fix: slurm host system access

made cobald be able to run slurm jobs, previously failing with
permission denied.

.gitignore

@@ -1,2 +1,3 @@
 .*.swp
 *.retry
+vars_auth.yml

base.yml

@@ -0,0 +1,27 @@
+---
+- hosts: all
+  tasks:
+  - name: "install epel repo" # for htop etc.
+    yum:
+      name: epel-release
+      state: present
+
+  - name: "install tools"
+    yum:
+      name: [ vim-enhanced, htop, screen, bind-utils, nmap-ncat, net-tools ]
+      state: present
+
+  - name: "screenrc native scrolling in tmux"
+    copy:
+      content: "termcapinfo xterm* ti@:te@\ntermcapinfo screen* ti@:te@\n"
+      dest: "{{item}}"
+    with_items:
+    - "~{{unpriv_user}}/.screenrc"
+    - "~root/.screenrc"
+
+  - name: "install ssh-key"
+    authorized_key:
+      user: "{{cfg_unpriv_user}}"
+      key: "{{cfg_ssh_key}}"
+      state: present
+

cobald.yml

@@ -0,0 +1,61 @@
+---
+- hosts: slurm, cobald
+  vars:
+    container_privileged: True
+    slurm_num_nodes: 10
+  tasks:
+  - name: "setup docker"
+    import_role: name=docker
+    tags: docker
+
+  - name: "get facts from existing cobald instance (i.e. hostname)"
+    include_role:
+      name: cobald
+      tasks_from: facts
+      apply:
+        tags: slurm, cobald, slurm-config
+    tags: slurm, cobald, slurm-config
+    vars:
+      container_name: cobald
+
+  - name: "setup slurm test environment in docker containers"
+    include_role:
+      name: slurm
+      apply:
+        tags: slurm
+    vars:
+      slurm_user: slurm # or root
+      slurm_user_accounts:
+      - name: cobald
+        dir: /var/lib/cobald
+      num_nodes: "{{slurm_num_nodes}}"
+      extra_nodes:
+      - name: cobald
+        hostname: "{{cobald_container_hostname}}" # from cobald/facts.yml above
+        # hostname is used as NodeHostname, which is used slurms "networking
+        # code" (https://bugs.schedmd.com/show_bug.cgi?id=8615).
+        # It works either way around, but one of NodeName or NodeHostname has
+        # to match the container name (-n flag, not --hostname) since when
+        # submitting tasks to the slurm controller, it matches access
+        # permissions against a reverse lookup of the submitting ip address.
+        # Docker always and unconfigureably resolves the container ip in any
+        # network to containername.netname, where containername is the
+        # containers runtime name (not hostname supplied!) and netname is
+        # the network name in host environment. We should run our own dns...
+      docker_network: slurm
+      slurm_hostsystem_cluster_access: True
+    when: '"slurm" in group_names'
+    tags: slurm, cobald, influxdb, slurm-config
+    # tags: cobald requires some slurm facts, so cobald tag is included here
+
+  - name: "install cobald"
+    include_role:
+      name: cobald
+      apply:
+        tags: cobald
+    vars:
+      cobald_slurm: True
+      container_name: cobald
+      # docker_network: slurm # overriden by vars/slurm.yml
+    when: '"cobald" in group_names'
+    tags: cobald, influxdb, singularity

htcondor.yml

@@ -0,0 +1,32 @@
+---
+- hosts: htcondor
+  tasks:
+  - name: "install htcondor repo"
+    yum:
+      name: https://research.cs.wisc.edu/htcondor/repo/8.9/htcondor-release-current.el7.noarch.rpm
+      state: present
+    tags: htcondor
+
+  - name: "install htcondor software "
+    yum:
+      name: htcondor-ce
+      state: present
+    tags: htcondor
+
+  - name: "remove minicondor configuration"
+    yum:
+      name: minicondor
+      state: absent
+    tags: htcondor
+
+  - name: "setup singularity"
+    import_role: name="singularity"
+    tags: singularity
+
+  - name: "setup docker"
+    import_role: name=docker
+    tags: docker
+
+  - name: "setup htcondor test environment in docker containers"
+    import_role: name=docker-htcondor
+    tags: htcondor-containered, htcondor

install.sh

@@ -0,0 +1,4 @@
+mkdir -p ./collections/ansible_collections/community
+# git clone --depth=1 -b 1.2.1 https://github.com/ansible-collections/community.grafana.git ./collections/ansible_collections/community/grafana
+git clone --depth=1 -b 1.2.1-extended https://github.com/ansible_community.grafana.git ./collections/ansible_collections/community/grafana
+

inv.yml

@@ -6,6 +6,14 @@ all:
       ssh_args: -o ControlMaster=auto -o ControlPersist=60s
       # ansible_host: 192.168.122.139
       unpriv_user: thoto
+      cfg_unpriv_user: thoto
+    ed-c7-2:
+      ansible_user: root
+      ansible_host: ed-c7-2.virt.uller.thoto.net
+      # ansible_host: 192.168.123.60 # + jumphost
+      ssh_args: -o ControlMaster=auto -o ControlPersist=60s
+      unpriv_user: thoto
+      cfg_unpriv_user: thoto
   children:
     htcondor:
       hosts:
@@ -13,6 +21,8 @@ all:
     slurm:
       hosts:
         ed-c7-1:
+        ed-c7-2:
     cobald:
       hosts:
         ed-c7-1:
+        ed-c7-2:

play.yml

@@ -1,62 +1,10 @@
 ---
-- hosts: all
+- name: base setup
-  tasks:
+  import_playbook: base.yml
-  - name: "install tools"
-    yum:
-      name: [ vim-enhanced, htop, screen, bind-utils, nmap-ncat, net-tools ]
-      state: present
 
-- hosts: htcondor
+- name: setup htcondor
-  pre_tasks:
+  import_playbook: htcondor.yml
-  - name: "install htcondor repo"
+  when: '"htcondor" in group_names'
-    yum:
-      name: https://research.cs.wisc.edu/htcondor/repo/8.9/htcondor-release-current.el7.noarch.rpm
-      state: present
-    tags: htcondor
 
-  - name: "install htcondor software "
+- name: setup slurm and cobald
-    yum:
+  import_playbook: cobald.yml
-      name: htcondor-ce
-      state: present
-    tags: htcondor
-
-  - name: "remove minicondor configuration"
-    yum:
-      name: minicondor
-      state: absent
-    tags: htcondor
-
-  - name: "setup singularity"
-    import_tasks: "singularity.yml"
-    tags: singularity
-
-  roles:
-  - name: "setup docker"
-    role: docker
-    tags: docker
-
-  - name: "setup htcondor test environment in docker containers"
-    role: docker-htcondor
-    tags:
-      - htcondor-containered
-      - htcondor
-
-- hosts: slurm
-  vars:
-    container_privileged: True
-    num_nodes: 3
-  roles:
-  - name: "setup docker"
-    role: docker
-    tags: docker
-  - name: "setup slurm test environment in docker containers"
-    role: slurm
-    vars:
-      slurm_user: slurm # or root
-    tags: slurm
-
-- hosts: cobald
-  roles:
-  - name: "install cobald"
-    role: cobald
-    tags: cobald

roles/cobald/defaults/main.yml

@@ -0,0 +1,6 @@
+cobald_domainname: cobald.local
+influx_admin_user: my-user
+influx_admin_pw: my-password
+influx_org: my-org
+influx_pubport: 28086
+influx_bucket: batleth

roles/cobald/files/28-sync-container-slurmd

@@ -0,0 +1,3 @@
+#!/bin/sh
+[ /slurm-singimage/slurmd.sif -nt /shared/slurmd.sif ] && \
+	cp /slurm-singimage/slurmd.sif /shared/slurmd.sif

roles/cobald/files/31-slurmd-configless

@@ -0,0 +1,3 @@
+#!/bin/sh
+slurmd --conf-server ${slurmctld} -D -N ${nodename} 2>/dev/null 1>/dev/null &
+tail --retry --pid $! -f ${SLURMD_LOG_PATH} ${SLURM_SCHED_LOG_PATH}

roles/cobald/files/cgroup.conf.noautomount

@@ -0,0 +1,11 @@
+###
+#
+# Slurm cgroup support configuration file
+#
+# See man slurm.conf and man cgroup.conf for further
+# information on cgroup configuration parameters
+#--
+CgroupAutomount=no
+
+ConstrainCores=no
+ConstrainRAMSpace=no

roles/cobald/files/cobald-config/config.yaml

@@ -0,0 +1,31 @@
+---
+pipeline:
+- __type__: cobald.controller.linear.LinearController
+  low_utilisation: 0.9
+  high_allocation: 0.9
+  rate: 0.10
+- !Limiter
+  minimum: 3
+- !TelegrafPipelineMonitor
+  poll: True
+- !TardisPoolFactory
+  configuration: /etc/cobald/tardis.yaml
+logging:
+  version: 1
+  root:
+    level: DEBUG
+    handlers: [console, file]
+  handlers:
+    console:
+      class: logging.StreamHandler
+      formatter: test
+      level: DEBUG
+      stream: ext://sys.stderr
+    file:
+      class: logging.handlers.RotatingFileHandler
+      formatter: test
+      level: WARNING
+      filename: /var/log/cobald/cobald-tardis.log
+  formatters:
+    test:
+      format: "  %(name)s %(message)s"

roles/cobald/files/cobald-config/tardis.yaml

@@ -0,0 +1,41 @@
+Plugins:
+  SqliteRegistry:
+    db_file: /tmp/drone_registry.db
+  TelegrafMonitoring:
+    host: ed-telegraf
+    port: 8094
+#BatchSystem:
+#  adapter: FakeBatchSystem
+#  allocation: 1.0
+#  utilisation: !PeriodicValue
+#    period: 60
+#    amplitude: 0.15
+#    offset: 0.80
+##    phase: 1.
+#    phase: 1.6
+#  machine_status: Available
+BatchSystem:
+  adapter: Slurm
+  max_age: 0.1
+  options:
+    partition: cobald
+Sites:
+- name: slurmtest
+  adapter: Slurm
+  quota: 20
+slurmtest:
+  # executor: ...
+  StatusUpdate: 0.1
+  MachineTypes:
+    - m1.a
+  MachineTypeConfiguration:
+    m1.a:
+      Walltime: 5
+      Partition: container
+      StartupCommand: /usr/local/bin/start-drone
+      # SubmitOptions: ...
+  MachineMetaData:
+    m1.a:
+      Cores: 3 # cores
+      Memory: 1 # GB
+      Disk: 4 # not passed

roles/cobald/files/cobald-entrypoint.sh

@@ -1,7 +1,7 @@
 #!/bin/sh
 
-[ -f /usr/local/lib/cobaldmodules/setup.py -a \
+for i in /usr/local/lib/entrypoints.d/* ; do
-	-d /usr/local/lib/cobaldmodules/cobaldmodules ] && \
+    [ -f $i ] && /bin/sh $i || break
-		pip3 install --no-deps --editable /usr/local/lib/cobaldmodules
+done
 
 exec "${@:-/bin/bash}"

