diff --git a/ansible/deploy_swarm_monitoring.yml b/ansible/deploy_swarm_monitoring.yml
new file mode 100644
index 0000000..d5e11f1
--- /dev/null
+++ b/ansible/deploy_swarm_monitoring.yml
@@ -0,0 +1,18 @@
+#
+# Copyright StackHPC, 2018
+#
+---
+- name: Deploy Swarm cluster monitoring
+  hosts: cluster
+  become: yes
+  roles:
+  - role: monasca_agent_docker
+  - role: monasca_fluentd
+  - role: prometheus-docker-node
+
+- name: Deploy Prometheus server
+  hosts: master
+  become: yes
+  roles:
+  - role: prometheus-server
+
diff --git a/ansible/group_vars/all/all b/ansible/group_vars/all/all
index 078c77c..a6964d5 100644
--- a/ansible/group_vars/all/all
+++ b/ansible/group_vars/all/all
@@ -15,7 +15,7 @@ alaska_cloud: alaska
 alaska_homedir: /alaska
 alaska_softiron: 10.4.99.101
 
-# OpenStack fully qualified project name
+# OpenStack fully qualified project name (used for Grafana with domain support)
 project_name: p3@default
 
 # Virtual IP address of the controller node
@@ -28,6 +28,24 @@ alaska_monitoring_server: 10.60.253.3
 monasca_agent_p3_username: p3-monasca-agent
 monasca_agent_p3_password: "{{ vault_monasca_agent_password }}"
 
