ood_core 0.17.0 → 0.17.5

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 266e24cd2e11c2c712b6128bdc3f82bf9781ae9885f8f0eb21439cb80c889c90
-  data.tar.gz: 01e682b6313468371076cdfd4ff2df2f3c06c661af9b4d7d7a65b7dcf3e2d836
+  metadata.gz: 88f0fb9eed8a53308dd19e73929e3c7af71ad1295738d69e0bf34d551a2ef0b3
+  data.tar.gz: 6437344171900c7b289ee3cf658550f050f95435e46b75eb1e34891e6ec9af08
 SHA512:
-  metadata.gz: 22721c9d368ec44533d93914f977576ee77786ad0926976fa24067c7f353104edd32baffac34723b730e7f711c5b7581cf4f72f6232d050c719edcfc1b3cb14f
-  data.tar.gz: 5d99a3c782aad5e420333653e51073cdb6535b7689053452da7ee7ef234292769e70d4010d751aacd4576daf531700136c2aa0ef7a2efc31edcb29e4c01c8be4
+  metadata.gz: b6aa7f5b93547e1c167870146f76766c70f2d26b19c706fd757880b70973281466674c5791bf62eafd5297b6733f20dc8e37128b4715374d764eb6c85d2b2919
+  data.tar.gz: 3930e176306b5354d2f451d744c89be9f61a714c31e8a0e4dcf6fc680ad7e70af6eb8769675157c7f7d247258f33d705bd146ac50e22553f8e064898b2ebebe6

data/CHANGELOG.md

@@ -7,6 +7,50 @@ and this project adheres to [Semantic Versioning](http://semver.org/spec/v2.0.0.
 
 ## [Unreleased]
 
+## [0.17.5] - 8-20-2021
+
+### Fixed
+
+- kubernetes jobs delete without waiting in [314](https://github.com/OSC/ood_core/pull/314).
+
+## [0.17.4] - 7-29-2021
+
+Functionally the same as [0.17.3] but with some CI updates.
+
+## [0.17.3] - 7-29-2021
+
+### Fixed
+
+- Fixed handling of pods in a startup phase in [303](https://github.com/OSC/ood_core/pull/303).
+
+### Added
+
+- Enable automatic population of supplemental groups in [305](https://github.com/OSC/ood_core/pull/305).
+
+## [0.17.2] - 7-14-2021
+
+### Fixed
+
+- Fixed k8s adapter to only show Running pods as running in [300](https://github.com/OSC/ood_core/pull/300).
+
+## [0.17.1] - 6-14-2021
+
+### Fixed
+
+- Fixed [278](https://github.com/OSC/ood_core/pull/278) where unschedulable pods will now show up as
+  queued_held status.
+
+### Changed
+
+- KUBECONFIG now defaults to /dev/null in the kubernetes adapter in [292](https://github.com/OSC/ood_core/pull/292).
+
+### Added
+
+- Sites can now set `batch_connect.ssh_allow` on the cluster to disable the buttons to start
+  a shell session to compute nodes in [289](https://github.com/OSC/ood_core/pull/289).
+- `POD_PORT` is now available to jobs in the kubernetes adapter in [290](https://github.com/OSC/ood_core/pull/290).
+- Kubernetes pods now support a startProbe in [291](https://github.com/OSC/ood_core/pull/291).
+
 ## [0.17.0] - 5-26-2021
 
 ### Fixed
@@ -336,7 +380,12 @@ and this project adheres to [Semantic Versioning](http://semver.org/spec/v2.0.0.
 ### Added
 - Initial release!
 
-[Unreleased]: https://github.com/OSC/ood_core/compare/v0.16.0...HEAD
+[Unreleased]: https://github.com/OSC/ood_core/compare/v0.17.5...HEAD
+[0.17.5]: https://github.com/OSC/ood_core/compare/v0.17.4...v0.17.5
+[0.17.4]: https://github.com/OSC/ood_core/compare/v0.17.3...v0.17.4
+[0.17.3]: https://github.com/OSC/ood_core/compare/v0.17.2...v0.17.3
+[0.17.2]: https://github.com/OSC/ood_core/compare/v0.17.1...v0.17.2
+[0.17.1]: https://github.com/OSC/ood_core/compare/v0.17.0...v0.17.1
 [0.17.0]: https://github.com/OSC/ood_core/compare/v0.16.1...v0.17.0
 [0.16.1]: https://github.com/OSC/ood_core/compare/v0.16.0...v0.16.1
 [0.16.0]: https://github.com/OSC/ood_core/compare/v0.15.1...v0.16.0