roles/cobald/files/grafana-dashboard.json

@@ -0,0 +1,589 @@
+{
+  "annotations": {
+    "list": [
+      {
+        "builtIn": 1,
+        "datasource": "-- Grafana --",
+        "enable": true,
+        "hide": true,
+        "iconColor": "rgba(0, 211, 255, 1)",
+        "name": "Annotations & Alerts",
+        "type": "dashboard"
+      }
+    ]
+  },
+  "editable": true,
+  "gnetId": null,
+  "graphTooltip": 0,
+  "id": 1,
+  "iteration": 1623317629899,
+  "links": [],
+  "panels": [
+    {
+      "collapsed": false,
+      "datasource": null,
+      "gridPos": {
+        "h": 1,
+        "w": 24,
+        "x": 0,
+        "y": 0
+      },
+      "id": 4,
+      "panels": [],
+      "title": "Row title",
+      "type": "row"
+    },
+    {
+      "aliasColors": {},
+      "bars": false,
+      "dashLength": 10,
+      "dashes": false,
+      "datasource": "InfluxDB",
+      "fieldConfig": {
+        "defaults": {
+          "unit": "none"
+        },
+        "overrides": []
+      },
+      "fill": 1,
+      "fillGradient": 0,
+      "gridPos": {
+        "h": 9,
+        "w": 12,
+        "x": 0,
+        "y": 1
+      },
+      "hiddenSeries": false,
+      "id": 2,
+      "interval": "1s",
+      "legend": {
+        "avg": false,
+        "current": false,
+        "max": true,
+        "min": true,
+        "show": true,
+        "total": false,
+        "values": true
+      },
+      "lines": true,
+      "linewidth": 1,
+      "maxDataPoints": 200,
+      "nullPointMode": "null",
+      "options": {
+        "alertThreshold": true
+      },
+      "percentage": false,
+      "pluginVersion": "7.5.7",
+      "pointradius": 2,
+      "points": false,
+      "renderer": "flot",
+      "seriesOverrides": [],
+      "spaceLength": 10,
+      "stack": false,
+      "steppedLine": false,
+      "targets": [
+        {
+          "groupBy": [
+            {
+              "params": [
+                "$__interval"
+              ],
+              "type": "time"
+            },
+            {
+              "params": [
+                "null"
+              ],
+              "type": "fill"
+            }
+          ],
+          "orderByTime": "ASC",
+          "policy": "default",
+          "query": "f_r = (r, accumulator) => ({\n    _value: accumulator._value + (\n      if r._value == \"AvailableState\" then 1\n      else if r._value == \"DownState\" then -1\n      else 0)\n    })\n\nnodes = from(bucket: \"batleth\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> keep(columns: [\"tardis_machine_name\", \"_time\"])\n  |> sort(columns: [\"_time\"], desc: true)\n  |> unique(column: \"tardis_machine_name\")\n//  |> yield()\n\noffset = from(bucket: \"batleth\")\n  |> range(start: 0, stop: v.timeRangeStart)\n  |> filter(fn: (r) => r._field == \"state\")\n  |> group(columns: [\"tardis_machine_name\", \"machine_type\"])\n  |> reduce(fn: f_r, identity: {_value: 0})\n  |> duplicate(column: \"_stop\", as: \"_time\")\n\nnew = from(bucket: \"batleth\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> filter(fn: (r) => r._field == \"state\")\n//  |> filter(fn: (r) => r.tardis_machine_name == \"${machine}\")\n  |> group(columns: [\"tardis_machine_name\", \"machine_type\"])\n  |> window(every: $__interval)\n  |> reduce(fn: f_r, identity: {_value: 0})\n  |> duplicate(column: \"_stop\", as: \"_time\")\n\nunion(tables: [offset, new])\n  |> window(every: inf)\n  |> cumulativeSum()\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> yield()\n",
+          "refId": "A",
+          "resultFormat": "time_series",
+          "select": [
+            [
+              {
+                "params": [
+                  "value"
+                ],
+                "type": "field"
+              },
+              {
+                "params": [],
+                "type": "mean"
+              }
+            ]
+          ],
+          "tags": []
+        }
+      ],
+      "thresholds": [],
+      "timeFrom": null,
+      "timeRegions": [],
+      "timeShift": null,
+      "title": "nodes running",
+      "tooltip": {
+        "shared": true,
+        "sort": 0,
+        "value_type": "individual"
+      },
+      "type": "graph",
+      "xaxis": {
+        "buckets": null,
+        "mode": "time",
+        "name": null,
+        "show": true,
+        "values": []
+      },
+      "yaxes": [
+        {
+          "format": "none",
+          "label": null,
+          "logBase": 1,
+          "max": null,
+          "min": null,
+          "show": true
+        },
+        {
+          "format": "short",
+          "label": null,
+          "logBase": 1,
+          "max": null,
+          "min": null,
+          "show": true
+        }
+      ],
+      "yaxis": {
+        "align": false,
+        "alignLevel": null
+      }
+    },
+    {
+      "aliasColors": {},
+      "bars": false,
+      "dashLength": 10,
+      "dashes": false,
+      "datasource": "InfluxDB",
+      "fieldConfig": {
+        "defaults": {},
+        "overrides": []
+      },
+      "fill": 1,
+      "fillGradient": 0,
+      "gridPos": {
+        "h": 9,
+        "w": 12,
+        "x": 12,
+        "y": 1
+      },
+      "hiddenSeries": false,
+      "id": 8,
+      "interval": "1s",
+      "legend": {
+        "avg": false,
+        "current": false,
+        "max": false,
+        "min": false,
+        "show": true,
+        "total": false,
+        "values": false
+      },
+      "lines": true,
+      "linewidth": 1,
+      "maxDataPoints": null,
+      "nullPointMode": "null",
+      "options": {
+        "alertThreshold": true
+      },
+      "percentage": false,
+      "pluginVersion": "7.5.7",
+      "pointradius": 2,
+      "points": false,
+      "renderer": "flot",
+      "seriesOverrides": [],
+      "spaceLength": 10,
+      "stack": false,
+      "steppedLine": false,
+      "targets": [
+        {
+          "groupBy": [
+            {
+              "params": [
+                "$__interval"
+              ],
+              "type": "time"
+            },
+            {
+              "params": [
+                "null"
+              ],
+              "type": "fill"
+            }
+          ],
+          "orderByTime": "ASC",
+          "policy": "default",
+          "query": "from(bucket: \"batleth\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> filter(fn: (r) => r._measurement == \"tardis_pipeline\")\n  |> filter(fn: (r) => r._field == \"demand\" or r._field == \"supply\")\n  |> drop(columns: [\"host\"])\n  |> aggregateWindow(every: $__interval, fn: mean)\n  |> yield()",
+          "queryType": "randomWalk",
+          "refId": "A",
+          "resultFormat": "time_series",
+          "select": [
+            [
+              {
+                "params": [
+                  "value"
+                ],
+                "type": "field"
+              },
+              {
+                "params": [],
+                "type": "mean"
+              }
+            ]
+          ],
+          "tags": []
+        }
+      ],
+      "thresholds": [],
+      "timeFrom": null,
+      "timeRegions": [],
+      "timeShift": null,
+      "title": "pipeline demand/supply (mean)",
+      "tooltip": {
+        "shared": true,
+        "sort": 0,
+        "value_type": "individual"
+      },
+      "type": "graph",
+      "xaxis": {
+        "buckets": null,
+        "mode": "time",
+        "name": null,
+        "show": true,
+        "values": []
+      },
+      "yaxes": [
+        {
+          "$$hashKey": "object:115",
+          "decimals": null,
+          "format": "short",
+          "label": "cpus",
+          "logBase": 1,
+          "max": null,
+          "min": null,
+          "show": true
+        },
+        {
+          "$$hashKey": "object:116",
+          "format": "short",
+          "label": null,
+          "logBase": 1,
+          "max": null,
+          "min": null,
+          "show": true
+        }
+      ],
+      "yaxis": {
+        "align": false,
+        "alignLevel": null
+      }
+    },
+    {
+      "aliasColors": {},
+      "bars": false,
+      "dashLength": 10,
+      "dashes": false,
+      "datasource": "InfluxDB",
+      "fieldConfig": {
+        "defaults": {},
+        "overrides": []
+      },
+      "fill": 1,
+      "fillGradient": 0,
+      "gridPos": {
+        "h": 8,
+        "w": 12,
+        "x": 0,
+        "y": 10
+      },
+      "hiddenSeries": false,
+      "id": 6,
+      "interval": "1s",
+      "legend": {
+        "avg": false,
+        "current": false,
+        "max": false,
+        "min": false,
+        "show": true,
+        "total": false,
+        "values": false
+      },
+      "lines": true,
+      "linewidth": 1,
+      "nullPointMode": "null",
+      "options": {
+        "alertThreshold": true
+      },
+      "percentage": false,
+      "pluginVersion": "7.5.7",
+      "pointradius": 2,
+      "points": false,
+      "renderer": "flot",
+      "seriesOverrides": [],
+      "spaceLength": 10,
+      "stack": false,
+      "steppedLine": false,
+      "targets": [
+        {
+          "groupBy": [
+            {
+              "params": [
+                "$__interval"
+              ],
+              "type": "time"
+            },
+            {
+              "params": [
+                "null"
+              ],
+              "type": "fill"
+            }
+          ],
+          "orderByTime": "ASC",
+          "policy": "default",
+          "query": "f = (r, accumulator) => ({\n    _value: accumulator._value + (if r._value == \"AvailableState\" then 1 else if r._value == \"DownState\" then -1 else 0)\n  })\n\nfrom(bucket: \"batleth\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> filter(fn: (r) => r._field == \"state\")\n  |> group(columns: [\"tardis_machine_name\"])\n  |> window(every: $__interval)\n  |> reduce(fn: f, identity: {_value: 0})\n  |> duplicate(column: \"_stop\", as: \"_time\")\n  |> window(every: inf, timeColumn: \"_time\")\n  |> yield()",
+          "queryType": "randomWalk",
+          "refId": "A",
+          "resultFormat": "time_series",
+          "select": [
+            [
+              {
+                "params": [
+                  "value"
+                ],
+                "type": "field"
+              },
+              {
+                "params": [],
+                "type": "mean"
+              }
+            ]
+          ],
+          "tags": []
+        }
+      ],
+      "thresholds": [],
+      "timeFrom": null,
+      "timeRegions": [],
+      "timeShift": null,
+      "title": "node fluctuation",
+      "tooltip": {
+        "shared": true,
+        "sort": 0,
+        "value_type": "individual"
+      },
+      "type": "graph",
+      "xaxis": {
+        "buckets": null,
+        "mode": "time",
+        "name": null,
+        "show": true,
+        "values": []
+      },
+      "yaxes": [
+        {
+          "format": "short",
+          "label": null,
+          "logBase": 1,
+          "max": null,
+          "min": null,
+          "show": true
+        },
+        {
+          "format": "short",
+          "label": null,
+          "logBase": 1,
+          "max": null,
+          "min": null,
+          "show": true
+        }
+      ],
+      "yaxis": {
+        "align": false,
+        "alignLevel": null
+      }
+    },
+    {
+      "aliasColors": {},
+      "bars": false,
+      "dashLength": 10,
+      "dashes": false,
+      "datasource": "InfluxDB",
+      "fieldConfig": {
+        "defaults": {},
+        "overrides": []
+      },
+      "fill": 1,
+      "fillGradient": 0,
+      "gridPos": {
+        "h": 8,
+        "w": 12,
+        "x": 12,
+        "y": 10
+      },
+      "hiddenSeries": false,
+      "id": 10,
+      "interval": "1s",
+      "legend": {
+        "avg": false,
+        "current": false,
+        "max": false,
+        "min": false,
+        "show": true,
+        "total": false,
+        "values": false
+      },
+      "lines": true,
+      "linewidth": 1,
+      "maxDataPoints": null,
+      "nullPointMode": "null",
+      "options": {
+        "alertThreshold": true
+      },
+      "percentage": false,
+      "pluginVersion": "7.5.7",
+      "pointradius": 2,
+      "points": false,
+      "renderer": "flot",
+      "seriesOverrides": [],
+      "spaceLength": 10,
+      "stack": false,
+      "steppedLine": false,
+      "targets": [
+        {
+          "groupBy": [
+            {
+              "params": [
+                "$__interval"
+              ],
+              "type": "time"
+            },
+            {
+              "params": [
+                "null"
+              ],
+              "type": "fill"
+            }
+          ],
+          "orderByTime": "ASC",
+          "policy": "default",
+          "query": "from(bucket: \"batleth\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> filter(fn: (r) => r._measurement == \"tardis_pipeline\")\n  |> filter(fn: (r) => r._field == \"utilisation\" or r._field == \"allocation\")\n  |> keep(columns: [\"_time\", \"_measurement\", \"_field\", \"_value\", \"tardis_machine_name\"])\n  |> aggregateWindow(every: $__interval, fn: mean)\n  |> yield()",
+          "queryType": "randomWalk",
+          "refId": "A",
+          "resultFormat": "time_series",
+          "select": [
+            [
+              {
+                "params": [
+                  "value"
+                ],
+                "type": "field"
+              },
+              {
+                "params": [],
+                "type": "mean"
+              }
+            ]
+          ],
+          "tags": []
+        }
+      ],
+      "thresholds": [],
+      "timeFrom": null,
+      "timeRegions": [],
+      "timeShift": null,
+      "title": "pipeline (utilization/allocation)",
+      "tooltip": {
+        "shared": true,
+        "sort": 0,
+        "value_type": "individual"
+      },
+      "type": "graph",
+      "xaxis": {
+        "buckets": null,
+        "mode": "time",
+        "name": null,
+        "show": true,
+        "values": []
+      },
+      "yaxes": [
+        {
+          "format": "short",
+          "label": null,
+          "logBase": 1,
+          "max": null,
+          "min": null,
+          "show": true
+        },
+        {
+          "format": "short",
+          "label": null,
+          "logBase": 1,
+          "max": null,
+          "min": null,
+          "show": true
+        }
+      ],
+      "yaxis": {
+        "align": false,
+        "alignLevel": null
+      }
+    }
+  ],
+  "refresh": "5s",
+  "schemaVersion": 27,
+  "style": "dark",
+  "tags": [],
+  "templating": {
+    "list": [
+      {
+        "allValue": null,
+        "current": {
+          "selected": false,
+          "text": "cobald-xvmcqc",
+          "value": "cobald-xvmcqc"
+        },
+        "datasource": "InfluxDB",
+        "definition": "from(bucket: \"batleth\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> keep(columns: [\"tardis_machine_name\", \"_time\"])\n  |> sort(columns: [\"_time\"], desc: true)\n  |> unique(column: \"tardis_machine_name\")",
+        "description": null,
+        "error": null,
+        "hide": 0,
+        "includeAll": false,
+        "label": null,
+        "multi": false,
+        "name": "machine",
+        "options": [],
+        "query": "from(bucket: \"batleth\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> keep(columns: [\"tardis_machine_name\", \"_time\"])\n  |> sort(columns: [\"_time\"], desc: true)\n  |> unique(column: \"tardis_machine_name\")",
+        "refresh": 2,
+        "regex": "",
+        "skipUrlSync": false,
+        "sort": 0,
+        "tagValuesQuery": "",
+        "tags": [],
+        "tagsQuery": "",
+        "type": "query",
+        "useTags": false
+      }
+    ]
+  },
+  "time": {
+    "from": "now-5m",
+    "to": "now"
+  },
+  "timepicker": {},
+  "timezone": "",
+  "title": "cobald",
+  "uid": "urDuvE6Gk",
+  "version": 2
+}

