Performance monitoring

You can view a wide range of statistical information through the VoiceAI Connect Enterprise Web UI. VoiceAI Connect Enterprise uses the open-source Grafana monitoring tool, which provides a default dashboards with various graphs displaying statistical information.

The statistical data is retrieved from VoiceAI Connect Enterprise's InfluxDB database server and displayed on the following default dashboards:

• VoiceAI default dashboard:

  • Disk usage per host

  • Memory usage per host

  • CPU usage per host

  • Currently active calls (gauge)

  • Active calls

  • Failed calls per Session Manager

  • Successful calls versus failed calls

• VoiceAI bots dashboard:

  • Total number of failed calls per bot (bar graph)

  • Total number of calls per bot (bar graph)

  • Active calls per bot

  • Bot messages and failures per bot

  • Text-to-speech characters per bot

  • Speech-to-text billing time per bot

• VoiceAI providers dashboard:

  • Speech-to-text requests and failures per provider

  • Text-to-speech requests and failures per provider

  • Text-to-speech delay (max. and average) per provider

If you want, you can also create your own dashboard and choose the statistical graphs that you want on them.

To view performance monitoring:

1. From the Navigation menu, click Monitoring.

2. To view a default dashboard or create a new dashboard:

   1. Click the hamburger icon in the top-left corner to open the navigation menu.

   2. Select Dashboards, and then in the right pane, select the desired default dashboard or click New to create one:

3. Some of the actions that you can do in Grafana:

   • On the top-right corner, a bar with buttons allows you to perform various operations: you can click to refresh the data periodically. You can click to change the time range for the statistical data. You can also save a snapshot of the dashboard.

   • You can edit the dashboard, or copy it and then edit the copy.

   • Hovering over a graph shows the data in numbers.

   • Hovering over a graph and then pressing the "v" key shows the graph in full screen.

   • Dragging the mouse over a period in a single graph view, zooms the graph to that time range.

The KPIs available for use in Grafana queries are listed in the following table:

Measurement	Field	Tag indexes	Description
disk	used_percent	host	Main partition usage in percentage
mem	available_percent	host	Total available RAM in percentage
cpu	usage_idle	host	Total CPU idle in percentage
status	CallAttemptsPerMinute	botId, ttsProviderId, sttProviderId, sessionManagerId	Call attempts per minute
status	SuccessfulCallsPerMinute	botId, ttsProviderId, sttProviderId, sessionManagerId	Successful calls per minute
status	CallEndingsPerMinute	botId, ttsProviderId, sttProviderId, sessionManagerId	Call endings per minute (success and failures)
status	STTFailuresPerMinute	botId, ttsProviderId, sttProviderId, sessionManagerId	STT failures per minute
status	TTSFailuresPerMinute	botId, ttsProviderId, sttProviderId, sessionManagerId	TTS failures per minute
status	BotFailuresPerMinute	botId, ttsProviderId, sttProviderId, sessionManagerId	Bot failures per minute
status	SBCFailuresPerMinute	botId, ttsProviderId, sttProviderId, sessionManagerId	SBC failures per minute
status	PlayUrlFailuresPerMinute	botId, ttsProviderId, sttProviderId, sessionManagerId	PlayUrl failures per minute
status	AverageCallDurationSeconds	botId, ttsProviderId, sttProviderId, sessionManagerId	Average call duration in seconds
status	IncomingMessagesPerMinute	botId, ttsProviderId, sttProviderId, sessionManagerId	Bot incoming messages per minute
status	OutgoingMessagesPerMinute	botId, ttsProviderId, sttProviderId, sessionManagerId	Bot outgoing messages per minute
status	AverageTTSDelayMilliseconds	botId, ttsProviderId, sttProviderId, sessionManagerId	Average TTS delay in milliseconds
status	MaxTTSDelayMilliseconds	botId, ttsProviderId, sttProviderId, sessionManagerId	Max TTS delay in milliseconds
status	TTSRequestsPerMinute	botId, ttsProviderId, sttProviderId, sessionManagerId	TTS requests per minute (not cached)
status	TTSCharactersPerMinute	botId, ttsProviderId, sttProviderId, sessionManagerId	TTS characters per minute (not cached)
status	CachedTTSRequestsPerMinute	botId, ttsProviderId, sttProviderId, sessionManagerId	Cached TTS requests per minute
status	CachedTTSCharactersPerMinute	botId, ttsProviderId, sttProviderId, sessionManagerId	Cached TTS characters per minute
status	STTRawTimeInSecondsPerMinute	botId, ttsProviderId, sttProviderId, sessionManagerId	STT raw time (in seconds) per minute
status	STTBillingTimeInSecondsPerMinute	botId, ttsProviderId, sttProviderId, sessionManagerId	STT billing time (in seconds) per minute
status	STTRequestsPerMinute	botId, ttsProviderId, sttProviderId, sessionManagerId	STT requests per minute
status	PlayUrlRequestsPerMinute	botId, ttsProviderId, sttProviderId, sessionManagerId	PlayUrl requests per minute (not cached)
status	CachedPlayUrlRequestsPerMinute	botId, ttsProviderId, sttProviderId, sessionManagerId	Cached PlayUrl requests per minute
status	AveragePlayUrlDelayMilliseconds	botId, ttsProviderId, sttProviderId, sessionManagerId	Average PlayUrl delay in milliseconds
status	MaxPlayUrlDelayMilliseconds	botId, ttsProviderId, sttProviderId, sessionManagerId	Max PlayUrl delay in milliseconds
status	ActiveCalls	botId, ttsProviderId, sttProviderId, sessionManagerId	Currently active calls
status	SuccessfulOutboundCallsPerMinute	botId, ttsProviderId, sttProviderId, sessionManagerId	Successful outbound calls per minute
status	OutboundCallEndingsPerMinute	botId, ttsProviderId, sttProviderId, sessionManagerId	Outbound call endings per minute (success and failures)
status	CompletedOutboundCallsPerMinute	botId, ttsProviderId, sttProviderId, sessionManagerId	Completed (answered and ended successfully) outbound calls per minute
status	ThrottledCallsPerMinute	botId, ttsProviderId, sttProviderId, sessionManagerId	Throttled calls per minute (by call throttling feature)