data/lib/ood_core/cluster.rb

@@ -148,12 +148,12 @@ module OodCore
     end
 
     # Whether this cluster supports SSH to batch connect nodes
-    # @return [Boolean] whether cluster supports SSH to batch connect node
+    # @return [Boolean, nil] whether cluster supports SSH to batch connect node
     def batch_connect_ssh_allow?
       return @batch_connect_ssh_allow if defined?(@batch_connect_ssh_allow)
-      return @batch_connect_ssh_allow = true if batch_connect_config.nil?
+      return @batch_connect_ssh_allow = nil if batch_connect_config.nil?
 
-      @batch_connect_ssh_allow = batch_connect_config.fetch(:ssh_allow, true)
+      @batch_connect_ssh_allow = batch_connect_config.fetch(:ssh_allow, nil)
     end
 
     # The comparison operator

data/lib/ood_core/job/adapters/kubernetes/batch.rb

@@ -14,6 +14,7 @@ class OodCore::Job::Adapters::Kubernetes::Batch
   attr_reader :config_file, :bin, :cluster, :mounts
   attr_reader :all_namespaces, :using_context, :helper
   attr_reader :username_prefix, :namespace_prefix
+  attr_reader :auto_supplemental_groups
 
   def initialize(options = {})
     options = options.to_h.symbolize_keys
@@ -25,6 +26,7 @@ class OodCore::Job::Adapters::Kubernetes::Batch
     @all_namespaces = options.fetch(:all_namespaces, false)
     @username_prefix = options.fetch(:username_prefix, '')
     @namespace_prefix = options.fetch(:namespace_prefix, '')
+    @auto_supplemental_groups = options.fetch(:auto_supplemental_groups, false)
 
     @using_context = false
     @helper = OodCore::Job::Adapters::Kubernetes::Helper.new
@@ -123,7 +125,7 @@ class OodCore::Job::Adapters::Kubernetes::Batch
       when "get"
         call_json_output('get', resource, id)
       when "delete"
-        call("#{namespaced_cmd} delete #{resource} #{id}")
+        call("#{namespaced_cmd} delete #{resource} #{id} --wait=false")
       end
     rescue NotFoundError
       {}
@@ -176,6 +178,19 @@ class OodCore::Job::Adapters::Kubernetes::Batch
     Etc.getgrgid(run_as_group).name
   end
 
+  def default_supplemental_groups
+    OodSupport::User.new.groups.sort_by(&:id).map(&:id).reject { |id| id < 1000 }
+  end
+
+  def supplemental_groups(groups = [])
+    sgroups = []
+    if auto_supplemental_groups
+      sgroups.concat(default_supplemental_groups)
+    end
+    sgroups.concat(groups.to_a)
+    sgroups.uniq.sort
+  end
+
   def default_env
     {
       USER: username,
@@ -183,6 +198,7 @@ class OodCore::Job::Adapters::Kubernetes::Batch
       HOME: home_dir,
       GROUP: group,
       GID: run_as_group,
+      KUBECONFIG: '/dev/null',
     }
   end
 
@@ -190,6 +206,7 @@ class OodCore::Job::Adapters::Kubernetes::Batch
   # create an id.
   def generate_id_yml(script)
     native_data = script.native
+    native_data[:container][:supplemental_groups] = supplemental_groups(native_data[:container][:supplemental_groups])
     container = helper.container_from_native(native_data[:container], default_env)
     id = generate_id(container.name)
     configmap = helper.configmap_from_native(native_data, id, script.content)

data/lib/ood_core/job/adapters/kubernetes/helper.rb

@@ -54,7 +54,9 @@ class OodCore::Job::Adapters::Kubernetes::Helper
       working_dir: container[:working_dir],
       restart_policy: container[:restart_policy],
       image_pull_policy: container[:image_pull_policy],
-      image_pull_secret: container[:image_pull_secret]
+      image_pull_secret: container[:image_pull_secret],
+      supplemental_groups: container[:supplemental_groups],
+      startup_probe: container[:startup_probe],
     )
   end
 