roles/cobald/files/influxdb-dashboard-cobald.json

@@ -0,0 +1,794 @@
+{
+ "meta": {
+  "version": "1",
+  "type": "dashboard",
+  "name": "cobald-Template",
+  "description": "template created from dashboard: cobald"
+ },
+ "content": {
+  "data": {
+   "type": "dashboard",
+   "attributes": {
+    "name": "cobald",
+    "description": ""
+   },
+   "relationships": {
+    "label": {
+     "data": []
+    },
+    "cell": {
+     "data": [
+      {
+       "type": "cell",
+       "id": "07900a722c363000"
+      },
+      {
+       "type": "cell",
+       "id": "07900a7236f63000"
+      },
+      {
+       "type": "cell",
+       "id": "07900a723cf63000"
+      },
+      {
+       "type": "cell",
+       "id": "07900a7243f63000"
+      },
+      {
+       "type": "cell",
+       "id": "079e694f29581000"
+      },
+      {
+       "type": "cell",
+       "id": "079e6e037c181000"
+      }
+     ]
+    },
+    "variable": {
+     "data": []
+    }
+   }
+  },
+  "included": [
+   {
+    "id": "07900a722c363000",
+    "type": "cell",
+    "attributes": {
+     "x": 4,
+     "y": 0,
+     "w": 4,
+     "h": 4
+    },
+    "relationships": {
+     "view": {
+      "data": {
+       "type": "view",
+       "id": "07900a722c363000"
+      }
+     }
+    }
+   },
+   {
+    "id": "07900a7236f63000",
+    "type": "cell",
+    "attributes": {
+     "x": 0,
+     "y": 0,
+     "w": 4,
+     "h": 4
+    },
+    "relationships": {
+     "view": {
+      "data": {
+       "type": "view",
+       "id": "07900a7236f63000"
+      }
+     }
+    }
+   },
+   {
+    "id": "07900a723cf63000",
+    "type": "cell",
+    "attributes": {
+     "x": 4,
+     "y": 4,
+     "w": 4,
+     "h": 4
+    },
+    "relationships": {
+     "view": {
+      "data": {
+       "type": "view",
+       "id": "07900a723cf63000"
+      }
+     }
+    }
+   },
+   {
+    "id": "07900a7243f63000",
+    "type": "cell",
+    "attributes": {
+     "x": 0,
+     "y": 4,
+     "w": 4,
+     "h": 4
+    },
+    "relationships": {
+     "view": {
+      "data": {
+       "type": "view",
+       "id": "07900a7243f63000"
+      }
+     }
+    }
+   },
+   {
+    "id": "079e694f29581000",
+    "type": "cell",
+    "attributes": {
+     "x": 8,
+     "y": 0,
+     "w": 4,
+     "h": 4
+    },
+    "relationships": {
+     "view": {
+      "data": {
+       "type": "view",
+       "id": "079e694f29581000"
+      }
+     }
+    }
+   },
+   {
+    "id": "079e6e037c181000",
+    "type": "cell",
+    "attributes": {
+     "x": 8,
+     "y": 4,
+     "w": 4,
+     "h": 4
+    },
+    "relationships": {
+     "view": {
+      "data": {
+       "type": "view",
+       "id": "079e6e037c181000"
+      }
+     }
+    }
+   },
+   {
+    "type": "view",
+    "id": "07900a722c363000",
+    "attributes": {
+     "name": "nodes running",
+     "properties": {
+      "shape": "chronograf-v2",
+      "queries": [
+       {
+        "text": "from(bucket: \"batleth\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> filter(fn: (r) => r._field == \"state\")\n  |> group()\n  |> window(every: 10s)\n  |> reduce(fn: (r, accumulator) => ({\n      _value: accumulator._value + (\n        if r._value == \"AvailableState\" then 1 \n        else if r._value == \"DownState\" then -1 \n        else 0)\n    }), identity: {_value: 0})\n  |> duplicate(column: \"_stop\", as: \"_time\")\n  |> window(every: inf) //, timeColumn: \"_time\")\n  |> cumulativeSum()\n//  |> reduce(fn: (r, accumulator) => ({r with x: r._value * 2}), identity: {x:0})\n//  |> map(fn: (r) => ({r with vnew: r._value*2}))\n//  |> integral(unit: 10s, timeColumn: \"_stop\")\n//  |> window(every: inf) //, timeColumn: \"_stop\")\n  |> yield()",
+        "editMode": "advanced",
+        "name": "",
+        "builderConfig": {
+         "buckets": [],
+         "tags": [
+          {
+           "key": "_measurement",
+           "values": [],
+           "aggregateFunctionType": "filter"
+          }
+         ],
+         "functions": [
+          {
+           "name": "mean"
+          }
+         ],
+         "aggregateWindow": {
+          "period": "auto",
+          "fillValues": false
+         }
+        }
+       }
+      ],
+      "axes": {
+       "x": {
+        "bounds": [
+         "",
+         ""
+        ],
+        "label": "",
+        "prefix": "",
+        "suffix": "",
+        "base": "10",
+        "scale": "linear"
+       },
+       "y": {
+        "bounds": [
+         "",
+         ""
+        ],
+        "label": "",
+        "prefix": "",
+        "suffix": "",
+        "base": "10",
+        "scale": "linear"
+       }
+      },
+      "type": "xy",
+      "legend": {},
+      "geom": "line",
+      "colors": [
+       {
+        "id": "9b960932-18d9-4f57-80ba-24998a06613d",
+        "type": "scale",
+        "hex": "#31C0F6",
+        "name": "Nineteen Eighty Four",
+        "value": 0
+       },
+       {
+        "id": "c1742651-0d5e-4148-b9c0-92beb642417a",
+        "type": "scale",
+        "hex": "#A500A5",
+        "name": "Nineteen Eighty Four",
+        "value": 0
+       },
+       {
+        "id": "756fa8cb-9d9c-4e45-9a4f-f2b106b0216a",
+        "type": "scale",
+        "hex": "#FF7E27",
+        "name": "Nineteen Eighty Four",
+        "value": 0
+       }
+      ],
+      "note": "",
+      "showNoteWhenEmpty": false,
+      "xColumn": "_time",
+      "generateXAxisTicks": [],
+      "xTotalTicks": 0,
+      "xTickStart": 0,
+      "xTickStep": 0,
+      "yColumn": "_value",
+      "generateYAxisTicks": [],
+      "yTotalTicks": 0,
+      "yTickStart": 0,
+      "yTickStep": 0,
+      "shadeBelow": false,
+      "position": "overlaid",
+      "timeFormat": "",
+      "hoverDimension": "auto",
+      "legendColorizeRows": true,
+      "legendOpacity": 1,
+      "legendOrientationThreshold": 100000000
+     }
+    }
+   },
+   {
+    "type": "view",
+    "id": "07900a7236f63000",
+    "attributes": {
+     "name": "Name this Cell",
+     "properties": {
+      "shape": "chronograf-v2",
+      "queries": [
+       {
+        "text": "from(bucket: \"batleth\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> filter(fn: (r) => r[\"tardis_machine_name\"] == \"678162c190d5\")\n  |> window(every: 10s)\n  |> count()\n  |> duplicate(column: \"_stop\", as: \"_time\")\n  |> window(every: inf)\n  |> yield()",
+        "editMode": "advanced",
+        "name": "",
+        "builderConfig": {
+         "buckets": [],
+         "tags": [
+          {
+           "key": "_measurement",
+           "values": [],
+           "aggregateFunctionType": "filter"
+          }
+         ],
+         "functions": [
+          {
+           "name": "mean"
+          }
+         ],
+         "aggregateWindow": {
+          "period": "auto",
+          "fillValues": false
+         }
+        }
+       }
+      ],
+      "axes": {
+       "x": {
+        "bounds": [
+         "",
+         ""
+        ],
+        "label": "",
+        "prefix": "",
+        "suffix": "",
+        "base": "10",
+        "scale": "linear"
+       },
+       "y": {
+        "bounds": [
+         "",
+         ""
+        ],
+        "label": "",
+        "prefix": "",
+        "suffix": "",
+        "base": "10",
+        "scale": "linear"
+       }
+      },
+      "type": "xy",
+      "legend": {},
+      "geom": "line",
+      "colors": [
+       {
+        "id": "2566435b-7ee0-4222-8ac0-b7f14ab783d9",
+        "type": "scale",
+        "hex": "#31C0F6",
+        "name": "Nineteen Eighty Four",
+        "value": 0
+       },
+       {
+        "id": "9263bcff-35a0-4025-bacd-68a1bef54784",
+        "type": "scale",
+        "hex": "#A500A5",
+        "name": "Nineteen Eighty Four",
+        "value": 0
+       },
+       {
+        "id": "2f04bd8c-2203-4be6-bc34-c25720d24379",
+        "type": "scale",
+        "hex": "#FF7E27",
+        "name": "Nineteen Eighty Four",
+        "value": 0
+       }
+      ],
+      "note": "",
+      "showNoteWhenEmpty": false,
+      "xColumn": "_time",
+      "generateXAxisTicks": [],
+      "xTotalTicks": 0,
+      "xTickStart": 0,
+      "xTickStep": 0,
+      "yColumn": "_value",
+      "generateYAxisTicks": [],
+      "yTotalTicks": 0,
+      "yTickStart": 0,
+      "yTickStep": 0,
+      "shadeBelow": false,
+      "position": "overlaid",
+      "timeFormat": "",
+      "hoverDimension": "auto",
+      "legendColorizeRows": true,
+      "legendOpacity": 1,
+      "legendOrientationThreshold": 100000000
+     }
+    }
+   },
+   {
+    "type": "view",
+    "id": "07900a723cf63000",
+    "attributes": {
+     "name": "node fluctuation",
+     "properties": {
+      "shape": "chronograf-v2",
+      "queries": [
+       {
+        "text": "f = (r, accumulator) => ({\n    _value: accumulator._value + (if r._value == \"AvailableState\" then 1 else if r._value == \"DownState\" then -1 else 0)\n  })\n\nfrom(bucket: \"batleth\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> filter(fn: (r) => r._field == \"state\")\n  |> group()\n  |> window(every: 10s)\n  |> reduce(fn: f, identity: {_value: 0})\n  |> duplicate(column: \"_stop\", as: \"_time\")\n  |> window(every: inf, timeColumn: \"_time\")\n  |> yield()",
+        "editMode": "advanced",
+        "name": "",
+        "builderConfig": {
+         "buckets": [],
+         "tags": [
+          {
+           "key": "_measurement",
+           "values": [],
+           "aggregateFunctionType": "filter"
+          }
+         ],
+         "functions": [
+          {
+           "name": "mean"
+          }
+         ],
+         "aggregateWindow": {
+          "period": "auto",
+          "fillValues": false
+         }
+        }
+       }
+      ],
+      "axes": {
+       "x": {
+        "bounds": [
+         "",
+         ""
+        ],
+        "label": "",
+        "prefix": "",
+        "suffix": "",
+        "base": "10",
+        "scale": "linear"
+       },
+       "y": {
+        "bounds": [
+         "",
+         ""
+        ],
+        "label": "",
+        "prefix": "",
+        "suffix": "",
+        "base": "10",
+        "scale": "linear"
+       }
+      },
+      "type": "xy",
+      "legend": {},
+      "geom": "line",
+      "colors": [
+       {
+        "id": "9b960932-18d9-4f57-80ba-24998a06613d",
+        "type": "scale",
+        "hex": "#31C0F6",
+        "name": "Nineteen Eighty Four",
+        "value": 0
+       },
+       {
+        "id": "c1742651-0d5e-4148-b9c0-92beb642417a",
+        "type": "scale",
+        "hex": "#A500A5",
+        "name": "Nineteen Eighty Four",
+        "value": 0
+       },
+       {
+        "id": "756fa8cb-9d9c-4e45-9a4f-f2b106b0216a",
+        "type": "scale",
+        "hex": "#FF7E27",
+        "name": "Nineteen Eighty Four",
+        "value": 0
+       }
+      ],
+      "note": "",
+      "showNoteWhenEmpty": false,
+      "xColumn": "_time",
+      "generateXAxisTicks": [],
+      "xTotalTicks": 0,
+      "xTickStart": 0,
+      "xTickStep": 0,
+      "yColumn": "_value",
+      "generateYAxisTicks": [],
+      "yTotalTicks": 0,
+      "yTickStart": 0,
+      "yTickStep": 0,
+      "shadeBelow": false,
+      "position": "overlaid",
+      "timeFormat": "",
+      "hoverDimension": "auto",
+      "legendColorizeRows": true,
+      "legendOpacity": 1,
+      "legendOrientationThreshold": 100000000
+     }
+    }
+   },
+   {
+    "type": "view",
+    "id": "07900a7243f63000",
+    "attributes": {
+     "name": "states",
+     "properties": {
+      "shape": "chronograf-v2",
+      "queries": [
+       {
+        "text": "from(bucket: \"batleth\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> filter(fn: (r) => r._field == \"state\")\n  |> group(columns: [\"_value\"], mode: \"by\")\n  |> duplicate(column: \"_value\", as: \"state\")\n  |> window(every: 10s)\n  |> count(column: \"state\")\n  |> rename(columns: {\"_value\": \"_field\", \"state\": \"_value\"})\n  |> group(columns: [\"_field\"])\n  |> yield()",
+        "editMode": "advanced",
+        "name": "",
+        "builderConfig": {
+         "buckets": [],
+         "tags": [
+          {
+           "key": "_measurement",
+           "values": [],
+           "aggregateFunctionType": "filter"
+          }
+         ],
+         "functions": [
+          {
+           "name": "mean"
+          }
+         ],
+         "aggregateWindow": {
+          "period": "auto",
+          "fillValues": false
+         }
+        }
+       }
+      ],
+      "axes": {
+       "x": {
+        "bounds": [
+         "",
+         ""
+        ],
+        "label": "",
+        "prefix": "",
+        "suffix": "",
+        "base": "10",
+        "scale": "linear"
+       },
+       "y": {
+        "bounds": [
+         "",
+         ""
+        ],
+        "label": "",
+        "prefix": "",
+        "suffix": "",
+        "base": "10",
+        "scale": "linear"
+       }
+      },
+      "type": "xy",
+      "legend": {},
+      "geom": "line",
+      "colors": [
+       {
+        "id": "9b960932-18d9-4f57-80ba-24998a06613d",
+        "type": "scale",
+        "hex": "#31C0F6",
+        "name": "Nineteen Eighty Four",
+        "value": 0
+       },
+       {
+        "id": "c1742651-0d5e-4148-b9c0-92beb642417a",
+        "type": "scale",
+        "hex": "#A500A5",
+        "name": "Nineteen Eighty Four",
+        "value": 0
+       },
+       {
+        "id": "756fa8cb-9d9c-4e45-9a4f-f2b106b0216a",
+        "type": "scale",
+        "hex": "#FF7E27",
+        "name": "Nineteen Eighty Four",
+        "value": 0
+       }
+      ],
+      "note": "",
+      "showNoteWhenEmpty": false,
+      "xColumn": "_stop",
+      "generateXAxisTicks": [],
+      "xTotalTicks": 0,
+      "xTickStart": 0,
+      "xTickStep": 0,
+      "yColumn": "_value",
+      "generateYAxisTicks": [],
+      "yTotalTicks": 0,
+      "yTickStart": 0,
+      "yTickStep": 0,
+      "shadeBelow": false,
+      "position": "overlaid",
+      "timeFormat": "",
+      "hoverDimension": "auto",
+      "legendColorizeRows": true,
+      "legendOpacity": 1,
+      "legendOrientationThreshold": 100000000
+     }
+    }
+   },
+   {
+    "type": "view",
+    "id": "079e694f29581000",
+    "attributes": {
+     "name": "pipeline (demand / supply)",
+     "properties": {
+      "shape": "chronograf-v2",
+      "queries": [
+       {
+        "text": "from(bucket: \"batleth\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> filter(fn: (r) => r._measurement == \"tardis_pipeline\")\n  |> filter(fn: (r) => r._field == \"demand\" or r._field == \"supply\")\n//  |> filter(fn: (r) => r._field == \"state\")\n//  |> group()\n//  |> window(every: 10s)\n//  |> duplicate(column: \"_stop\", as: \"_time\")\n//  |> window(every: inf) //, timeColumn: \"_time\")\n//  |> cumulativeSum()\n  |> yield()",
+        "editMode": "advanced",
+        "name": "",
+        "builderConfig": {
+         "buckets": [],
+         "tags": [
+          {
+           "key": "_measurement",
+           "values": [],
+           "aggregateFunctionType": "filter"
+          }
+         ],
+         "functions": [
+          {
+           "name": "mean"
+          }
+         ],
+         "aggregateWindow": {
+          "period": "auto",
+          "fillValues": false
+         }
+        }
+       }
+      ],
+      "axes": {
+       "x": {
+        "bounds": [
+         "",
+         ""
+        ],
+        "label": "",
+        "prefix": "",
+        "suffix": "",
+        "base": "10",
+        "scale": "linear"
+       },
+       "y": {
+        "bounds": [
+         "",
+         ""
+        ],
+        "label": "",
+        "prefix": "",
+        "suffix": "",
+        "base": "10",
+        "scale": "linear"
+       }
+      },
+      "type": "xy",
+      "legend": {},
+      "geom": "line",
+      "colors": [
+       {
+        "id": "4ef29481-ecf3-4a09-b0f5-e34e8d3e50b5",
+        "type": "scale",
+        "hex": "#31C0F6",
+        "name": "Nineteen Eighty Four",
+        "value": 0
+       },
+       {
+        "id": "719de04a-f70f-4c54-a1bb-982a9d13dbae",
+        "type": "scale",
+        "hex": "#A500A5",
+        "name": "Nineteen Eighty Four",
+        "value": 0
+       },
+       {
+        "id": "8d5c1f25-3801-4cdd-ad40-b8e2e78342a0",
+        "type": "scale",
+        "hex": "#FF7E27",
+        "name": "Nineteen Eighty Four",
+        "value": 0
+       }
+      ],
+      "note": "",
+      "showNoteWhenEmpty": false,
+      "xColumn": "_time",
+      "generateXAxisTicks": [],
+      "xTotalTicks": 0,
+      "xTickStart": 0,
+      "xTickStep": 0,
+      "yColumn": "_value",
+      "generateYAxisTicks": [],
+      "yTotalTicks": 0,
+      "yTickStart": 0,
+      "yTickStep": 0,
+      "shadeBelow": false,
+      "position": "overlaid",
+      "timeFormat": "",
+      "hoverDimension": "auto",
+      "legendColorizeRows": true,
+      "legendOpacity": 1,
+      "legendOrientationThreshold": 100000000
+     }
+    }
+   },
+   {
+    "type": "view",
+    "id": "079e6e037c181000",
+    "attributes": {
+     "name": "pipeline (Utilization / Allocation)",
+     "properties": {
+      "shape": "chronograf-v2",
+      "queries": [
+       {
+        "text": "from(bucket: \"batleth\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> filter(fn: (r) => r._measurement == \"tardis_pipeline\")\n  |> filter(fn: (r) => r._field == \"utilisation\" or r._field == \"allocation\")\n//  |> filter(fn: (r) => r._field == \"state\")\n//  |> group()\n//  |> window(every: 10s)\n//  |> duplicate(column: \"_stop\", as: \"_time\")\n//  |> window(every: inf) //, timeColumn: \"_time\")\n//  |> cumulativeSum()\n  |> yield()",
+        "editMode": "advanced",
+        "name": "",
+        "builderConfig": {
+         "buckets": [],
+         "tags": [
+          {
+           "key": "_measurement",
+           "values": [],
+           "aggregateFunctionType": "filter"
+          }
+         ],
+         "functions": [
+          {
+           "name": "mean"
+          }
+         ],
+         "aggregateWindow": {
+          "period": "auto",
+          "fillValues": false
+         }
+        }
+       }
+      ],
+      "axes": {
+       "x": {
+        "bounds": [
+         "",
+         ""
+        ],
+        "label": "",
+        "prefix": "",
+        "suffix": "",
+        "base": "10",
+        "scale": "linear"
+       },
+       "y": {
+        "bounds": [
+         "",
+         ""
+        ],
+        "label": "",
+        "prefix": "",
+        "suffix": "",
+        "base": "10",
+        "scale": "linear"
+       }
+      },
+      "type": "xy",
+      "legend": {},
+      "geom": "line",
+      "colors": [
+       {
+        "id": "4ef29481-ecf3-4a09-b0f5-e34e8d3e50b5",
+        "type": "scale",
+        "hex": "#31C0F6",
+        "name": "Nineteen Eighty Four",
+        "value": 0
+       },
+       {
+        "id": "719de04a-f70f-4c54-a1bb-982a9d13dbae",
+        "type": "scale",
+        "hex": "#A500A5",
+        "name": "Nineteen Eighty Four",
+        "value": 0
+       },
+       {
+        "id": "8d5c1f25-3801-4cdd-ad40-b8e2e78342a0",
+        "type": "scale",
+        "hex": "#FF7E27",
+        "name": "Nineteen Eighty Four",
+        "value": 0
+       }
+      ],
+      "note": "",
+      "showNoteWhenEmpty": false,
+      "xColumn": "_time",
+      "generateXAxisTicks": [],
+      "xTotalTicks": 0,
+      "xTickStart": 0,
+      "xTickStep": 0,
+      "yColumn": "_value",
+      "generateYAxisTicks": [],
+      "yTotalTicks": 0,
+      "yTickStart": 0,
+      "yTickStep": 0,
+      "shadeBelow": false,
+      "position": "overlaid",
+      "timeFormat": "",
+      "hoverDimension": "auto",
+      "legendColorizeRows": true,
+      "legendOpacity": 1,
+      "legendOrientationThreshold": 100000000
+     }
+    }
+   }
+  ]
+ },
+ "labels": []
+}

roles/cobald/files/influxdb.repo

@@ -0,0 +1,6 @@
+[influxdb]
+name = InfluxDB Repository - RHEL \$releasever
+baseurl = https://repos.influxdata.com/rhel/\$releasever/\$basearch/stable
+enabled = 1
+gpgcheck = 1
+gpgkey = https://repos.influxdata.com/influxdb.key

roles/cobald/files/init-cobaldmodules.sh

@@ -0,0 +1,5 @@
+#!/bin/sh
+
+[ -f /usr/local/lib/cobaldmodules/setup.py -a \
+	-d /usr/local/lib/cobaldmodules/cobaldmodules ] && \
+		pip3 install --no-deps --editable /usr/local/lib/cobaldmodules

roles/cobald/files/slurm-slurmd.def

@@ -0,0 +1,31 @@
+Bootstrap: docker-daemon
+From: slurm:slurmd
+
+%files
+	31-slurmd-configless /etc/docker-init.d/31-slurm-configless
+	/container/volumes/munge/munge.key /etc/munge/munge.key
+	cgroup.conf.noautomount /etc/slurm/cgroup.conf
+
+%post
+	rm /etc/docker-init.d/30-slurmd
+	chmod 755 /etc/docker-init.d/31-slurm-configless
+
+%startscript
+	if [ -z "${1}" -o -z "${2}" ] ; then
+		echo "undefined variables slurmctld or nodename"
+		exit 1
+	fi
+	export slurmctld="${1}"
+	export nodename="${2}"
+	echo ${slurmctld} ${nodename} ${SLURMD_LOG_PATH} ${SLURM_SCHED_LOG_PATH}
+	exec /usr/local/sbin/entrypoint.sh /usr/local/sbin/docker-init
+
+%runscript
+	if [ -z "${1}" -o -z "${2}" ] ; then
+		echo "undefined variables slurmctld or nodename"
+		exit 1
+	fi
+	export slurmctld="${1}"
+	export nodename="${2}"
+	echo ${slurmctld} ${nodename} ${SLURMD_LOG_PATH} ${SLURM_SCHED_LOG_PATH}
+	exec /usr/local/sbin/entrypoint.sh /usr/local/sbin/docker-init

roles/cobald/files/start-drone

@@ -0,0 +1,59 @@
+#!/bin/sh
+#SBATCH -D /shared
+export
+echo $@
+nodename=$(hostname | awk '{ print "drone" substr($1,match($1, "([[:digit:]]+)")) }')
+
+SHUTDOWN_DONE=0
+
+function handler_quit(){
+	[ $SHUTDOWN_DONE -ne 0 ] && return
+	set -x
+	echo "drain container"
+	scontrol update NodeName=${nodename} State=DRAIN Reason="cobald node quit"
+	shutdown_jobs=$(squeue -w ${nodename} --noheader -O jobid)
+	[ -n "${shutdown_jobs}" ] && scancel ${shutdown_jobs}
+	#scancel -w ${nodename}
+	i=$(( $(scontrol show config | grep KillWait | \
+			sed 's/^KillWait.*= \([0-9]*\) sec/\1/') - 2 ))
+	while [ -n "$(squeue -w ${nodename} --noheader -O jobid)" -o ${i} -lt 1 ]
+	do
+		i=$(( ${i} - 1 ))
+		sleep 1
+	done
+	scancel -s KILL -w ${nodename}  # hard kill all remaining jobs
+	echo "shutdown container"
+	scontrol update NodeName=${nodename} State=DOWN Reason=shutdown
+	singularity instance stop slurm-drone
+	scontrol update NodeName=${nodename} State=FUTURE
+	umount /inner-cgroup/freezer
+	umount /inner-cgroup
+	SHUTDOWN_DONE=1
+	exit 0
+}
+
+# set -x
+
+trap handler_quit EXIT
+
+echo "mounting cgroups"
+mkdir /inner-cgroup
+mount -t tmpfs none /inner-cgroup
+mkdir /inner-cgroup/freezer/
+mount --bind /sys/fs/cgroup/freezer/slurm/ /inner-cgroup/freezer/
+mount -o remount,ro /inner-cgroup
+
+echo "starting ${nodename}"
+scontrol update NodeName=${nodename} State=RESUME # revoke last DRAIN
+scontrol update NodeName=${nodename} State=FUTURE
+singularity instance start \
+	-B /inner-cgroup/:/sys/fs/cgroup/ \
+	--writable-tmpfs /shared/slurmd.sif slurm-drone \
+	slurm-ctl ${nodename}
+# scontrol update NodeName=${nodename} NodeHostname=${SLURM_JOB_ID}
+scontrol update NodeName=${nodename} NodeHostname=${TardisDroneUuid}
+if [ $? -eq 0 ] ; then
+	echo "container started, sleeping $(( 60 * ${SLURM_Walltime} - 2 ))"
+	sleep $(( 60 * ${SLURM_Walltime} - 2 ))
+fi
+handler_quit

