Is job id and session id same ?

Thanks for reaching out, Aditya. I'm an AI assistant and I'm here to help. Give me a moment while I research this topic.

how can i get the session id ?

You're asking how to get the LiveKit session ID — here are the ways you can retrieve it, depending on your needs. • Use the Analytics API: call the List Sessions endpoint to get sessions for a project; each session object includes a

sessionId

. a. Example:

curl -H "Authorization: Bearer $TOKEN" \
     "<https://cloud-api.livekit.io/api/project/$PROJECT_ID/sessions>"

• From server-side room/job context: in Agents

JobContext

you can access room identifiers (e.g.,

ctx.job.room.name

or

ctx.room.name

) to correlate with sessions. • From session/detail endpoints or participant lists: session details and participant connection timestamps returned by the analytics/session endpoints let you correlate participants or tracks to a specific

sessionId

. Would you like the exact Analytics API response format or help locating the

sessionId

for a specific project or job? Sources: Analytics API | LiveKit docs | Job lifecycle | LiveKit docs

is there 1:1 relation between job and session id ?

Short answer: No — job IDs and session IDs are not guaranteed to be 1:1. • A job ID identifies a single agent execution (one agent invocation). • A session ID (room session) identifies the overall LiveKit communication session for a room. Relationships: • One agent job → one job ID and one agent session context (1:1 for that agent execution). • One room session → can contain many agent jobs (many job IDs map to the same session ID). • If your setup dispatches a single agent per room, you'll see a 1:1 correspondence by design, but that is not enforced by the system. Would you like help mapping job IDs to session IDs from your Analytics API response or agent logs? Sources: Agents (Node.js) | livekit/livekit_composite | DeepWiki | Job lifecycle | LiveKit docs | Analytics API | LiveKit docs | Agent dispatch | LiveKit docs | Observability data hooks | LiveKit docs

is there a no way to get the session id when the session starts or ends ?

You're asking whether you can obtain a LiveKit session ID when a session starts or ends — yes. Here are the practical ways to get it. • Analytics API (recommended): Call the List Sessions endpoint to see active sessions; each session object includes

sessionId

. Use the session list at session start and the session detail endpoint after end to get

sessionId

, start/end times, and final metadata. ◦ Example:

curl -H "Authorization: Bearer $TOKEN" \
      "<https://cloud-api.livekit.io/api/project/$PROJECT_ID/sessions>"

• Correlate from agent/job context: In Agents, use

JobContext

(e.g.,

ctx.room.name

or

ctx.job.room.name

) to identify the room; then query the Analytics API to find the matching

sessionId

. Agents also expose lifecycle hooks (e.g.,

AgentSession

close event) you can use to trigger the lookup at session end. • Room lifecycle events: Listen for room disconnection reasons such as

ROOM_CLOSED

(emitted when all participants leave). When that event fires, query the Analytics session detail for the room to obtain the sessionId and end timestamp. Do you want a short script to call the Analytics API and map

roomName

to

sessionId

, or help wiring this into your agent/session event handler? Sources: Analytics API | LiveKit docs

what is the difference between shutdown callback and on session end hook ?

You're asking for the difference between the shutdown callback and the on-session-end hook — here’s a concise comparison and when to use each. • on_session_end (Python only) ◦ Trigger: Runs when a session ends (part of the session termination flow). ◦ Purpose: Produce a final session report or export structured session data (transcripts, events, metadata). ◦ Access: Full session data via

ctx.make_session_report()

. ◦ Defined via:

@server.rtc_session(on_session_end=...)

. ◦ Use it to: Save analytics, export transcripts, or persist session-level records. • Shutdown callback (Python & Node.js) ◦ Trigger: Runs after

session.shutdown()

/

ctx.shutdown()

when the session is closing, before the process exits. The framework waits (default up to 60s) for these callbacks to finish. ◦ Purpose: Cleanup resources — close DB connections, flush logs, save transient state. ◦ Access: Session is already closing; less direct access to live session internals. ◦ Defined via:

ctx.add_shutdown_callback(...)

. ◦ Use it to: Ensure graceful cleanup and resource release. How they fit together: • In Python you can use both:

on_session_end

to generate/export the final session report, then shutdown callbacks to perform cleanup. • In Node.js use shutdown callbacks for cleanup since

on_session_end

is not available. Would you like a short Python example showing both wired together? Sources: Job lifecycle | LiveKit docs | Observability data hooks | LiveKit docs