Hello,

Unfortunately we are unable to provide code support, I would recommend adding verbose logging and error handling so you can find the root of the instabilities.

Thank you, I know you not provide code support!

Just wondering looking at logs can you see if the resources are ok? Do I need share the here something? logs screenshot or project id?

Looking at your metrics graphs for the service, you have never even remotely went close to using your max allowed resecures.

Yeah, odd not sure why it stops responding when the app auth fails. wondering if the database drop the connection somehow.

You are hardcoding the database URL, please use reference variables to reference the private URL.

https://docs.railway.app/guides/variables#referencing-another-services-variable

Is the same issue, I hardcoded to see if that helps.

Please go back to referencing the private URL.

Like this? (attachment )

Looks good.

FWIW Service names aren't sensitive information.

Thank you for you support! Everthing else look ok to you?

From the Railway side of things, yep!

brody, why should the use reference variables to reference the private URL make a difference?

It is a best practice, using the private network to connect to your database means no database <--> service egress fees!

good, but I guess in terms of stability or the lake of them does not matter for our issue, looks like.

Yep, I would make sure to properly handle errors and add verbose logging so you can narrow down your issue.

brody,

It seems like the connection between nodejs and postgresql sometimes is lost, or it is postgresql that have issues... not 100% sure yet.

But today I restarted postgresql and the app back to work again.

Every day I have to restart our nodejs 4 or more times, today I just restarted postgresql to see what If that work.

I will continue monitor the issue. Do you mind to check if out postgresql is been ok?

Hello,

There's nothing wrong with the database from its logs.

https://docs.strapi.io/dev-docs/configurations/database#database-pooling-options

Please take a look at the Caution banner here, perhaps this is your issue.

I've also looked at the probes for the private network and there are no issues there, so whatever this is, it would not be a platform error.

Thank you, link you shared is from strapi any reason?

Please see the sentence that followed the link for the reasoning..

> https://docs.strapi.io/dev-docs/configurations/database#database-pooling-options

> Please take a look at the Caution banner here, perhaps this is your issue.

okay just wanted to make sure you that you are not thinking that we use strapi cms, because we use Payload cms. so I was confused. thank you.

unresponsive again... I don't get it.

Right but it could be a similar issue as they are mentioning in their docs.

already did 2 hours ago