roles/cobald/files/telegraf.Dockerfile

@@ -0,0 +1,8 @@
+FROM centos:7
+
+COPY influxdb.repo /etc/yum.repos.d/influxdb.repo
+
+RUN yum -y install telegraf &&\
+    yum clean all && rm -rf /var/cache/yum
+
+CMD telegraf

roles/cobald/library/TODO.md

@@ -0,0 +1,11 @@
+# Tests
+    ANSIBLE_LIBRARY=. ansible -m influx_bucket -a "base='http://192.168.122.140:28086' org='my-org' auth_token='87-fEnSlQldFi1T_CLHsrHxH-T9VKey-qzUbVH6tmR2QzL4oZzbUPwzS1wzOoIkyfmyGbRv75yLjYfztxziivw==' name='bucky' description='test 123'" localhost -vvv
+    ANSIBLE_LIBRARY=. ansible -m influx_token -a "base='http://192.168.122.140:28086' org='my-org' auth_token='87-fEnSlQldFi1T_CLHsrHxH-T9VKey-qzUbVH6tmR2QzL4oZzbUPwzS1wzOoIkyfmyGbRv75yLjYfztxziivw==' key='foo' description='test 123' permissions=\"{{'[{\\\"action\\\": \\\"write\\\",\\\"resource\\\": {\\\"type\\\": \\\"buckets\\\"} }]'|from_json}}\"" localhost -vvv
+    ANSIBLE_LIBRARY=. ansible -m influx_dashboard -create -a "base='http://192.168.122.140:28086' org='my-org' token='2Mji-PvTzgn2oie5p36pJ-vxqWCnxczMWGrnYz2nUHj6Q6XvdIGiLPmK4DjX16KGhOjxQ5dWymDusE8qJrhFFg==' data='{{lookup(\"file\", \"../files/influxdb-dashboard-cobald.json\")}}'" localhost -vvv
+Missing: lot of stuff, e.g. missing tokens, invalid data like bucket instead of buckets ...
+
+# TODO
+* tests
+* state (present/absent)
+* `module_utils/urls.py` (https://github.com/ansible/ansible/blob/devel/lib/ansible/module_utils/urls.py)
+* see module notes

roles/cobald/tasks/dev.yml

@@ -1,58 +0,0 @@
-- yum:
-    name: git
-    state: present
-
-- git:
-    repo: https://github.com/thoto/cobald
-    dest: "~{{unpriv_user}}/cobald-src"
-    version: bugfix/mixed_construction_methods
-  become: yes
-  become_user: "{{unpriv_user}}"
-  register: cobald_git_pull
-
-- git:
-    repo: https://github.com/MatterMiners/tardis
-    dest: "~{{unpriv_user}}/tardis-src"
-    version: master
-  become: yes
-  become_user: "{{unpriv_user}}"
-  register: tardis_git_pull
-
-- name: "get unpriv_user {{unpriv_user}} uid and gid"
-  getent:
-    database: passwd
-    key: "{{unpriv_user}}"
-
-- name: run pip install
-  docker_container:
-    image: cobald
-    name: "cobald-src-{{item.name}}-install"
-    volumes:
-    - "~{{unpriv_user}}/{{item.name}}-src:/usr/local/src/{{item.name}}:rw"
-    state: started
-    detach: False
-    cleanup: True
-    user: "{{getent_passwd[unpriv_user][1]}}:{{getent_passwd[unpriv_user][2]}}"
-    entrypoint: ""
-    command: |
-      bash -c 'HOME=/tmp pip3 install --editable /usr/local/src/{{item.name}}'
-  with_items:
-  - name: cobald
-    run: "{{cobald_git_pull.changed}}"
-  - name: tardis
-    run: "{{tardis_git_pull.changed}}"
-  when: item.run
-
-- docker_container:
-    name: cobald-dev
-    image: cobald
-    volumes:
-    - "~{{unpriv_user}}/cobald:/etc/cobald"
-    - "~{{unpriv_user}}/cobald/modules:/usr/local/src/cobaldmodules"
-    - "~{{unpriv_user}}/cobald-src:/usr/local/src/cobald:ro"
-    - "~{{unpriv_user}}/tardis-src:/usr/local/src/tardis:ro"
-    state: started
-    detach: True
-    cleanup: True
-    interactive: True
-    command: /bin/bash

roles/cobald/tasks/dockerimage-generic.yml

@@ -0,0 +1,34 @@
+- file:
+    path: "/container/docker-images/cobald.{{cobald_image_tag|default('latest')}}/"
+    state: directory
+    owner: "{{unpriv_user}}"
+    group: docker
+
+- template:
+    src: cobald.Dockerfile
+    dest: "/container/docker-images/cobald.{{cobald_image_tag|default('latest')}}/Dockerfile"
+    owner: "{{unpriv_user}}"
+    group: docker
+  register: cobald_cp_dockerfile
+
+- copy:
+    src: "{{item}}"
+    dest: "/container/docker-images/cobald.{{cobald_image_tag|default('latest')}}/{{item}}"
+    owner: "{{unpriv_user}}"
+    group: docker
+    mode: 0755
+  with_items:
+  - cobald-entrypoint.sh
+  - init-cobaldmodules.sh
+  register: cobald_cp_files
+
+- docker_image:
+    name: "cobald"
+    tag: "{{cobald_image_tag|default('latest')}}"
+#    pull: False
+    build:
+      pull: False
+      path: "/container/docker-images/cobald.{{cobald_image_tag|default('latest')}}/"
+    source: build
+    force_source: "{{cobald_cp_dockerfile.changed or cobald_cp_files.changed}}"
+

roles/cobald/tasks/facts.yml

@@ -0,0 +1,11 @@
+- block:
+  - docker_container_info:
+      name: "{{ container_name | mandatory }}"
+    register: cobald_container_info
+
+  - set_fact:
+      cobald_container_hostname: |-
+        {{cobald_container_info.container.Config.Hostname | default('cobald-'+
+          lookup('password', '/dev/null chars=ascii_lowercase length=6')) }}
+  when: cobald_container_hostname is not defined
+

roles/cobald/tasks/grafana.yml

@@ -0,0 +1,54 @@
+- name: create influx token for grafana
+  influx_token:
+    base: "http://localhost:{{influx_pubport}}"
+    org: "my-org"
+    auth_token: "{{influx_admin_token}}"
+    description: grafana read access
+    key: grafana
+    permissions:
+    - action: read
+      resource:
+        type: buckets
+  register: influx_grafana_token
+
+- name: run grafana
+  docker_container:
+    name: ed-grafana
+    image: docker.io/grafana/grafana:7.5.7
+    hostname: ed-grafana
+    domainname: cobald.local
+    networks:
+    - name: "{{cobald_docker_network}}"
+    networks_cli_compatible: True
+    published_ports:
+    - "3000:3000"
+    state: started
+    detach: True
+    cleanup: True
+
+- wait_for:
+    host: localhost
+    port: 3000
+
+- community.grafana.grafana_datasource:
+    grafana_url: http://localhost:3000
+    grafana_user: admin
+    grafana_password: admin
+    name: InfluxDB
+    ds_type: influxdb
+    ds_url: "{{influx_url}}"
+    additional_json_data:
+      defaultBucket: "{{influx_bucket}}"
+      organization: "{{influx_org}}"
+      version: Flux
+    additional_secure_json_data:
+      token: "{{influx_grafana_token.token}}"
+
+- community.grafana.grafana_dashboard:
+    grafana_url: http://localhost:3000
+    grafana_user: admin
+    grafana_password: admin
+    state: present
+    commit_message: updated by ansible
+    overwrite: yes
+    json_data: "{{lookup('file', 'grafana-dashboard.json')|from_json}}"

roles/cobald/tasks/influxdb.yml

@@ -0,0 +1,76 @@
+- name: run influxdb in docker container
+  docker_container:
+    name: ed-influxdb
+    image: docker.io/library/influxdb:2.0
+    hostname: "{{cobald_influx_hostname}}"
+    domainname: "{{cobald_domainname}}"
+    networks:
+    - name: "{{ cobald_docker_network }}"
+    networks_cli_compatible: True
+    published_ports:
+    - "{{influx_pubport}}:8086"
+    volumes:
+    - "ed-influxdb-data:/var/lib/influxdb2"
+    - "ed-influxdb-config:/etc/influxdb2"
+    - "/container/volumes/influxdb-backup/:/backup"
+    env:
+      DOCKER_INFLUXDB_INIT_MODE: setup
+      DOCKER_INFLUXDB_INIT_USERNAME: "{{influx_admin_user}}"
+      DOCKER_INFLUXDB_INIT_PASSWORD: "{{influx_admin_pw}}"
+      DOCKER_INFLUXDB_INIT_ORG: "{{influx_org}}"
+      DOCKER_INFLUXDB_INIT_BUCKET: my-bucket
+    state: started
+    detach: True
+    cleanup: True
+
+- name: add ansible connection to influxdb container
+  add_host:
+    name: ed-influxdb
+    ansible_connection: docker
+    ansible_docker_extra_args: "-H=ssh://{{ansible_host}}"
+  changed_when: False
+
+- name: wait for influx to run
+  raw: until curl http://localhost:8086 ; do sleep 1 ; done
+  changed_when: False
+  delegate_to: ed-influxdb
+
+- name: fetch influxdb auth token
+  raw: influx auth list --user my-user --hide-headers --json
+  register: influx_token_fetch
+  changed_when: False
+  delegate_to: ed-influxdb
+
+- name: set influxdb admin token
+  set_fact:
+    influx_admin_token:
+      "{{(influx_token_fetch.stdout | from_json | first).token}}"
+
+- name: create influxdb bucket for cobald
+  influx_bucket:
+    base: "http://localhost:{{influx_pubport}}"
+    org: "my-org"
+    auth_token: "{{influx_admin_token}}"
+    name: "{{influx_bucket}}"
+
+- name: create influxdb dashboard
+  influx_dashboard:
+    base: "http://localhost:{{influx_pubport}}"
+    org: "my-org"
+    auth_token: "{{influx_admin_token}}"
+    data: "{{lookup('file', 'influxdb-dashboard-cobald.json')}}"
+  when: influxdb_dashboard | default(True)
+
+- name: create influxdb write access token for telegraf
+  influx_token:
+    base: "http://localhost:{{influx_pubport}}"
+    org: "my-org"
+    auth_token: "{{influx_admin_token}}"
+    description: cobald tardis telegraf monitoring plugin
+    key: telegraf_cobaldtardis
+    permissions:
+    - action: write
+      resource:
+        type: buckets
+        name: "{{influx_bucket}}"
+  register: influx_telegraf_token

roles/cobald/tasks/main.yml

@@ -1,53 +1,149 @@
-- file:
+- include_vars: cobald-slurm.yml
-    path: "/container/{{item}}/cobald/"
+  when: cobald_slurm | default(False)
+  tags: always
+
+- name: build cobald:slurm docker image
+  include_role:
+    name: slurm
+    tasks_from: dockerimage
+  vars:
+    slurm_image_prefix: cobald
+    image_name: "{{cobald_image_tag}}"
+    dockerfile: "{{ lookup('template', 'cobald.Dockerfile') }}"
+    files_list:
+    - cobald-entrypoint.sh
+    - init-cobaldmodules.sh
+    - start-drone
+    - 28-sync-container-slurmd
+    files: "
+      {%- set files = [] -%} {%- for i in files_list -%}
+        {%- set files = files.append(
+          { 'dest': i, 'content': lookup('file', i) }) -%}
+        {%- endfor %}{{ files }}"
+  when: cobald_slurm | default(False)
+
+- name: build generic cobald docker image
+  include_tasks: dockerimage-generic.yml
+  when: not (cobald_slurm | default(False))
+
+- name: make cobald data volume
+  file:
+    path: "/container/volumes/cobald/"
     state: directory
     owner: "{{unpriv_user}}"
     group: docker
-  loop:
-  - docker-images
-  - volumes
 
