Getting 502 on healthy deployments

Multiple deployments set up with blueprint are deploying successfully, but are not available. Both the url and my own custom domain fails.

Please help

Hi @franzwilhelm,

We’ve applied a fix here and your services should be available. We’ll be investigating why this happened, and taking steps to prevent it from happening again. Very sorry for the disruption here!


You did manage to fix the existing issue, but it seems to me the root cause has not been found as the issue is back again, this time on another service.

We’ve setup an external health check on ( to monitor the health so we know when the issue has been resolved on your end. The render dashboard health check must be calling the internal url, not the public facing one, and this should be considered a major bug.

1 Like

Now suddenly (without any notice from the render team), the service is healthy again. The total length of the outage since we setup an external health check was 18 hours and 15 minutes. We’ve changed nothing on our end

Hi Franz, I’m sorry about this. We’re digging into the root cause and will let you know ASAP.


Same error for few days already.
Error 502 after deployment. healthy server.
Just come and goes.

Current error id: Request ID: 6ed31ab32d238b51-SEA

We’ve just deployed a fix for this issue. We’ll share more details soon, but there was a specific case when the first build for a new service failed, and it prevented a necessary networking component from being created. Our systems now ensure the creation of the network components regardless of the initial build status.

If you’re still running into 502s please share the service ID or URL here and we’ll dig further.

still getting the same issue after new deployment
Request ID: 6ed34ad0c82f3c2e-SEA

Hi @hpannecoucke,

Are these 502s happening for you primarily around deploys? I noticed that your service has a disk attached, note that disks do prevent zero downtime deploys, so some amount of brief 502s are expected during new deployments. More info on that here:

If you require zero downtime deploys, you may want to look into using the Strapi with Postgresql option instead:

If you’re seeing different behavior with the errors, unrelated to deploy downtime, please let me know.

Seems like your fix didn’t work…

Issue back again after new blueprint service created that failed the first time. Check out request id: 6f09621e30941c0e-FRA and 6f0963e1f0fa0b49-FRA for instance

And now suddenly it gave me 403 Forbidden from Cloudflare:

Hi @franzwilhelm,

Really sorry that you’re still running into issues here. We’re taking a look.

I decided to delete and recreate my deployment. The id of the old service was srv-c8tmjb7d17c33hbb41p0, for you to investigate further. After deleting and recreating, my service magically works again…

I’m getting 403s as well for my app:
Sometimes it works, sometimes it doesn’t, without any new deployments…

@franzwilhelm Does it work for you reliably now that you deleted and recreated your deployment?
Did you delete the whole web service on Render and created it from scratch?

Ok, for now it seems to work from all subdomains without any 403s…

Yes I did delete it and recreate it. As Render’s team has already identified, this mainly happens when a service fails on its first deployment. Even though it gets successfully deployed after that, the load balancer is badly configured on Render’s side, meaning you’re helpless. The only reason to get it working again is to wait for the render team to manually reconfigure the load balancer for you, or delete and recreate the service. Render has no monitoring in house on this (health checks only checks internal urls), so you have to contact them .

Alright, thanks for the info!

I am also having this issue. My server seems to be okay I can interact with it through the shell and my code works locally, but when I deploy every page on my website returns a 502 error.

Render URL:
Request ID: 70cd56a909259864-SEA
Website URL: