Canary deployment is spawning wrong machines (region and memory)

[nicolas-besnard]

I’m testing the Canary release for my app, and it’s not working as expected.

1. Machines created are not in the original region
2. Machines created are using less memory than the original one

My deployment is failing because the new machine doesn’t have enough memory

Is anyone using Canary and successfully deployed something? By looking at the code that implements this feature, I have the feeling that it can be caused by my app only having 1 machine

[kurt]

Will you share your fly.toml? Is there a [processes] config there?

[nicolas-besnard]

No, I don’t have any [processes] defined

app = "app"

kill_signal = "SIGINT"
kill_timeout = 5
processes = []
primary_region = "ewr"

[deploy]
  strategy = "canary"

[experimental]
  allowed_public_ports = []
  auto_rollback = true

[[services]]
  internal_port = 8080
  processes = ["app"]
  protocol = "tcp"
  script_checks = []

  [services.concurrency]
    hard_limit = 25
    soft_limit = 20
    type = "connections"

  [[services.ports]]
    force_https = true
    handlers = ["http"]
    port = 80

  [[services.ports]]
    handlers = ["tls", "http"]
    port = 443

  [[services.tcp_checks]]
    grace_period = "1s"
    interval = "15s"
    restart_limit = 0
    timeout = "2s"

  # https://fly.io/docs/reference/configuration/#services-http_checks
  [[services.http_checks]]
    interval = "10s"
    grace_period = "5s"
    method = "get"
    path = "/healthcheck"
    protocol = "http"
    timeout = "2s"
    tls_skip_verify = false
    [services.http_checks.headers]