-- copy:
+- name: copy cobald config
-    src: cobald.Dockerfile
+  copy:
-    dest: /container/docker-images/cobald/Dockerfile
-    owner: "{{unpriv_user}}"
-    group: docker
-  register: cobald_cp_dockerfile
-
-- copy:
-    src: cobald-entrypoint.sh
-    dest: /container/docker-images/cobald/cobald-entrypoint.sh
-    owner: "{{unpriv_user}}"
-    group: docker
-    mode: 0755
-  register: cobald_cp_files
-
-- docker_image:
-    name: "cobald"
-#    pull: False
-    build:
-      pull: False
-      path: "/container/docker-images/cobald/"
-    source: build
-    force_source: "{{cobald_cp_dockerfile.changed or cobald_cp_files.changed}}"
-
-- copy:
     src: cobald-config/
-    dest: /container/volumes/cobald
+    dest: "~{{unpriv_user}}/cobald/"
+    force: False
     owner: "{{unpriv_user}}"
     group: docker
-  when: False
+    mode: "0644"
+
+- name: ensure network for cobald container exists
+  docker_network:
+    name: "{{cobald_docker_network}}"
+    state: present
 
 # docker run -v $(pwd)/cobald-config-host:/etc/cobald -v $(pwd)/cobald:/cobald --rm -it cobald  bash
 
-- docker_container:
+- name: install git
-    name: cobald
+  yum:
-    image: cobald
+    name: git
+    state: present
+
+- name: make directories for cobald configuration and modules
+  file:
+    path: "{{item}}"
+    owner: "{{unpriv_user}}"
+    group: "{{unpriv_user}}"
+    mode: "a=rx,u=rwx"
+    state: directory
+  with_items:
+  - "~{{unpriv_user}}/cobald/modules"
+  - "~{{unpriv_user}}/cobald"
+
+- name: clone cobald code from git
+  git:
+    repo: https://github.com/thoto/cobald
+    dest: "~{{unpriv_user}}/cobald-src"
+    version: bugfix/mixed_construction_methods
+    update: no  # FIXME
+  become: yes
+  become_user: "{{unpriv_user}}"
+  register: cobald_git_pull
+
+- name: clone tardis code from git
+  git:
+    repo: https://github.com/MatterMiners/tardis
+    dest: "~{{unpriv_user}}/tardis-src"
+    version: master
+    update: no  # FIXME
+  become: yes
+  become_user: "{{unpriv_user}}"
+  register: tardis_git_pull
+
+- name: "get unpriv_user {{unpriv_user}} uid and gid"
+  getent:
+    database: passwd
+    key: "{{unpriv_user}}"
+
+- name: run pip install on cobald and tardis
+  docker_container:
+    image: "cobald:{{cobald_image_tag|default('latest')}}"
+    name: "cobald-src-{{item.name}}-install"
     volumes:
-    - /container/volumes/cobald:/etc/cobald:ro
+    - "~{{unpriv_user}}/{{item.name}}-src:/usr/local/src/{{item.name}}:rw"
+    state: started
+    detach: False
+    cleanup: True
+    user: "{{getent_passwd[unpriv_user][1]}}:{{getent_passwd[unpriv_user][2]}}"
+    entrypoint: ""
+    command: |
+      bash -c 'HOME=/tmp pip3 install --editable /usr/local/src/{{item.name}}'
+  with_items:
+  - name: cobald
+    run: "{{cobald_git_pull.changed}}"
+  - name: tardis
+    run: "{{tardis_git_pull.changed}}"
+  when: item.run
+
+- import_tasks: telegraf.yml
+
+- name: get cobald hostname
+  include_tasks: facts.yml
+  when: cobald_container_hostname is not defined
+
+- name: build singularity container
+  include_tasks:
+    file: singularity.yml
+    apply:
+      tags: singularity
+  tags: singularity
+
+- name: run cobald container
+  docker_container:
+    name: "{{ container_name | default('cobald') }}"
+    image: "cobald:{{cobald_image_tag|default('latest')}}"
+    hostname: "{{cobald_container_hostname}}"
+    domainname: "{{ cobald_domainname | default('cobald.local')}}"
+    volumes: "{{default_mounts + cobald_mounts }}"
+    networks:
+    - name: "{{cobald_docker_network}}"
+    networks_cli_compatible: True
     state: started
     detach: True
     cleanup: True
-  when: False
+    interactive: True
-
+    # command: python3 -m cobald.daemon /etc/cobald/config.yaml
-- include_tasks: dev.yml
+  vars:
+    default_mounts: "{{cobald_slurm_mounts | default([])}}"
+    cobald_mounts:
+    - "~{{unpriv_user}}/cobald:/etc/cobald"
+    # - /container/volumes/cobald:/etc/cobald:ro
+    - "/container/docker-images/sing-slurmd/build/:/slurm-singimage/:ro"
+    - "~{{unpriv_user}}/cobald/modules:/usr/local/src/cobaldmodules"
+    - "~{{unpriv_user}}/cobald-src:/usr/local/src/cobald:ro"
+    - "~{{unpriv_user}}/tardis-src:/usr/local/src/tardis:ro"

roles/cobald/tasks/singularity.yml

@@ -0,0 +1,48 @@
+- name: setup singularity
+  import_role: name="singularity"
+  tags: singularity
+
+- name: make singularity image build directory
+  file:
+    state: directory
+    path: "{{item}}"
+    owner: "{{unpriv_user}}"
+    group: "docker"
+    mode: "0755"
+  loop:
+    - /container/docker-images/sing-slurmd
+    - /container/docker-images/sing-slurmd/cache
+    - /container/docker-images/sing-slurmd/build
+
+- name: copy slurm singularity container files
+  copy:
+    src: "{{item}}"
+    dest: "/container/docker-images/sing-slurmd/{{item}}"
+    owner: "{{unpriv_user}}"
+    group: "docker"
+  loop:
+    - slurm-slurmd.def
+    - 31-slurmd-configless
+    - cgroup.conf.noautomount
+  register: cobald_copy_sing_files
+
+- name: remove old container
+  file:
+    path: /container/docker-images/sing-slurmd/build/slurmd.sif
+    state: absent
+  when: cobald_copy_sing_files.changed
+
+- name: build container
+  shell:
+    chdir: /container/docker-images/sing-slurmd/
+    cmd: SINGULARITY_TMPDIR=/container/docker-images/sing-slurmd/cache
+      singularity build --disable-cache
+      /container/docker-images/sing-slurmd/build/slurmd.sif
+      /container/docker-images/sing-slurmd/slurm-slurmd.def
+    creates: /container/docker-images/sing-slurmd/build/slurmd.sif
+  register: cobald_sing_build
+
+- debug: msg="{{[cobald_sing_build.stdout, cobald_sing_build.stderr]}}"
+  tags: [ never, debug ]
+
+# TODO: trigger copy in cobald container when slurmd.sif rebuilt

roles/cobald/tasks/telegraf.yml

@@ -0,0 +1,69 @@
+- name: setup directories for telegraf
+  file:
+    path: "/container/{{item}}/telegraf/"
+    state: directory
+    owner: "{{unpriv_user}}"
+    group: docker
+  loop:
+  - docker-images
+  - volumes
+
+- name: copy telegraf Dockerfile
+  copy:
+    src: telegraf.Dockerfile
+    dest: /container/docker-images/telegraf/Dockerfile
+    owner: "{{unpriv_user}}"
+    group: docker
+  register: cobald_cp_telegraf_dockerfile
+
+- name: copy telegraf repo file
+  copy:  # telegraf is found in influxdb repo
+    src: influxdb.repo
+    dest: /container/docker-images/telegraf/influxdb.repo
+    owner: "{{unpriv_user}}"
+    group: docker
+
+- name: docker image for telegraf
+  docker_image:
+    name: "ed-telegraf"
+    build:
+      pull: False
+      path: "/container/docker-images/telegraf/"
+    source: build
+    force_source: "{{cobald_cp_telegraf_dockerfile.changed}}"
+
+- import_tasks: influxdb.yml
+  tags: influxdb
+
+- name: generate telegraf config
+  template:
+    src: telegraf.conf.j2
+    dest: /container/volumes/telegraf/telegraf.conf
+    owner: "{{unpriv_user}}"
+    group: docker
+  vars:
+    influx_token: "{{influx_telegraf_token.token}}"
+    influx_url: "http://{{cobald_influx_hostname}}:8086"
+  register: telegraf_config_gen
+
+- name: run telegraf container
+  docker_container:
+    name: ed-telegraf
+    image: ed-telegraf
+    hostname: telegraf
+    domainname: "{{ cobald_domainname }}"
+    networks:
+    - name: "{{ cobald_docker_network }}"
+      aliases: ["ed-telegraf"]
+    volumes:
+    - "/container/volumes/telegraf/telegraf.conf:/etc/telegraf/telegraf.conf:ro"
+    state: started
+    recreate: "{{ telegraf_config_gen.changed | default(False) | bool }}"
+    detach: True
+#   cleanup: True
+    networks_cli_compatible: True
+
+- import_tasks: grafana.yml
+  vars:
+    influx_url: "http://{{cobald_influx_hostname}}:8086"
+  tags: influxdb

roles/cobald/templates/cobald.Dockerfile

@@ -1,4 +1,4 @@
-FROM docker.io/library/centos:7
+FROM {{ cobald_docker_base_image | default("docker.io/library/centos:7") }}
 
 RUN yum update -y && \
 	yum install -y python3 git && pip3 install --upgrade pip && \
@@ -11,7 +11,9 @@ RUN git clone $REPOCOBALD /usr/local/src/cobald && \
 	git clone $REPOTARDIS /usr/local/src/tardis
 
 RUN mkdir /etc/cobald /var/log/cobald && \
-	useradd -m -d /var/lib/cobald --no-log-init --system cobald
+	( getent passwd cobald > /dev/null || \
+		useradd -m -d /var/lib/cobald --no-log-init --system cobald ) && \
+	chown cobald:cobald /var/log/cobald
 
 #RUN mkdir /cobald && python3 -m venv /cobald && source /cobald/bin/activate &&\
 #	pip3 install --upgrade pip && pip3 install cobald
@@ -40,12 +42,32 @@ VOLUME /usr/local/src/cobaldmodules
 
 VOLUME /etc/cobald
 
-COPY cobald-entrypoint.sh /usr/local/sbin/docker-entrypoint.sh
+RUN mkdir -p /usr/local/lib/entrypoints.d/
 
-RUN chmod 755 /usr/local/sbin/docker-entrypoint.sh
+COPY init-cobaldmodules.sh /usr/local/lib/entrypoints.d/50-init-cobaldmodules.sh
 
-ENTRYPOINT /usr/local/sbin/docker-entrypoint.sh
+RUN chmod 755 /usr/local/lib/entrypoints.d/50-init-cobaldmodules.sh
+
+COPY start-drone /usr/local/bin/start-drone
+COPY 28-sync-container-slurmd /etc/docker-init.d/28-sync-container-slurmd
+RUN chmod 755 /usr/local/bin/start-drone /etc/docker-init.d/28-sync-container-slurmd
+
+RUN echo -e "#!/bin/sh\npython3 -m cobald.daemon /etc/cobald/config.yaml" >> /etc/docker-init.d/70-cobald && chmod 755 /etc/docker-init.d/70-cobald
+
+{% if cobald_docker_default_command | default(True) -%}
+COPY cobald-entrypoint.sh /usr/local/sbin/cobald-entrypoint.sh
+
+RUN chmod 755 /usr/local/sbin/cobald-entrypoint.sh
+
+ENTRYPOINT [ "/usr/local/sbin/cobald-entrypoint.sh" ]
+
+RUN yum -y install iproute &&\
+    yum clean all && rm -rf /var/cache/yum
 
 USER cobald
 
-CMD "python3 -m cobald.daemon /etc/cobald/config.yaml"
+STOPSIGNAL SIGINT
+
+# CMD "python3 -m cobald.daemon /etc/cobald/config.yaml"
+CMD /etc/docker-init.d/60-cobald
+{%- endif %}

roles/cobald/vars/cobald-slurm.yml

@@ -0,0 +1,12 @@
+cobald_image_tag: slurm
+cobald_docker_base_image: "{{slurm.base_image}}"
+cobald_docker_default_command: False
+cobald_docker_network: "{{slurm.network}}"
+cobald_domainname: "{{slurm.domain}}"
+cobald_slurm_mounts: "{{slurm.mounts}}"
+#- /container/volumes/slurm/:/etc/slurm/:rw
+##- "{{slurm_cfg_path | mandatory}}:/etc/slurm/:rw"
+#- /container/volumes/munge/munge.key:/etc/munge/munge.key:rw
+## - "{{slurm_munge_path | mandatory}}:/etc/munge/munge.key:rw"
+#- slurm-shared:/shared/:rw
+## - "{{slurm_shared_path | mandatory}}:{{slurm_shared_target | default('/shared')}}:rw"

roles/cobald/vars/main.yml

@@ -0,0 +1,2 @@
+cobald_docker_network: "{{docker_network}}"
+cobald_influx_hostname: "ed-influxdb"

roles/singularity/tasks/main.yml

@@ -8,7 +8,7 @@
     value: "15000"
     sysctl_file: /etc/sysctl.d/90-max_net_namespaces.conf
 