@@ -149,7 +151,7 @@ class OodCore::Job::Adapters::Kubernetes::Helper
     {
       id: json_data.dig(:metadata, :name).to_s,
       job_name: name_from_metadata(json_data.dig(:metadata)),
-      status: pod_status_from_json(json_data),
+      status: OodCore::Job::Status.new(state: pod_status_from_json(json_data)),
       job_owner: job_owner_from_json(json_data, ns_prefix),
       submission_time: submission_time(json_data),
       dispatch_time: dispatch_time(json_data),
@@ -239,15 +241,21 @@ class OodCore::Job::Adapters::Kubernetes::Helper
   def submission_time(json_data)
     status = json_data.dig(:status)
     start = status.dig(:startTime)
+    creation = json_data.dig(:metadata, :creationTimestamp)
 
-    if start.nil?
+    if !creation.nil?
+      str = creation
+    elsif !start.nil?
+      str = start
+    else
       # the pod is in some pending state limbo
       conditions = status.dig(:conditions)
+      return nil if conditions.nil?
+      return nil if conditions.size == 0
       # best guess to start time is just the first condition's
       # transition time
       str = conditions[0].dig(:lastTransitionTime)
-    else
-      str = start
+      return nil if str.nil?
     end
 
     DateTime.parse(str).to_time.to_i
@@ -255,11 +263,21 @@ class OodCore::Job::Adapters::Kubernetes::Helper
 
   def pod_status_from_json(json_data)
     phase = json_data.dig(:status, :phase)
+    conditions = json_data.dig(:status, :conditions)
+    container_statuses = json_data.dig(:status, :containerStatuses)
+    unschedulable = conditions.to_a.any? { |c| c.dig(:reason) == "Unschedulable" }
+    ready = !container_statuses.to_a.empty? && container_statuses.to_a.all? { |s| s.dig(:ready) == true }
+    started = !container_statuses.to_a.empty? && container_statuses.to_a.any? { |s| s.fetch(:state, {}).key?(:running) }
+    return "running" if ready
+    return "queued" if phase == "Running" && started
+
     state = case phase
-            when "Running"
-              "running"
             when "Pending"
-              "queued"
+              if unschedulable
+                "queued_held"
+              else
+                "queued"
+              end
             when "Failed"
               "suspended"
             when "Succeeded"
@@ -269,8 +287,6 @@ class OodCore::Job::Adapters::Kubernetes::Helper
             else
               "undetermined"
             end
-
-    OodCore::Job::Status.new(state: state)
   end
 
   def terminated_state(status)

data/lib/ood_core/job/adapters/kubernetes/resources.rb

@@ -33,13 +33,36 @@ module OodCore::Job::Adapters::Kubernetes::Resources
     end
   end
 
+  class TCPProbe
+    attr_accessor :port, :initial_delay_seconds, :failure_threshold, :period_seconds
+
+    def initialize(port, data)
+      data ||= {}
+      @port = port
+      @initial_delay_seconds = data[:initial_delay_seconds] || 2
+      @failure_threshold = data[:failure_threshold] || 5
+      @period_seconds = data[:period_seconds] || 5
+    end
+
+    def to_h
+      {
+        port: port,
+        initial_delay_seconds: initial_delay_seconds,
+        failure_threshold: failure_threshold,
+        period_seconds: period_seconds,
+      }
+    end
+  end
+
   class Container
     attr_accessor :name, :image, :command, :port, :env, :memory, :cpu, :working_dir,
-                  :restart_policy, :image_pull_policy, :image_pull_secret, :supplemental_groups
+                  :restart_policy, :image_pull_policy, :image_pull_secret, :supplemental_groups,
+                  :startup_probe
 
     def initialize(
         name, image, command: [], port: nil, env: {}, memory: "4Gi", cpu: "1",
-        working_dir: "", restart_policy: "Never", image_pull_policy: nil, image_pull_secret: nil, supplemental_groups: []
+        working_dir: "", restart_policy: "Never", image_pull_policy: nil, image_pull_secret: nil, supplemental_groups: [],
+        startup_probe: {}
       )
       raise ArgumentError, "containers need valid names and images" unless name && image
 
