Osquery uses basic SQL commands to leverage a relational data-model to describe a device.

osquery

image.png

Hey all :wave:
I'm trying to deploy fleet, but i cannot complete the setup.
I get the following error:
&gt; We were unable to configure Fleet. If your Fleet server is behind a proxy, please ensure the server can be reached.
Can't find anything inside the logs. What can i do to troubleshoot/fix this?

Greetings <@U07FRRTT09J>!
What is your setup or infrastructure like? How are you deploying Fleet?

on <http://fly.io|fly.io> with the fleetdm/fleet:latest image + mysql + redis + s3

```app = 'fleet'
primary_region = 'fra'

[build]
[deploy]
  release_command = "fleet prepare db"

[env]
  FLEET_MYSQL_ADDRESS = '[...]:3306'
  FLEET_MYSQL_DATABASE = 'fleet'
  FLEET_MYSQL_USERNAME = 'fleet'
  FLEET_REDIS_ADDRESS = '[...]:6379'
  FLEET_REDIS_USERNAME = 'default'
  FLEET_SERVER_TLS = 'false'
  FLEET_S3_SOFTWARE_INSTALLERS_BUCKET = 'fleet'
  FLEET_LOGGING_DEBUG = 'true'
  FLEET_DOMAIN = 'fleet.[...]'

[http_service]
  internal_port = 8080
  force_https = true
  auto_stop_machines = 'off'
  auto_start_machines = true
  min_machines_running = 1
  processes = ['app']

  [[http_service.checks]]
    interval = '30s'
    timeout = '10s'
    grace_period = '20s'
    method = 'GET'
    path = '/healthz'

[[vm]]
  memory = '1gb'
  cpu_kind = 'shared'
  cpus = 1```

i have put placeholders here, but the actual domain is properly set up with correct certificate too

```FROM fleetdm/fleet:latest
CMD fleet serve```
copied the Dockerfile and fly.toml from somewhere in here, slightly adjusted the fly.toml (release command and some ENV vars)

Hey <@U07FRRTT09J>, thanks for trying this out!
Asking around with the crew here and not too familiar with <http://fly.io|fly.io>. We're not able to support the deployment directly. My hunch is there are some missing "wrapper" configurations. Can attempt to convert from our <https://github.com/fleetdm/fleet/blob/main/render.yaml|Render Blueprint.> Guessing you're trying to get Fleet going and recommend our <https://fleetdm.com/docs/deploy/deploy-fleet#render|Deployment Page> (Render and AWS Terraform). We'll keep this in mind though and if anyone finds something, we'll be sure to update.

As a workaround, could attempt to use <https://fleetdm.com/docs/using-fleet/fleetctl-cli#get-more-info-about-a-command|fleetctl> and doing `fleetctl setup`.

Thanks! I used fleetctl setup and it threw an actual error:
&gt; {
&gt;   "message": "Validation Failed",
&gt;   "errors": [
&gt;     {
&gt;       "name": "Could not create user. Password is over the 48 characters limit. If the password is under 48 characters, please check the auth_salt_key_size in your Fleet server config.",
&gt;       "reason": "password too long"
&gt;     }
&gt;   ],
&gt;   "uuid": "1a6b885c-3c0a-45a6-8150-f8a1811527d4"
&gt; }
&gt; Error: error setting up Fleet: setup received status 422 Validation Failed: password too long

I still think this kind of error should be propagated to the UI, or at least be logged