Allocations not becoming 'Healthy' for Services when using health_check = "checks"

[Vincent]

Hello,

We're trying out the rolling updates and noticed that the new allocations are not reporting their health status as expected.

$ nomad alloc-status 20e7ab79

...

Latest Deployment

ID          = faa23fa7

Status      = running

Description = Deployment is running

Deployed

Task Group  Desired  Placed  Healthy  Unhealthy

vh          4        2       0        0

Allocations

ID        Node ID   Task Group  Version  Desired  Status    Created At

20e7ab79  4d9c2742  vh          34       run      running   08/17/17 16:06:32 EDT

40a6a05f  b7132742  vh          34       run      running   08/17/17 16:06:32 EDT

f293f968  17af2742  vh          33       run      running   08/17/17 16:04:21 EDT

741ecd6d  f6922742  vh          33       run      running   08/17/17 16:04:21 EDT

46b28ac0  4d9c2742  vh          33       stop     complete  08/17/17 16:03:45 EDT

d7ba846e  b7132742  vh          33       stop     complete  08/17/17 16:03:45 EDT

$ nomad alloc-status 20e7ab79

...

ID                  = 20e7ab79

Node ID             = 4d9c2742

Job ID              = vh

Job Version         = 34

Client Status       = running

Client Description  = <none>

Desired Status      = run

Created At          = 08/17/17 16:06:32 EDT

Deployment ID       = faa23fa7

Deployment Health   = unset

The new allocations will eventually become 'unhealthy' after it exceeds healthy_deadline.

Our allocation tasks are all running and service checks are healthy (validated through Consul that they're passing!) but the alloc-status still shows "Deployment Health   = unset"

      update {

        max_parallel     = 2

        min_healthy_time = "10s"

        healthy_deadline = "5m"

        health_check     = "checks"

      }

      service {

        name = "my service name"

        port = "http"

          tags = [

              "urlpublic-frontend-${NOMAD_META_env}.domain.net/",

              "${NOMAD_META_env}",

              "version-${NOMAD_META_version}",

              "git-commit-${NOMAD_META_git_commit_id}"]

        check {

          name = "health endpoint"

          type = "http"

          port = "management"

          path = "/health"

          interval = "10s"

          timeout = "30s"

        }

        check {

          name = "app landing page"

          type = "http"

          port = "http"

          path = "/"

          interval = "10s"

          timeout = "30s"

        }

      }

Using health_check = "task_states" does work as expected!

Is there something we're missing or failing to set for our service checks?

Thanks,

VH

[Alex Dadgar]

Hey Vincent,

You are getting hit by a bug that has since been fixed. The fix will find its way into Nomad 0.6.1 which should be released soon. Sorry about that!

Thanks,
Alex Dadgar

--
This mailing list is governed under the HashiCorp Community Guidelines - https://www.hashicorp.com/community-guidelines.html. Behavior in violation of those guidelines may result in your removal from this mailing list.
 
GitHub Issues: https://github.com/hashicorp/nomad/issues
IRC: #nomad-tool on Freenode
---
You received this message because you are subscribed to the Google Groups "Nomad" group.
To unsubscribe from this group and stop receiving emails from it, send an email to nomad-tool+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/nomad-tool/80cdffb0-3973-49f7-8abf-6e7b1a7de35c%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Vincent]

Ah!! Thanks for the reply! I'll start keeping a closer eye on the issues.

VH