Duplicated index in allocation name

[pznamensky]

Nomad version

Nomad v0.9.4 (a81aa84)

Operating system and Environment details

CentOS Linux release 7.7.1908 (Core)
docker-ce-18.03.1

Issue

Similar to #3698
After updating a nomad job some allocations got the same name.

nomad-bm alloc -json | jq '.[] | .Name + " - " + .TaskStates[].State' | grep app | sort | uniq -c
      1 "app.app[0] - running"
      1 "app.app[10] - running"
      1 "app.app[11] - running"
      2 "app.app[1] - running" // duplicate name
      1 "app.app[2] - running"
      2 "app.app[3] - running" // duplicate name
      1 "app.app[4] - running"
      2 "app.app[5] - running" // duplicate name
      1 "app.app[6] - running"
      2 "app.app[7] - running" // duplicate name
      1 "app.app[8] - running"
      1 "app.app[9] - running"

Reproduction steps

1. create a job
2. make changes in nomad file
3. run nomad run with that changes
4. check allocation names

Job file (if appropriate)

job "app" {
  datacenters = ["dc1"]
  type = "service"

  update {
    max_parallel = 2
    min_healthy_time = "10s"
    healthy_deadline = "3m"
    canary = 12
    auto_promote = true
  }
  group "app" {
    count = 16
    restart {
      attempts = 10
      interval = "5m"
      delay = "20s"
      mode = "delay"
    }
    task "app" {
      driver = "docker"
      config {
        image = "app:master"
      }
      resources {
        network {
          port "http" {}
        }
      }
      service {
        name = "app"
        port = "http"
        check {
          type     = "tcp"
          interval = "10s"
          timeout  = "5s"
        }
      }
    }
  }
}

[tgross]

Hi @pznamensky! The allocation indexes aren't guaranteed to be unique. The servers make a best-effort attempt to make this value unique from 0 to count-1 but when you have canaries on the task group, there are reused indexes since you have multiple job versions running at the same time.

If you need a unique number for interpolating a value in your application, you can get this by combining the job version and the alloc index (which together should be unique).

[tgross]

I'm going to close this issue as a duplicate of #5879.

[pznamensky]

@tgross thanks for the explanation!

[pznamensky]

@tgross, sorry for coming back to it, but what is the purpose of ALLOC INDEX? Since there is no guarantees it is unique, probably it would be reasonable to delete it at all. What do you think?

[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.