I need to know if my application is running on multiple instances
We are using NestJS CRON, but we started to receive two emails, it happened back on heroku because app was running on multiple dynos. I wanted to confirm is this the same case on railway?
23 Replies
When deploying there is a overlap where you will have two up at one time
Leme find the docs
There's this.
Just. Need to find the section about the deployment overlapp cus you can configure it via env vars
Hmmm I asume the overlap is still there even if your not using health check.
Cannot find the docs related to the overlap sorry π¦
Somone else might be able to.
Yeah I saw two deployments active I removed one now but my question now is shouldn't it automatically remove the old one.
It should yeah
Can any of the technical team member check and see what caused this?
This isn't necessarily true.
@Muhammad Farhan Tahir just fyi, there's a dedicated channel for you as you're on the teams plan. #chainlist
I will use that channel from now own.
You don't have to but I was just making sure you knew about it! π
Thanks, I actually wasn't aware of it.
But as I have opened up this post, let's continue here for this one.
As for the issue at hand, did this happen when you had two active deployments? And could you share the project ID with me?
Shall I share the project ID here ? and no I am not exactly sure if it happened during two active deployments, I am trying to reproduce but I remember it happened before on heroku and we had two dynos running which was causing it.
Yeah, you can share the project ID. Permissions are scoped to the user.
So no one will be able to do anything with just the ID.
4cf41ad7-2e62-4a26-80c0-0693d49c57d4
And was this in the production environment or staging?
staging
We haven't made our prod active yet as we are further exploring railway though liking every bit of it.
By the way, I just reproduced it and no its not tied to multiple active deployments.
as right on API side I have one active deployment.
You have an active deployment in the production environment from 5 days ago. Any chance that's causing the issue?
No as we use redis for queuing and staging has separate instance.
Right. Let me raise this internally and see if someone can look into why it may have happened.
I verified env variables as well.
Thanks
It might have taken long time for us to reconcile the active deploys, I am seeing only two deploys live. Are you still observing this behavior?
Yes.
Anything to share on this one?
Forgive me! It was a hectic day with the service degredations, I am going to investigate this further.