (3478, 19302)
OWASP/certificate errors	NETWORK/SECURITY — SSL interception	Check if proxy is intercepting TLS
Failed to set remote description	MEDIA — SDP negotiation failure	May be codec mismatch or NAT issue
IceConnectionState: failed / IceConnectionState: disconnected	NETWORK — media path broken after connecting	Indicates mid-call network disruption
peerConnection closed unexpectedly	NETWORK/MEDIA — connection dropped	Correlate with timestamp for gap analysis

ICE state machine — what it means:

The ICE (Interactive Connectivity Establishment) connection state progresses through several states. Abnormal states indicate network problems:

ICE State	Meaning
checking	Attempting to find a media path — normal during setup
connected	Media path found — call can proceed
completed	Best media path selected — healthy
disconnected	Temporary media path loss — may self-recover
failed	No media path could be established — call will drop
closed	Connection deliberately torn down

A state sequence of checking → failed without ever reaching connected definitively indicates a network/firewall issue blocking media.

---

Step 5 — Timeline Analysis for Network Issues

Measure connection timing:

Use timestamps to evaluate how long key operations take:

Operation	Expected	Concern
Login → streaming client connected	< 5 seconds	> 10s = slow network or DNS
requesting getUserMedia → returning media from getUserMedia	< 2 seconds	> 5s = device/driver issue
stanza instance connected → Streaming service -> CP Connected	< 3 seconds	> 5s = WebSocket latency
Gap between repeated Will retry GCBA authentication	Increasing backoff (4s, 8s, 16s…)	Normal retry behavior
IceConnectionState: checking → IceConnectionState: connected	< 3 seconds	> 8s = NAT/firewall causing slow ICE negotiation

Check for time gaps:

Look for unexplained gaps (> 5 seconds) between sequential log entries during an active call.

These suggest:

• Network drops
• CPU/browser freezes
• WebSocket disconnection
• Mid-call ICE restart triggered by connectivity loss

Mid-call disconnection pattern:

If a call drops mid-conversation, look for this sequence:

IceConnectionState: disconnected
[webrtc-sdk] attempting ICE restart
IceConnectionState: failed
peerConnection closed

This indicates the media path was lost after being established, suggesting network instability rather than a configuration issue. Compare timestamps against any known network events (VPN disconnection, Wi-Fi handoff, etc.).

---

Step 6 — Check the Region / Environment

Search terms:

• pure.cloud
• hostRootUrl
• mypurecloud

The log should show the correct Genesys Cloud region:

Region Code	AWS Region	Primary Domain
cac1	ca-central-1 (Canada)	cac1.pure.cloud
use1	us-east-1 (US East — Virginia)	mypurecloud.com
use2	us-east-2 (US East 2 — Ohio, GovCloud)	use2.us-gov-pure.cloud
usw2	us-west-2 (US West — Oregon)	usw2.pure.cloud
euw1	eu-west-1 (EU — Ireland)	mypurecloud.ie
euw2	eu-west-2 (EU — London)	euw2.pure.cloud
euc1	eu-central-1 (EU — Frankfurt)	mypurecloud.de
aps1	ap-southeast-2 (Asia Pacific — Sydney)	mypurecloud.com.au
apne1	ap-northeast-1 (Asia Pacific — Tokyo)	mypurecloud.jp

Confirm this matches the expected organization region.

If the region is wrong, a DNS bypass, proxy misconfiguration, or VPN split-tunnel policy may be routing the user to the incorrect environment. This causes authentication failures and connectivity issues because the user’s organization does not exist in that region.

---

Step 7 — Genesys Cloud Admin Checks

7a — Conversation Details: Timeline View

Navigate to Analytics → Analytics Workspace in the left navigation. Search by the affected agent’s name and filter to the relevant time window.

Open an affected interaction and view the Timeline tab. This shows a visual breakdown of each call segment and how it ended.

