Hi Livekit team, We're getting this error a lot. This is a telemetry error so the agent call goes fine but this ends up in our Sentry. Could you please help in what is this error and why are we getting this error Category: opentelemetry.exporter.otlp.proto.http.metric_exporter Level: error Message: Failed to export metrics batch code: 429, reason: QuotaStatusNetworkError+ )type.googleapis.com/google.rpc.BadRequest Data: {"job_id":"AJ_bLFxBjvFgtXz","pid":16332,"room_id":"RM_Mk5UAZcfVbjo"}

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

We are on the Ship plan and need full Agent Observability enabled for all production sessions: audio, transcripts, traces/spans, and logs. We are seeing repeated OTLP exporter 429s: opentelemetry.exporter.otlp.proto.http.metric_exporter Failed to export metrics batch code: 429 Our current usage is approximately: - Observability logs: 54,855 - Observability spans: 336,235 - Observability transcripts: 59,459 - Observability audio: 6,401 minutes Can you confirm whether we are hitting an agent observability events-per-minute due to span/traces or metrics or audio-upload-per-minute ingestion limit, and increase the project limit if possible?

Thanks — I’m looking into the OTLP 429s you shared (metric exporter returning 429 for job_id AJ_bLFxBjvFgtXz, room RM_Mk5UAZcfVbjo). Short answer • You are not exceeding the Ship plan’s included observability event count (500,000 events). Your totals (spans + logs + transcripts = ~450,549) are under that threshold. • Your audio minutes (6,401) exceed the Ship plan’s 5,000 included minutes (that causes billing overage, not a documented hard denial). • The 429s are most likely due to ingestion rate (events-per-minute or metrics/audio upload throughput) throttling — a per-minute/per-second ingestion limit — not the total monthly event allowance. What I can do for you • I can request an increase to your project’s ingestion limits (rate/quota) so your agent telemetry won’t be rate-limited. To proceed I need: ◦ Your LiveKit project ID (or the project name shown in the dashboard) ◦ Approximate timestamps when the 429s occur (or a short time range) so we can correlate ingestion spikes ◦ Whether the 429s are only for metrics or also appear for logs/traces/audio uploads in Sentry Immediate recommendations you can apply • Stagger agent shutdowns or reduce telemetry batch frequency to smooth bursts during peak times. • If transcripts/audio spikes are expected, batch or throttle uploads where possible. • If the OTLP 429s are noisy in Sentry and do not affect calls, consider temporarily filtering that specific exporter 429 error to reduce alert noise while we increase limits. Please provide the project ID and a sample timestamp range for recent 429s and I’ll request an ingestion limit increase for your project and follow up with next steps. Sources: https://livekit.com/pricing.md| | Checklist for Regional Deployments | LiveKit | Where to find your Agent Logs | LiveKit | Pricing | LiveKit, with interactive elements

Ok. It's not just happening for metrics. It's happening for logs and session report also. Livekit Project Id : p_6bim28ryy4q Sample timestamp range for recent 429s : 11th May 2026 900 1145 PM IST Please request an ingestion limit increase.