API service deployment unhealthy

[simone-kalbermatter]

After deploying a self-hosted E2B cluster following the Self-Hosting Guide, the api job in Nomad is constantly reported as unhealthy, despite the container starting up and logging no significant errors.

stdout:

{"level":"info","timestamp":"2025-03-07T20:18:47.535001655Z","message":"Starting API service...","service":"orchestration-api","internal":true,"pid":1,"commit_sha":"4cab1b6d"}
{"level":"info","timestamp":"2025-03-07T20:18:47.554439137Z","message":"initializing API store and services","service":"orchestration-api","internal":true,"pid":1}
{"level":"info","timestamp":"2025-03-07T20:18:47.554769512Z","message":"created Supabase client","service":"orchestration-api","internal":true,"pid":1}
{"level":"info","timestamp":"2025-03-07T20:18:47.554895967Z","message":"No Posthog API key provided, silencing logs","service":"orchestration-api","internal":true,"pid":1}
Host for gRPC not set, using dummy connection
{"level":"info","timestamp":"2025-03-07T20:18:47.555351Z","message":"Starting DNS server","service":"orchestration-api","internal":true,"pid":1}
{"level":"info","timestamp":"2025-03-07T20:18:47.555997555Z","message":"Running the initial node sync","service":"orchestration-api","internal":true,"pid":1}
{"level":"info","timestamp":"2025-03-07T20:18:47.558889651Z","message":"http service starting","service":"orchestration-api","internal":true,"pid":1,"port":50001}
{"level":"info","timestamp":"2025-03-07T20:19:07.556846988Z","message":"API internal status","service":"orchestration-api","internal":true,"pid":1,"sandboxes_count":0,"nodes_count":0,"nodes":[]}
{"level":"info","timestamp":"2025-03-07T20:19:27.556118427Z","message":"API internal status","service":"orchestration-api","internal":true,"pid":1,"sandboxes_count":0,"nodes_count":0,"nodes":[]}

stderr:

2025/03/07 20:18:48 Failed to connect to otel collector, not using OTEL for logs: context deadline exceeded
2025/03/07 20:18:54 Failed to connect to otel collector, not using OTEL for logs: context deadline exceeded

I'm not sure how to debug this, any help would be much appreciated!

[linear[bot]]

E2B-1729 API service deployment unhealthy

[jakubno]

It looks like there's an issue with your orchestrator or client node. The API server can't start without connecting to an orchestrator. Could you please check this and let me know?

[simone-kalbermatter]

The problem was probably because I tried to deploy this from the main branch. Should the deployment be made from the e2b-prod or e2b-production branch?

[jakubno]

It's possible there was some issue, can you check if orchestrator is running and what's in orchestrator's logs?

[ayshsens]

same issue with main branch https://github.com/e2b-dev/infra/issues/793

can you share a stable version which can be used? Is last release stable https://github.com/e2b-dev/infra/releases/tag/v0.1.4

[sitole]

Hi, looks like this issue is caused in the Ubuntu machine image we were using during the Packer build. There was an issue with Consul networking that was causing the services from different machines. We have a fixed version currently deployed in main (more info can be found here https://github.com/e2b-dev/infra/issues/731#issuecomment-2981840507).

Please try to rebuild machine images and redeploy if the issue is resolved.

[wangzihaoha]

same issue with main branch主分支有同样的问题 #793

can you share a stable version which can be used?可以分享一个可以使用的稳定版本吗？ Is last release stable最新发布版本是否稳定 https://github.com/e2b-dev/infra/releases/tag/v0.1.4

Hello, I would like to know if you have successfully deployed E2B.

[wangzihaoha]

The problem was probably because I tried to deploy this from the main branch.问题可能是因为我尝试从主分支部署它。 Should the deployment be made from the e2b-prod or e2b-production branch?部署应该从 e2b-prod 还是 e2b-production 分支进行？

Hello, I would like to know if you have successfully deployed E2B.

[jakubno]

Hello, I would like to know if you have successfully deployed E2B.

Could you please create a separate issue and include a description of the issue you're having?

[jakubno]

This issue should be resolved. If you continue to experience similar problems, please feel free to reopen the case.