Healthcheck interval seemingly being ignored

Troubleshooting and help
1

Hello,

I’m noticing that after I’ve deployed today the health check endpoint is being pinged every second despite the interval being set to 300 seconds (5 minutes). Am I missing something or is there an issue?

I’ve also posted this in the Slack community workspace. I wasn’t sure which to post to, so I posted to both. I apologize if that is considered a no-no.

Thanks,
Mike

2025-06-17 15:16:18,690 | INFO     | 140640202918784[Task-111] | slack_app       | Health check.
INFO:     15.204.221.39:29092 - "GET /health_check HTTP/1.1" 200 OK
2025-06-17 15:16:19,692 | INFO     | 140640202918784[Task-112] | slack_app       | Health check.
INFO:     15.204.221.39:29104 - "GET /health_check HTTP/1.1" 200 OK
2025-06-17 15:16:20,694 | INFO     | 139680381873024[Task-126] | slack_app       | Health check.
INFO:     15.204.221.39:29108 - "GET /health_check HTTP/1.1" 200 OK
2025-06-17 15:16:21,696 | INFO     | 140640202918784[Task-113] | slack_app       | Health check.
INFO:     15.204.221.39:25746 - "GET /health_check HTTP/1.1" 200 OK
2025-06-17 15:16:22,698 | INFO     | 139680381873024[Task-127] | slack_app       | Health check.
INFO:     15.204.221.39:25752 - "GET /health_check HTTP/1.1" 200 OK
2025-06-17 15:16:23,701 | INFO     | 140640202918784[Task-114] | slack_app       | Health check.
INFO:     15.204.221.39:25768 - "GET /health_check HTTP/1.1" 200 OK
2025-06-17 15:16:24,703 | INFO     | 140640202918784[Task-115] | slack_app       | Health check.

Screenshot from 2025-06-17 12-14-07
Screenshot from 2025-06-17 12-14-07814×331 23.9 KB

2

Hey,

During startup, the Interval is ignored and Koyeb will run health checks very often so we can transition the Deployment to Healthy as fast as possible.
Once the Deployment is stable, the specified Interval will be used instead.

Hope that makes sense!

3

Ah, yeah, that makes sense. Thanks for clearing that up for me. Out of curiosity, for how long does it usually keep doing it? My service is showing a healthy status and it has been up for 35 minutes.

4

Hey!

Indeed, that’s longer than expected. It seems that you’re hitting an issue where we run that update later than it should.

We’re looking into it