Key things to check:

Field	What to check
Disconnect type	Endpoint = agent/softphone side disconnected; Peer = far-end (caller) disconnected; System = platform-level disconnect
Duration vs. Handle time	A large gap between when the call connected and when audio began may indicate a media establishment failure
Segments	Multiple short segments on a single interaction may indicate reconnections or unexpected re-routing
Wrap-up	Confirm wrap-up was applied — missing wrap-up after a short call may mean the agent was dropped unexpectedly

A pattern of repeated Endpoint disconnects on short calls for a specific agent — particularly without wrap-up — is a strong indicator the softphone is dropping calls, pointing to a connectivity issue on their end.

7b — Genesys Cloud Quality Management: Call Recordings

If call recordings are enabled, review recordings for:

• One-way audio — agent cannot hear caller or vice versa → likely ICE failure where one media direction was established
• Audio cutting out at consistent intervals → may indicate a load balancer or proxy with keepalive timeout dropping the media stream
• Complete silence from call start → ICE negotiation failed entirely
• Echo or feedback → not typically a network issue; local acoustic problem

Even if the recording appears silent, the recording being present confirms that the call was connected at the signaling level. If a recording does not exist at all, the call was never established (signaling failure).

7c — Genesys Cloud Status Page

Check status.mypurecloud.com for any active or recent incidents affecting:

• Voice / WebRTC services
• Streaming (WebSocket) services
• The specific region

If there is an active incident in the user’s region, that may be the cause. Document the incident number for correlation before proceeding with deeper network investigation.

---

Step 8 — Network Readiness Assessment Tool

If the log review or admin checks point to a potential network issue on the agent’s side, the Network Readiness Assessment Tool can be used to get a concrete measurement of their network conditions. This test must be run by the agent on their own machine and network.

Note: The Genesys Cloud Network Readiness Assessment Tool will be deprecated on April 1, 2026.

Ask the agent to follow these steps and share the results with you.

How to Run the Test

1. From the affected agent’s PC, open a browser
2. Navigate to: http://canada.pna.inindca.com:8080/myspeed/purecloud_presales_version_1.html
3. Enter the organization name (e.g. maximuscanada)
4. Download and run the .exe file
5. Allow the test to complete fully — do not close the browser

Tip: Ask the agent to run the test at the same time of day that the issue typically occurs, to capture realistic network conditions.

Interpreting the VoIP Results

Metric	Good	Warning	Fail
Jitter (Up/Down)	< 10ms	10–30ms	> 30ms — network congestion or poor QoS
Packet Loss (Up/Down)	0%	0.1–1%	> 1% — significant quality impact
Packet Discards	0%	< 1%	> 1% — buffer/QoS issue
Packet Order	100%	95–99%	< 95% — routing instability
Round Trip Time (RTT)	< 80ms	80–150ms	> 150ms — latency issue, check routing
MOS Score (Up/Down)	4.0+	3.5–4.0	< 3.5 — users will notice quality degradation

• All green → Network quality is not the issue. Focus investigation on Genesys config or the agent’s device/headset.
• Any Warning or Fail → The agent’s network is a contributing factor. Refer findings to their IT/network team.

Interpreting the Firewall Results

All ports should show ✔ Passed. Key ports to watch:

Port(s)	Protocol	Purpose	Impact if Blocked
443	TCP	WebSocket / TLS signaling	Softphone won’t connect
5060 / 5061	TCP	SIP signaling	Calls can’t be established
3478 / 19302	TCP & UDP	STUN/TURN (NAT traversal)	One-way or no audio
16384–65535	UDP	RTP media (voice traffic)	No audio or choppy audio
53	TCP & UDP	DNS resolution	General connectivity failures

• All passed → Firewall is not blocking Genesys traffic.
• Any failed → Pass the specific failed port(s) and protocol to the agent’s network/security team for remediation.

