BentoML

<@U03NQ3FCY6A> if you can help me again with this, it would be much appreciated :pray:

`resource "aws_instance" "app_server" {`
  `launch_template {`
    `id = aws_launch_template.lt.id`
  `}`
  `provisioner "local-exec" {`
    `command = &lt;&lt;-EOT`
        `attempt_counter=0`
        `max_attempts=80`
        `printf 'waiting for server to start'`
        `until $(curl --output /dev/null --silent --head --fail http://${self.public_ip}); do`
            `if [ $attempt_counter -eq $max_attempts ];then`
              `echo "Max attempts reached"`
              `exit 1`
            `fi`

            `printf '.'`
            `attempt_counter=$(($attempt_counter+1))`
            `sleep 15`
        `done`
        `EOT`
  `}`
`}`

It's a script constantly checks if the instance has been launched and the service is working

Yes, and the instance is launched and becomes active but the service fails to start

it's giving up after 80 tries when it cannot curl the public IP

And it's weird because it runs locally with docker

can u verify that
```http://${self.public_ip})```
is accessible? do you need to specify a port? (like `3000` or whatever)

i'm not too sure about AWS + Terraform but what if u leave out the `curl` command in the first place? will the service still start?

The thing is that it was working before like this

```http://${self.public_ip})```
is accessible but the service is not running

I'll try to get logs into the runners / api servers to see if there's any obvious errors

hmm. are the EC2 instances already spun up?

you can probably log into the EC2 console and check

<https://github.com/bentoml/aws-ec2-deploy#connect-to-the-ec2-instance>

Ok, just started the whole build process again and ended up with the same result.. The instance is created and running but the service fails to start

Done, just found the logs and it says `error processing tar file: write .....: no space left on device`

Now i need to find out how to change the deployment_config to launch the instance with a bigger volume

Do you know if there's any configuration option that defines the volume size before running bentoctl apply?

Or do i have to modify the terraform script to create one?

Thank you so much again for helping! :pray:

The problem was the volume size. We were missing 3gb on the new instance and i had to play with the terraform files

But i learned how to check the log files now on the new instance at least, haha :joy: thank you!