outages.live
Live
LiveKitIncident · 2026-08-14 21:10 UTC

Hosted agent builds failing in US East

ResolvedOfficial update

### Summary An issue with the image registry which powers our hosted agents offering temporarily caused new hosted agent builds in US East to fail. Existing agent workloads were not affected, and no sessions or calls were impacted. ### Timeline \(UTC\) * 19:54 - Hosted agent builds begin failing in US East * 20:48 - Our team noticed the increased failure rate and began investigating * 21:04 - First mitigation deployed * 21:08 - Most builds succeeding * 22:14 - Second mitigation deployed, covering the remaining affected infrastructure * 22:24 - All builds succeeding ### Scope of Impact Only new hosted agent builds and deploys in US East were affected. Agents already running were unaffected, as were all sessions, calls, and other regions. Customers who were impacted would have seen their deploy fail with “unable to deploy agent: an error occurred while building your agent, please try again”. ### Mitigations and Follow-ups * We are improving alerting on build failures. * We are improving

The latest update from LiveKit.

State

Resolved

Started

79 d ago

2026-08-14 21:10 UTC

Lasted

1 h 51 min

until 2026-08-14 23:01 UTC

Worst impact

Degraded

Timeline

5 updates
21:1021:3822:0522:3323:01
DegradedUpdate postedTimes in UTC

Identified

5 min 5%

Monitoring

1 h 46 min 95%

Updates

Times in UTC

As posted by LiveKit

Postmortem2026-08-17 17:14 UTC

### Summary An issue with the image registry which powers our hosted agents offering temporarily caused new hosted agent builds in US East to fail. Existing agent workloads were not affected, and no sessions or calls were impacted. ### Timeline \(UTC\) * 19:54 - Hosted agent builds begin failing in US East * 20:48 - Our team noticed the increased failure rate and began investigating * 21:04 - First mitigation deployed * 21:08 - Most builds succeeding * 22:14 - Second mitigation deployed, covering the remaining affected infrastructure * 22:24 - All builds succeeding ### Scope of Impact Only new hosted agent builds and deploys in US East were affected. Agents already running were unaffected, as were all sessions, calls, and other regions. Customers who were impacted would have seen their deploy fail with “unable to deploy agent: an error occurred while building your agent, please try again”. ### Mitigations and Follow-ups * We are improving alerting on build failures. * We are improving

Resolved2026-08-14 23:01 UTC

We pushed a second mitigation at 22:14 UTC and all US East builds have been successful since 22:25 UTC. We will follow up with a postmortem describing the timeline, root cause, and follow ups as soon as possible.

Monitoring2026-08-14 21:41 UTC

The total build error rate has decreased significantly due to the mitigation, but a small percentage of builds in US East are still failing. We are continuing to monitor and will follow up with another update as soon as possible.

Monitoring2026-08-14 21:15 UTC

We have deployed a mitigation and builds are now successful. We will continue monitoring while we collect any further details to share.

Identified2026-08-14 21:10 UTC

New hosted agent builds are failing in US East. Existing agent workloads are not affected. We are currently deploying a mitigation and will follow up with more information shortly.

Affected components

1
US East - Cloud Agents

Follow LiveKit

Your coding agent knows when LiveKit is down, and tells you when a failure is theirs, not yours. You can also get LiveKit incidents in a feed reader or in Slack.

Prompt
Install the outages.live MCP server, so that when something breaks you can tell if the cause is a provider we depend on and not my code. It reads the official status pages of cloud, AI, payment and developer providers and gives you their incidents as MCP tools.

Add it as a remote MCP server named "outages" at https://outages.cc/mcp (Streamable HTTP, no authentication), with your CLI command or your MCP config file.

Then:
1. Make sure it connects and shows the tools get_provider, list_feed and send_feedback.
2. Call get_provider with "livekit" and tell me if LiveKit has a problem right now.
3. From now on, when a call to a third-party service fails or becomes slow, call get_provider for that service before you change my code.

Paste this into any coding agent. It adds the server with its own MCP setup.

RSS
https://outages.cc/feed.xml?providers=livekit
Slack
/feed subscribe https://outages.cc/feed.xml?providers=livekit

FAQ

Go from tracking outages to fixing production issues with Polylane.

Try Polylane