-- name: "enable user thoto for fakeroot access"
+- name: "enable user {{unpriv_user}} for fakeroot access"
   lineinfile:
     line: "{{unpriv_user}}:4294836224:65536"
     dest: "{{item}}"

roles/slurm/defaults/main.yml

@@ -1,5 +1,9 @@
-container_privileged: False
 slurm_user: slurm
 slurm_log_path_ctld: /var/log/slurm/slurmctld.log
 slurm_log_path_d: /var/log/slurm/slurmd.log
 slurm_log_path_sched: /var/log/slurm/slurmsched.log
+slurm_prefix: slurm
+slurm_domain: slurm.local
+container_privileged: False
+docker_network: slurm
+slurm_image_prefix: slurm

roles/slurm/files/docker-init

@@ -0,0 +1,18 @@
+#!/bin/bash
+function trp_term(){
+        echo pkill -P $pids
+        for j in $pids ; do
+                pkill -P $j
+                kill -SIGTERM $j
+        done
+}
+trap trp_term SIGINT SIGTERM
+pids=""
+for i in /etc/docker-init.d/* ; do
+        [ ! -f $i ] && break
+        $i &
+        pids="$pids $!"
+done
+wait $pids
+
+# TODO: call start scripts like "foo.sh start" and "foo.sh stop" to avoid pkill

roles/slurm/files/entry-munge.sh

@@ -0,0 +1,7 @@
+#!/usr/bin/env bash
+set -e
+
+if [ -f "/etc/munge/munge.key" ] ; then
+	chown munge:munge /etc/munge/munge.key
+	chmod 600 /etc/munge/munge.key
+fi

roles/slurm/files/entrypoint.sh

@@ -1,9 +1,8 @@
 #!/usr/bin/env bash
 set -e
 
-if [ -f "/etc/munge/munge.key" ] ; then
+for i in /usr/local/lib/entrypoints.d/* ; do
-	chown munge:munge /etc/munge/munge.key
+    [ -f $i ] && /bin/sh $i || break
-	chmod 600 /etc/munge/munge.key
+done
-fi
 
-exec "$@"
+exec "${@:-/bin/bash}"

roles/slurm/files/slurm-base.Dockerfile

@@ -0,0 +1,35 @@
+FROM docker.io/library/centos:7
+
+RUN yum install -y epel-release && \
+	yum install -y slurm && \
+	yum clean all && rm -rf /var/cache/yum
+
+RUN	yum install -y less iproute bind-utils nmap-ncat net-tools && \
+	yum clean all && rm -rf /var/cache/yum
+
+RUN mkdir -p /usr/local/lib/entrypoints.d/
+
+COPY --chown=root:root entry-munge.sh /usr/local/lib/entrypoints.d/10-munge.sh
+COPY --chown=root:root entrypoint.sh /usr/local/sbin/entrypoint.sh
+
+RUN chmod 755 /usr/local/lib/entrypoints.d/10-munge.sh && \
+	chmod 755 /usr/local/sbin/entrypoint.sh
+
+ENTRYPOINT [ "/usr/local/sbin/entrypoint.sh" ]
+
+ARG slurmuser=slurm
+ENV slurmuser=${slurmuser}
+
+RUN useradd -d /var/lib/slurm -m --no-log-init --system $slurmuser &&\
+	slurm-setuser -u $slurmuser -g $slurmuser -y
+
+COPY docker-init /usr/local/sbin/docker-init
+RUN mkdir /etc/docker-init.d  && chmod 755 /usr/local/sbin/docker-init
+COPY start-scripts/10-munge /etc/docker-init.d/10-munge
+RUN chmod 755 /etc/docker-init.d/10-munge
+
+ARG moreusers
+RUN function mu { [ -z "$1" ] || useradd -d $2 -m --no-log-init --system $1 ;};\
+	echo "${moreusers}" | tr ',' '\n' | while read i ; do mu $i ; done
+
+CMD /usr/local/sbin/docker-init

roles/slurm/files/slurm.Dockerfile

@@ -1,43 +0,0 @@
-FROM docker.io/library/centos:7 as base
-
-RUN yum install -y epel-release && \
-	yum install -y slurm && \
-	yum clean all && rm -rf /var/cache/yum
-
-RUN	yum install -y less iproute bind-utils nmap-ncat net-tools && \
-	yum clean all && rm -rf /var/cache/yum
-
-COPY entrypoint.sh /usr/local/sbin/entrypoint.sh
-
-RUN chown root:root /usr/local/sbin/entrypoint.sh && \
-	chmod 755 /usr/local/sbin/entrypoint.sh
-
-ENTRYPOINT [ "/usr/local/sbin/entrypoint.sh" ]
-
-ARG slurmuser=slurm
-ENV slurmuser=${slurmuser}
-
-RUN useradd -d /var/lib/slurm -m --no-log-init --system $slurmuser &&\
-	slurm-setuser -u $slurmuser -g $slurmuser -y
-
-ENV SLURMCTLD_LOG_PATH="/var/log/slurm/slurmctld.log"
-ENV SLURMD_LOG_PATH="/var/log/slurm/slurmd.log"
-ENV SLURM_SCHED_LOG_PATH="/var/log/slurm/slurmsched.log"
-
-FROM base as slurmd
-
-RUN	yum install -y slurm-slurmd && \
-	yum clean all && rm -rf /var/cache/yum
-
-CMD bash -c 'cat <({ su -s /bin/sh -c "munged -F" munge & \
-	slurmd -D 2>/dev/null 1>/dev/null & \
-	tail --retry --pid $! -f ${SLURMD_LOG_PATH} ${SLURM_SCHED_LOG_PATH} & })'
-
-FROM base as slurmctld
-
-RUN	yum install -y slurm-slurmctld && \
-	yum clean all && rm -rf /var/cache/yum
-
-CMD bash -c 'cat <({ su -s /bin/sh -c "munged -F" munge & \
-	su -s /bin/sh -c "slurmctld -D" ${slurmuser} 2>/dev/null 1>/dev/null & \
-	tail --retry --pid $! -f ${SLURMCTLD_LOG_PATH} ${SLURM_SCHED_LOG_PATH} & })'

roles/slurm/files/slurmctld.Dockerfile

@@ -1,32 +1,11 @@
-FROM docker.io/library/centos:7 as base
+FROM slurm:base
-
-RUN yum install -y epel-release && \
-	yum install -y slurm && \
-	yum clean all && rm -rf /var/cache/yum
-
-RUN	yum install -y less iproute bind-utils nmap-ncat net-tools && \
-	yum clean all && rm -rf /var/cache/yum
-
-COPY entrypoint.sh /usr/local/sbin/entrypoint.sh
-
-RUN chown root:root /usr/local/sbin/entrypoint.sh && \
-	chmod 755 /usr/local/sbin/entrypoint.sh
-
-ENTRYPOINT [ "/usr/local/sbin/entrypoint.sh" ]
-
-ARG slurmuser=slurm
-ENV slurmuser=${slurmuser}
-
-RUN useradd -d /var/lib/slurm -m --no-log-init --system $slurmuser &&\
-	slurm-setuser -u $slurmuser -g $slurmuser -y
 
 RUN	yum install -y slurm-slurmctld && \
 	yum clean all && rm -rf /var/cache/yum
 
+COPY start-scripts/20-slurmctld /etc/docker-init.d/20-slurmctld
+RUN chmod 755 /etc/docker-init.d/20-slurmctld
+
 ENV SLURMCTLD_LOG_PATH="/var/log/slurm/slurmctld.log"
 ENV SLURMD_LOG_PATH="/var/log/slurm/slurmd.log"
 ENV SLURM_SCHED_LOG_PATH="/var/log/slurm/slurmsched.log"
-
-CMD bash -c 'cat <({ su -s /bin/sh -c "munged -F" munge & \
-	su -s /bin/sh -c "slurmctld -D" ${slurmuser} 2>/dev/null 1>/dev/null & \
-	tail --retry --pid $! -f ${SLURMCTLD_LOG_PATH} ${SLURM_SCHED_LOG_PATH} & })'

roles/slurm/files/slurmd.Dockerfile

@@ -1,32 +1,14 @@
-FROM docker.io/library/centos:7
+FROM slurm:base
-
-RUN yum install -y epel-release && \
-	yum install -y slurm && \
-	yum clean all && rm -rf /var/cache/yum
-
-RUN	yum install -y less iproute bind-utils nmap-ncat net-tools && \
-	yum clean all && rm -rf /var/cache/yum
-
-COPY entrypoint.sh /usr/local/sbin/entrypoint.sh
-
-RUN chown root:root /usr/local/sbin/entrypoint.sh && \
-	chmod 755 /usr/local/sbin/entrypoint.sh
-
-ENTRYPOINT [ "/usr/local/sbin/entrypoint.sh" ]
-
-ARG slurmuser=slurm
-ENV slurmuser=${slurmuser}
-
-RUN useradd -d /var/lib/slurm -m --no-log-init --system $slurmuser &&\
-	slurm-setuser -u $slurmuser -g $slurmuser -y
 
 RUN yum install -y slurm-slurmd && \
 	yum clean all && rm -rf /var/cache/yum
 
+COPY start-scripts/30-slurmd /etc/docker-init.d/30-slurmd
+RUN chmod 755 /etc/docker-init.d/30-slurmd
+
 ENV SLURMCTLD_LOG_PATH="/var/log/slurm/slurmctld.log"
 ENV SLURMD_LOG_PATH="/var/log/slurm/slurmd.log"
 ENV SLURM_SCHED_LOG_PATH="/var/log/slurm/slurmsched.log"
 
-CMD bash -c 'cat <({ su -s /bin/sh -c "munged -F" munge & \
+RUN yum install -y singularity && \
-	slurmd -D 2>/dev/null 1>/dev/null & \
+	yum clean all && rm -rf /var/cache/yum
-	tail --retry --pid $! -f ${SLURMD_LOG_PATH} ${SLURM_SCHED_LOG_PATH} & })'

roles/slurm/files/start-scripts/10-munge

@@ -0,0 +1,2 @@
+#!/bin/sh
+exec su -s /bin/sh -c "munged -F" munge

roles/slurm/files/start-scripts/20-slurmctld

@@ -0,0 +1,4 @@
+#!/bin/sh
+su -s /bin/sh -c "slurmctld -D" ${slurmuser} 2>/dev/null 1>/dev/null &
+tail --retry --pid $! -f ${SLURMCTLD_LOG_PATH} ${SLURM_SCHED_LOG_PATH}
+

roles/slurm/files/start-scripts/30-slurmd

@@ -0,0 +1,4 @@
+#!/bin/sh
+slurmd -D 2>/dev/null 1>/dev/null &
+tail --retry --pid $! -f ${SLURMD_LOG_PATH} ${SLURM_SCHED_LOG_PATH}
+

roles/slurm/handlers/main.yml

@@ -1,3 +1,3 @@
 - name: reconfigure slurm 
-  command:
+  shell:
-    cmd: docker container exec -it slurm-ctl scontrol reconfigure
+    cmd: "docker container exec -it {{slurm_prefix}}-ctl scontrol reconfigure || docker container restart {{slurm_prefix}}-ctl && docker container exec -it {{slurm_prefix}}-ctl scontrol reconfigure"

roles/slurm/tasks/dockerimage.yml

@@ -1,31 +1,29 @@
-- file:
+- name: build slurm base docker image
-    path: "/container/docker-images/{{item}}"
+  include_tasks: dockerimage_build.yml
-    state: directory
+  vars:
-    owner: "{{unpriv_user}}"
+    slurm_image_prefix: "{{slurm_base_image_prefix | default('slurm') }}"
-    group: docker
+    image_name: base
+    dockerfile: "{{lookup('file', 'slurm-base.Dockerfile')}}"
+    files:
+    - dest: entrypoint.sh
+      content: "{{ lookup('file', 'entrypoint.sh') }}"
+    - dest: entry-munge.sh
+      content: "{{ lookup('file', 'entry-munge.sh') }}"
+    - dest: docker-init
+      content: "{{ lookup('file', 'docker-init') }}"
+    - dest: start-scripts/10-munge
+      content: "{{ lookup('file', 'start-scripts/10-munge') }}"
+    image_args:
+      moreusers: >-
+        {% for a in slurm_user_accounts | default([]) -%}
+        {{a['name']}} {{a['dir']}}{{loop.last | ternary('',',')}}
+        {%- endfor %}
+  when: not slurm_baseimg_build_chg | default(False)
 
-- copy:
+- set_fact:
-    src: "{{item}}.Dockerfile"
+    slurm_baseimg_build_chg:
-    dest: "/container/docker-images/{{item}}/Dockerfile"
+      "{{(slurm_baseimg_build_chg | default(False)) or
-    owner: "{{unpriv_user}}"
+         slurm_img_build.changed}}"
-    group: docker
-  register: slurm_cp_dockerfile
-
-- copy:
-    src: "entrypoint.sh"
-    dest: "/container/docker-images/{{item}}/entrypoint.sh"
-    owner: root
-    group: root
-    mode: u=rwx,g=rx,o=rx
-  register: slurm_cp_entrypt
-
-- docker_image:
-    name: "slurm-{{item}}"
-#    pull: False
-    build:
-      pull: False
-      path: "/container/docker-images/{{item}}"
-#      target: "{{item}}" # unsupported on old docker-py versions as in el7 
-    source: build
-    force_source: "{{slurm_cp_dockerfile.changed or slurm_cp_entrypt.changed}}"
 
