Example Redis deployment doesn't work on 1.3.0 without consul

[mr-karan]

Nomad version

Output from nomad version

Nomad v1.3.0 (52e95d64113e01be05d585d8b4c07f6f19efebbc)

Operating system and Environment details

uname -a                         
Linux pop-os 5.17.5-76051705-generic #202204271406~1651504840~22.04~63e51bd SMP PREEMPT Mon May 2 15: x86_64 x86_64 x86_64 GNU/Linux

lsb_release -a
No LSB modules are available.
Distributor ID:	Pop
Description:	Pop!_OS 22.04 LTS
Release:	22.04
Codename:	jammy

Issue

The example deployment from nomad init doesn't seem to run anymore on a single node Nomad acting as server+client (and no consul).

Reproduction steps

1. Save the file as nomad.hcl

datacenter = "dc1"
data_dir   = "/opt/nomad/data"

log_level = "DEBUG"

bind_addr = "0.0.0.0"

server {
  enabled          = true
  bootstrap_expect = 1
}

client {
  enabled = true
}

plugin "docker" {
  config {
    allow_privileged = true
    volumes {
      enabled = true
    }
    extra_labels = ["job_name", "job_id", "task_group_name", "task_name", "namespace", "node_name", "node_id"]
  }
}

2. Run nomad with:

nomad agent -dev -config=nomad.hcl

3. Run the example file

nomad init
nomad run example.nomad

4. Visit the UI for failed alloc

Expected Result

As of 1.2.6 until yesterday this worked normally, so it should also continue to work on 1.3.0?

Actual Result

It seems to check whether consul is present on the client or not and errors out with Constraint ${attr.consul.version} semver >= 1.7.0 filtered 1 node

When I go to the client attributes list (at http://localhost:4646/ui/clients/c8927149-1161-f27c-7079-1adce6d56697), I don't see any consul block. Which makes sense because this consul block doesn't exist in the agent config file. The same issue happens even if I run nomad agent -dev so the agent config is not the issue here, I suppose.

Job file (if appropriate)

Same as nomad init

[mr-karan]

Ah, my bad. The example file contains a default service stanza without any provider. When I explicitly set provider=nomad, it works as expected. Should we update the default init file ?

[jrasell]

Hi @mr-karan and thanks for raising this.

This is an oversight on my part; Nomad 1.3 adds implicit constraints on task groups that utilise Consul service registration so they only get placed on nodes where Consul has been fingerprinted. I had not account for the non-short example job which includes this block.

I think updating the default file to set the provider to Nomad is the correct choice, however, i'll raise this internally to gather opinions also.

[github-actions]

I'm going to lock this issue because it has been closed for 120 days ⏳. This helps our maintainers find and focus on the active issues.
If you have found a problem that seems similar to this, please open a new issue and complete the issue template so we can capture all the details necessary to investigate further.