Summary of Outcomes

VoIP	Firewall	Action
Good	Good	Network is clear — investigate Genesys config, headset, or PC performance
Poor	Good	Recommend QoS review or wired connection — share failing metrics with network team
Good	Failures	Provide blocked port list to network/security team
Poor	Failures	Both QoS and firewall changes needed before Genesys Cloud can function reliably

Symptom-to-Metric Quick Reference

Symptom	Check these metrics first
Choppy / robotic audio	Jitter, Packet Loss
One-way audio	UDP ports (3478, 16384–65535), STUN/TURN ports
Calls dropping	RTT, Packet Loss, TCP 443/5061
Delay or talk-over	RTT (round trip time)
Softphone won’t connect	TCP 443, TCP 5060/5061, DNS (port 53)
Overall poor quality	MOS Score (< 3.5 = problem)

---

Step 9 — Determining the Source of the Issue

Before concluding where the issue lies, use the following checks to build a complete picture. The goal is to determine whether Genesys is functioning normally and the issue originates outside the platform.

Check	How
Is there an active platform incident?	Check status.mypurecloud.com for the relevant region
Is it isolated to one agent?	Check Analytics for other agents at the same office/location
Is it affecting multiple agents in the same location?	Suggests a shared network issue rather than individual configuration
Is it affecting agents on different networks?	If agents at other sites or working remotely are also affected, it is more likely a platform issue
When did it start?	Cross-reference with the Genesys status page and any known network/infrastructure changes at that time

---

Quick Reference: Log Indicators and What They Mean

Indicator	What it suggests	Next step
webRtcForceTurn: true	Genesys cannot establish a direct media path — likely a network/firewall issue on the agent’s side	Note in ticket; agent’s IT team should check firewall rules (UDP 16384–65535)
IceConnectionState: failed	Media path could not be established after signaling connected	Network issue on agent side — confirm no platform incident, then refer to IT
IceConnectionState: disconnected → failed	Media path was lost mid-call	Network instability — correlate with call timestamps
Repeated streaming client reconnections	WebSocket signaling is dropping and reconnecting	Agent-side network instability or proxy interference
WebSocket close/error events	Signaling path was interrupted	Could be network or platform — check status page first
Long gap between attempting to connect → connected	Slow connection establishment	DNS resolution or network latency issue
Empty availableDevices array	No audio devices detected in the browser	Not a network issue — browser/device problem on agent’s machine
Failed to set remote description	SDP negotiation failed during call setup	May be network NAT issue or platform-side problem — check status page
403 missing.permissions	Agent lacks a required permission	Not network — check role configuration in Admin
TransitionAborted	Ember.js UI route transition interrupted	Benign — ignore unless frequent
Failed to authenticate gcba	GCBA feature not configured	Non-critical — ignore
gux-button-slot errors	UI rendering issue	Cosmetic — ignore
One-way audio in recording	One direction of the media path failed	ICE-level issue; note call ID and refer to Genesys support if platform appears healthy

---

Escalation

The issue appears to be on the agent’s network/device (not a platform issue)

Log indicators: webRtcForceTurn: true, IceConnectionState: failed, repeated streaming client reconnections, or audio device enumeration failures.

Action: Document findings and refer to the agent’s IT support team. Provide:

• The relevant log entries with timestamps
• The specific indicators found (e.g., webRtcForceTurn: true, ICE failures)
• The affected agent’s name and the time window

Network requirements to pass along to their IT team if needed:

Protocol	Port	Purpose
TCP	443	HTTPS, WebSocket signaling, TURN over TLS
UDP / TCP	3478	STUN/TURN
UDP	16384–65535	WebRTC media (RTP/SRTP)

The issue appears to be platform-side

Log indicators: system disconnect type, errors not related to network or device, issue affecting multiple agents across different networks, or correlation with a status page incident.

Action: Open a Genesys Support case. Include:

• The Log Capture file
• Conversation IDs of affected calls (from Analytics → Analytics Workspace)
• The org name and region
• A summary of findings from this review