+- name: "build slurm base docker image {{image_name}}"
+  include_tasks: dockerimage_build.yml

roles/slurm/tasks/dockerimage_build.yml

@@ -0,0 +1,43 @@
+- name: create directories for docker image build
+  file:
+    path: "/container/docker-images/{{slurm_image_prefix}}-{{image_name}}/{{item}}"
+    state: directory
+    owner: "{{unpriv_user}}"
+    group: docker
+  loop: "{{ [''] + (files | map(attribute='dest') | map('dirname') |
+                    unique | select | list) }}"
+
+- name: "copy Dockerfile {{slurm_image_prefix}}:{{image_name}}"
+  copy:
+    content: "{{dockerfile}}"
+    dest: "/container/docker-images/{{slurm_image_prefix}}-{{image_name}}/Dockerfile"
+    owner: "{{unpriv_user}}"
+    group: docker
+  register: slurm_cp_dockerfile
+
+- name: copy requisite files
+  copy:
+    content: "{{ item.content }}"
+    dest: "/container/docker-images/{{slurm_image_prefix}}-{{image_name}}/{{item.dest}}"
+    owner: root
+    group: root
+    mode: u=rwx,g=rx,o=rx
+  loop: "{{ files | default([]) }}"
+  loop_control:
+    label: "{{ item.dest }}"
+  register: slurm_cp_files
+
+- name: "build docker image {{slurm_image_prefix}}:{{image_name}}"
+  docker_image:
+    name: "{{slurm_image_prefix}}"
+    tag: "{{image_name}}"
+#    pull: False
+    build:
+      args: "{{image_args | default(omit)}}"
+      pull: False
+      path: "/container/docker-images/{{slurm_image_prefix}}-{{image_name}}/"
+    source: build
+    force_source: "{{slurm_cp_dockerfile.changed or
+                     slurm_cp_files.changed or
+                     slurm_baseimg_build_chg | default(False) }}"
+  register: slurm_img_build

roles/slurm/tasks/host-config.yml

@@ -0,0 +1,35 @@
+# TODO: this does not work quite right since slurm-ctl does not reach the host
+# system. sinfo, scontrol etc. work but srun does not!
+
+- name: "get addresses from docker network"
+  docker_network_info:
+    name: "{{ docker_network }}"
+  register: slurm_network_data
+
+- name: link host slurm config
+  file:
+    path: "/etc/slurm/slurm.conf"
+    src: "/container/volumes/slurm/slurm.conf"
+    force: True
+    state: link
+    backup: True
+
+- name: create slurm user
+  user:
+    name: slurm
+    system: True
+
+- name: place entry of slurm-ctl in host /etc/hosts
+  lineinfile:
+    line: "{{slurm_network_data.network.Containers | dict2items 
+      | json_query('[?value.Name==`slurm-ctl`].value.IPv4Address') | first
+      | ipaddr('address') }}\tslurm-ctl"
+    regexp: "^(\\S*)(\\s*)slurm-ctl$"
+    path: /etc/hosts
+    backup: True
+
+- name: start munge locally
+  service:
+    name: munge
+    enabled: True
+    state: started

roles/slurm/tasks/main.yml

@@ -3,10 +3,27 @@
     name: [ slurm, slurm-doc ]
     state: present
 
-- include_tasks: dockerimage.yml
+- name: build docker images for slurm
+  include_tasks:
+    file: dockerimage.yml
   loop:
-    - slurmctld
+  - name: slurmctld
-    - slurmd
+    dockerfile: "{{ lookup('file', 'slurmctld.Dockerfile') }}"
+    files:
+    - dest: start-scripts/20-slurmctld
+      content: "{{ lookup('file', 'start-scripts/20-slurmctld') }}"
+  - name: slurmd
+    dockerfile: "{{ lookup('file', 'slurmd.Dockerfile') }}"
+    files:
+    - dest: start-scripts/30-slurmd
+      content: "{{ lookup('file', 'start-scripts/30-slurmd') }}"
+  vars:
+    image_name: "{{image.name | default(omit) }}"
+    dockerfile: "{{image.dockerfile | default(omit) }}"
+    files: "{{image.files | default(omit) }}"
+  loop_control:
+    loop_var: image
+    label: "{{ image.name }}"
 
 - name: generate munge key
   shell:
@@ -20,7 +37,8 @@
     group: munge
     mode: u=rw,g=,o=
 
-- file:
+- name: create munge key directory for containers
+  file:
     path: /container/volumes/munge
     state: directory
     owner: munge
@@ -35,10 +53,18 @@
     src: /etc/munge/munge.key
     dest: /container/volumes/munge/munge.key
 
-- file:
+- name: make slurm directory
+  file:
     path: /container/volumes/slurm/
     state: directory
 
+- name: "create docker network to make service discovery work"
+  docker_network:
+    name: "{{ docker_network }}"
+    state: present
+  register: slurm_network_data
+  tags: slurm-config
+
 - name: upload slurm config
   template:
     force: true
@@ -47,59 +73,66 @@
   loop:
     - slurm.conf
     - cgroup.conf
+  vars:
+    slurm_exec_node_cores: 3
+    slurm_exec_node_mem: 5000  # RealMemory=5964
+    slurm_alloc_nodes_default:
+    - name: "{{slurm_prefix+'-submit1'}}"
+    - name: "{{ inventory_hostname }}"
+      addr: "{{ slurm_network_data.network.IPAM.Config[0].Gateway }}"
+    alloc_nodes: "{{ slurm_alloc_nodes_default + extra_nodes | default([])}}"
+    partitions:
+    - name: cobald
+      nodeprefix: drone
+      num_nodes: 10
+      node_cores: 3
+      node_mem: 4900
+      port: 16818
+      initstate: FUTURE
   notify: reconfigure slurm
-  tags: [ slurm-config ]
+  tags: slurm-config
-
-- name: "create docker network to make service discovery work"
-  docker_network:
-    name: slurm
-    state: present
 
 - name: "create docker volume for shared access between nodes"
   docker_volume:
     name: slurm-shared
     state: present
 
-- set_fact:
-    slurm_nodes: # default nodes: controller and submit machine
-    - machine: ctl
-      image: slurm-slurmctld
-    - machine: submit1
-      image: slurm-slurmd
-      extra_mounts:
-      - "/home/{{unpriv_user}}/job3/:/mnt/:rw"
-  tags: [ slurm-config ]
-    
-# TODO: reserve some address using docker_network_info and assign as aux
-# address to enable slurmctld to get a static address in order to be
-# reachable from slurm running on docker host to enable submitting jobs.
-
 - name: run slurm docker containers
   docker_container:
-    name: "slurm-{{item.machine}}"
+    name: "{{ slurm_prefix }}-{{ item.machine }}"
-    hostname: "slurm-{{item.machine}}"
+    hostname: "{{ slurm_prefix }}-{{ item.machine }}"
-    domainname: "slurm.local"
+    domainname: "{{ slurm_domain }}"
-    volumes: "{{default_mounts + ( item.extra_mounts | default([]) ) }}"
+    volumes: "{{ slurm_default_mounts + ( item.extra_mounts | default([]) ) }}"
+    ports: "{{ item.exposed_ports | default([]) }}"
     networks:
-    - name: "slurm"
+    - name: "{{ docker_network }}"
+      aliases: "{{ item.aliases | default(omit) }}"
     env:
-      slurmuser: "{{slurm_user}}"
+      slurmuser: "{{ slurm_user }}"
-    image: "{{item.image}}"
+    image: "{{ item.image }}"
     state: started
     detach: True
     cleanup: True
     privileged: "{{ container_privileged | bool }}"
     networks_cli_compatible: True
-  vars:
+    interactive: True
-    default_mounts:
+  vars:  # see vars/main.yml
-    - /container/volumes/slurm/:/etc/slurm/:rw
+    slurm_nodes_all: "{{ slurm_nodes_exec + slurm_nodes_std }}"
-    - /container/volumes/munge/munge.key:/etc/munge/munge.key:rw
+  loop: "{{ slurm_nodes_all }}"
-    - slurm-shared:/shared/:rw
+  loop_control:
-    slurm_nodes_all: | # add execute nodes
+    label: "{{slurm_prefix}}-{{ item.machine }}"
-      {% for i in range(1, 4) -%}
+  tags: slurm-config
-        {% set _ = slurm_nodes.extend([
+
-          {'machine':'exec%s'|format(i), 'image': 'slurm-slurmd'}]) -%}
+- name: configure host system to integrate into slurm cluster
-      {%- endfor %}
+  import_tasks: host-config.yml
-      {{ slurm_nodes }}
+  when: slurm_hostsystem_cluster_access | default(False)
-  loop: "{{slurm_nodes_all}}"
+
-  tags: [ slurm-config ]
+- name: export facts about slurm cluster to be used by other modules
+  set_fact:
+    slurm:
+      user: "{{slurm_user}}"
+      domain: "{{slurm_domain}}"
+      base_image: "slurm:base"
+      mounts: "{{slurm_default_mounts}}"
+      network: "{{docker_network}}"
+  tags: always

roles/slurm/templates/slurm.conf.j2

@@ -9,6 +9,8 @@ ControlMachine=slurm-ctl
 AuthType=auth/munge
 #CheckpointType=checkpoint/none
 CryptoType=crypto/munge
+CommunicationParameters=NoAddrCache
+SlurmctldParameters=enable_configless
 #DisableRootJobs=NO
 #EnforcePartLimits=NO
 #Epilog=
@@ -103,8 +105,10 @@ Waittime=0
 #MaxMemPerCPU=0
 #SchedulerTimeSlice=30
 SchedulerType=sched/backfill
-SelectType=select/linear
+# SelectType=select/linear
+SelectType=select/cons_res
 #SelectTypeParameters=
+SelectTypeParameters=CR_CORE
 #
 #
 # JOB PRIORITY
@@ -163,6 +167,27 @@ SlurmSchedLogFile={{slurm_log_path_sched}}
 #
 #
 # COMPUTE NODES
-NodeName=slurm-exec[1-{{num_nodes}}] CPUs=2 CoresPerSocket=2 State=UNKNOWN
+NodeName=slurm-exec[1-{{ num_nodes }}] CPUs={{ slurm_exec_node_cores }} {{''
-NodeName=slurm-submit1 CPUs=1 State=UNKNOWN
+	}} RealMemory={{ slurm_exec_node_mem }} {{''
-PartitionName=debug Nodes=slurm-exec[1-{{num_nodes}}] AllocNodes=slurm-submit1 Default=YES MaxTime=INFINITE State=UP
+	}} CoresPerSocket={{ slurm_exec_node_cores }} State=UNKNOWN
+{% for p in partitions | default([]) %}
+NodeName={{ p.nodeprefix }}[1-{{ p.num_nodes }}] CPUs={{ p.node_cores }} {{''
+	}} RealMemory={{ p.node_mem }} {{''
+	}} CoresPerSocket={{ p.node_cores }} {{''
+	}} {%- if p.port is defined %} Port={{ p.port}} {% endif %}{{''
+	}} State={{ p.initstate | default('UNKNOWN') }}
+{% endfor %}
+{% for i in alloc_nodes -%}
+NodeName={{i.name}}
+	{%- if i.hostname is defined %} NodeHostname={{i.hostname}} {% endif %}
+	{%- if i.addr is defined %} NodeAddr={{i.addr}} {% endif %}
+	State=UNKNOWN
+{% endfor %}
+PartitionName=container Nodes=slurm-exec[1-{{num_nodes}}] {{ ''
+	}} AllocNodes={{alloc_nodes |map(attribute='name') | join(',')}} {{ ''
+	}} Default=YES MaxTime=INFINITE State=UP
+{% for p in partitions | default([]) %}
+PartitionName={{ p.name }} Nodes={{ p.nodeprefix }}[1-{{ p.num_nodes }}] {{ ''
+	}} AllocNodes={{alloc_nodes |map(attribute='name') | join(',')}} {{ ''
+	}} MaxTime=INFINITE State=UP
+{% endfor %}

roles/slurm/vars/main.yml

@@ -0,0 +1,21 @@
+slurm_nodes_std: # default nodes: controller and submit machine
+- machine: ctl
+  image: slurm:slurmctld
+  exposed_ports: [ "6817:6817/tcp" ]
+- machine: submit1
+  image: slurm:slurmd
+  extra_mounts:
+  - "/home/{{unpriv_user}}/job3/:/mnt/:rw"
+slurm_nodes_exec: | # extend range to execute nodes list
+  {% set slurm_nodes_exec = [] %}
+  {% for i in range(1, num_nodes+1) -%}
+    {% set _ = slurm_nodes_exec.extend([
+      {'machine':'exec%s'|format(i), 'image': 'slurm:slurmd',
+       'aliases':['drone%s'|format(i)]}]) -%}
+  {%- endfor %}
+  {{ slurm_nodes_exec }}
+slurm_default_mounts:
+- /container/volumes/slurm/:/etc/slurm/:rw
+- /container/volumes/munge/munge.key:/etc/munge/munge.key:rw
+- slurm-shared:/shared/:rw
+