@@ -55,6 +78,7 @@ module OodCore::Job::Adapters::Kubernetes::Resources
       @image_pull_policy = image_pull_policy.nil? ? "IfNotPresent" : image_pull_policy
       @image_pull_secret = image_pull_secret
       @supplemental_groups = supplemental_groups.nil? ? [] : supplemental_groups
+      @startup_probe = TCPProbe.new(@port, startup_probe)
     end
 
     def ==(other)
@@ -69,7 +93,8 @@ module OodCore::Job::Adapters::Kubernetes::Resources
         restart_policy == other.restart_policy &&
         image_pull_policy == other.image_pull_policy &&
         image_pull_secret == other.image_pull_secret &&
-        supplemental_groups == other.supplemental_groups
+        supplemental_groups == other.supplemental_groups &&
+        startup_probe.to_h == other.startup_probe.to_h
     end
   end
 

data/lib/ood_core/job/adapters/kubernetes/templates/pod.yml.erb

@@ -20,14 +20,7 @@ spec:
     runAsUser: <%= run_as_user %>
     runAsGroup: <%= run_as_group %>
     runAsNonRoot: true
-    <%- if spec.container.supplemental_groups.empty? -%>
-    supplementalGroups: []
-    <%- else -%>
-    supplementalGroups:
-    <%- spec.container.supplemental_groups.each do |supplemental_group| -%>
-    - "<%= supplemental_group %>"
-    <%- end -%>
-    <%- end -%>
+    supplementalGroups: <%= spec.container.supplemental_groups %>
     fsGroup: <%= fs_group %>
   hostNetwork: false
   hostIPC: false
@@ -52,6 +45,10 @@ spec:
       valueFrom:
         fieldRef:
           fieldPath: metadata.namespace
+    <%- unless spec.container.port.nil? -%>
+    - name: POD_PORT
+      value: "<%= spec.container.port %>"
+    <%- end -%>
     <%- spec.container.env.each_pair do |name, value| -%>
     - name: <%= name %>
       value: "<%= value %>"
@@ -65,6 +62,12 @@ spec:
     <%- unless spec.container.port.nil? -%>
     ports:
     - containerPort: <%= spec.container.port %>
+    startupProbe:
+      tcpSocket:
+        port: <%= spec.container.startup_probe.port %>
+      initialDelaySeconds: <%= spec.container.startup_probe.initial_delay_seconds %>
+      failureThreshold: <%= spec.container.startup_probe.failure_threshold %>
+      periodSeconds: <%= spec.container.startup_probe.period_seconds %>
     <%- end -%>
     <%- if !all_mounts.empty? || (!configmap.nil? && configmap.mounts?) -%>
     volumeMounts:

data/lib/ood_core/version.rb

@@ -1,4 +1,4 @@
 module OodCore
   # The current version of {OodCore}
-  VERSION = "0.17.0"
+  VERSION = "0.17.5"
 end

data/ood_core.gemspec

@@ -15,7 +15,7 @@ Gem::Specification.new do |spec|
   spec.license       = "MIT"
 
   spec.files         = `git ls-files -z`.split("\x0").reject do |f|
-    f.match(%r{^(test|spec|features)/})
+    f.match(%r{^(test|spec|features|.github)/})
   end
   spec.bindir        = "exe"
   spec.executables   = spec.files.grep(%r{^exe/}) { |f| File.basename(f) }

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: ood_core
 version: !ruby/object:Gem::Version
-  version: 0.17.0
+  version: 0.17.5
 platform: ruby
 authors:
 - Eric Franz
@@ -10,7 +10,7 @@ authors:
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2021-05-26 00:00:00.000000000 Z
+date: 2021-08-20 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: ood_support
@@ -160,7 +160,6 @@ executables: []
 extensions: []
 extra_rdoc_files: []
 files:
-- ".github/workflows/test.yml"
 - ".gitignore"
 - ".rspec"
 - CHANGELOG.md

data/.github/workflows/test.yml

@@ -1,30 +0,0 @@
-name: Unit Tests
-
-on:
-  push:
-    branches:
-      - master
-  pull_request:
-    branches:
-      - master
-
-jobs:
-  tests:
-    runs-on: ubuntu-latest
-
-    steps:
-      - name: checkout
-        uses: actions/checkout@v2
-
-      - name: Setup Ruby using Bundler
-        uses: ruby/setup-ruby@v1
-        with:
-          ruby-version: "2.7.1"
-          bundler-cache: true
-          bundler: "2.1.4"
-
-      - name: install gems
-        run: bundle install
-
-      - name: test
-        run: bundle exec rake spec