+# Monasca Fluentd config
+monasca_fluentd_log_api_uri: http://{{ controller_vip }}:5607
+monasca_fluentd_keystone_uri: http://{{ controller_vip }}:5000
+monasca_fluentd_username: "{{ monasca_agent_p3_username }}"
+monasca_fluentd_password: "{{ monasca_agent_p3_password }}"
+monasca_fluentd_project_domain_id: default
+monasca_fluentd_project_name: p3
+
+# Monasca Agent (Docker) config
+monasca_agent_docker_forwarder_port: 17120
+monasca_agent_docker_log_level: INFO
+monasca_agent_docker_api_uri: http://{{ controller_vip }}:8082/v2.0
+monasca_agent_docker_keystone_uri: http://{{ controller_vip }}:5000/v3
+monasca_agent_docker_username: "{{ monasca_agent_p3_username }}"
+monasca_agent_docker_password: "{{ monasca_agent_p3_password }}"
+monasca_agent_docker_project_name: p3
+
+
 # Local Grafana admin account for configuring Grafana
 grafana_admin_username: grafana-admin
 grafana_admin_password: "{{ vault_grafana_admin_password }}"
diff --git a/ansible/roles/monasca_agent_docker/tasks/main.yml b/ansible/roles/monasca_agent_docker/tasks/main.yml
new file mode 100644
index 0000000..46ec759
--- /dev/null
+++ b/ansible/roles/monasca_agent_docker/tasks/main.yml
@@ -0,0 +1,116 @@
+#
+# Copyright StackHPC, 2018
+#
+---
+- name: Start monasca-agent-forwarder container
+  docker_container:
+    name: monasca-agent-forwarder
+    pull: true
+    image: stackhpc/agent-forwarder:latest
+    state: started
+    volumes:
+      - plugins:/etc/monasca/agent/conf.d:ro
+    ports:
+      - "{{ monasca_agent_docker_forwarder_port }}:{{ monasca_agent_docker_forwarder_port }}"
+    env:
+      LOG_LEVEL: "{{ monasca_agent_docker_log_level }}"
+      OS_AUTH_URL: "{{ monasca_agent_docker_keystone_uri }}"
+      OS_USERNAME: "{{ monasca_agent_docker_username }}"
+      OS_PASSWORD: "{{ monasca_agent_docker_password }}"
+      OS_USER_DOMAIN_NAME: Default
+      OS_PROJECT_NAME: "{{ monasca_agent_docker_project_name }}"
+      OS_PROJECT_DOMAIN_NAME: Default
+      MONASCA_URL: "{{ monasca_agent_docker_api_uri }}"
+      SERVICE_TYPE: monitoring
+      ENDPOINT_TYPE: public
+      REGION_NAME: RegionOne
+      AGENT_HOSTNAME: "{{ ansible_hostname }}"
+      FORWARDER_URL: "http://monasca_agent-forwarder:{{ monasca_agent_docker_forwarder_port }}"
+      FORWARDER_PORT: "{{ monasca_agent_docker_forwarder_port }}"
+
+- name: Create Monasca collector plugin directory
+  file:
+    path: /etc/monasca/agent/conf.d/
+    state: directory
+    owner: root
+    group: root
+    mode: 0755
+
+- name: Template Monasca collector plugins
+  template:
+    src: "{{ item }}.j2"
+    dest: "/etc/monasca/agent/conf.d/{{ item }}"
+    owner: root
+    mode: 0644
+  with_items:
+    - cpu.yaml
+    - docker.yaml
+    - disk.yaml
+    - ib_network.yaml
+    - load.yaml
+    - memory.yaml
+    - network.yaml
+
+- name: Wait for forwarder service
+  wait_for:
+    port: "{{ monasca_agent_docker_forwarder_port }}"
+    delay: 1
+
+- name: Start monasca-agent-collector container
+  docker_container:
+    name: monasca-agent-collector
+    pull: true
+    links:
+      - monasca-agent-forwarder
+    image: stackhpc/agent-collector:latest
+    state: started
+    env:
+      DOCKER: True
+      LOG_LEVEL: "{{ monasca_agent_docker_log_level }}"
+      OS_AUTH_URL: "{{ monasca_agent_docker_keystone_uri }}"
+      OS_USERNAME: "{{ monasca_agent_docker_username }}"
+      OS_PASSWORD: "{{ monasca_agent_docker_password }}"
+      OS_USER_DOMAIN_NAME: Default
+      OS_PROJECT_NAME: "{{ monasca_agent_docker_project_name }}"
+      OS_PROJECT_DOMAIN_NAME: Default
+      MONASCA_URL: "{{ monasca_agent_docker_api_uri }}"
+      SERVICE_TYPE: monitoring
+      ENDPOINT_TYPE: public
+      REGION_NAME: RegionOne
+      AGENT_HOSTNAME: "{{ ansible_hostname }}"
+      FORWARDER_URL: "http://monasca-agent-forwarder:{{ monasca_agent_docker_forwarder_port }}"
+      FORWARDER_PORT: "{{ monasca_agent_docker_forwarder_port }}"
+    volumes:
+      - "/:/rootfs"
+      - "/var/run:/var/run:rw"
+      - "/sys:/sys:ro"
+      - "/var/lib/docker/:/var/lib/docker:ro"
+      - "/dev/disk/:/dev/disk:ro"
+      - "/etc/monasca/agent/conf.d:/etc/monasca/agent/conf.d:ro"
+
+- name: Start monasca-agent-statsd container
+  docker_container:
+    name: monasca-agent-statsd
+    pull: true
+    links:
+      - monasca-agent-forwarder
+    image: stackhpc/agent-statsd:latest
+    state: started
+    ports:
+      - "8125:8125/udp"
+    env:
+      DOCKER: True
+      LOG_LEVEL: "{{ monasca_agent_docker_log_level }}"
+      OS_AUTH_URL: "{{ monasca_agent_docker_keystone_uri }}"
+      OS_USERNAME: "{{ monasca_agent_docker_username }}"
+      OS_PASSWORD: "{{ monasca_agent_docker_password }}"
+      OS_USER_DOMAIN_NAME: Default
+      OS_PROJECT_NAME: "{{ monasca_agent_docker_project_name }}"
+      OS_PROJECT_DOMAIN_NAME: Default
+      MONASCA_URL: "{{ monasca_agent_docker_api_uri }}"
+      SERVICE_TYPE: monitoring
+      ENDPOINT_TYPE: public
+      REGION_NAME: RegionOne
+      AGENT_HOSTNAME: "{{ ansible_hostname }}"
+      FORWARDER_URL: "http://monasca-agent-forwarder:{{ monasca_agent_docker_forwarder_port }}"
+      FORWARDER_PORT: "{{ monasca_agent_docker_forwarder_port }}"
diff --git a/ansible/roles/monasca_agent_docker/templates/cpu.yaml.j2 b/ansible/roles/monasca_agent_docker/templates/cpu.yaml.j2
new file mode 100644
index 0000000..2aa6a36
--- /dev/null
+++ b/ansible/roles/monasca_agent_docker/templates/cpu.yaml.j2
@@ -0,0 +1,4 @@
+init_config: null
+instances:
+- built_by: System
+  name: cpu_stats
diff --git a/ansible/roles/monasca_agent_docker/templates/disk.yaml.j2 b/ansible/roles/monasca_agent_docker/templates/disk.yaml.j2
new file mode 100644
index 0000000..a4bb1dd
--- /dev/null
+++ b/ansible/roles/monasca_agent_docker/templates/disk.yaml.j2
@@ -0,0 +1,6 @@
+init_config: null
+instances:
+- built_by: System
+  device_blacklist_re: .*freezer_backup_snap.*
+  ignore_filesystem_types: iso9660,tmpfs
+  name: disk_stats
diff --git a/ansible/roles/monasca_agent_docker/templates/docker.yaml.j2 b/ansible/roles/monasca_agent_docker/templates/docker.yaml.j2
new file mode 100644
index 0000000..2f3f3eb
--- /dev/null
+++ b/ansible/roles/monasca_agent_docker/templates/docker.yaml.j2
@@ -0,0 +1,5 @@
+init_config:
+  docker_root: '/'
+
+instances:
+  - url: 'unix://var/run/docker.sock'
diff --git a/ansible/roles/monasca_agent_docker/templates/ib_network.yaml.j2 b/ansible/roles/monasca_agent_docker/templates/ib_network.yaml.j2
new file mode 100644
index 0000000..7ffd569
--- /dev/null
+++ b/ansible/roles/monasca_agent_docker/templates/ib_network.yaml.j2
@@ -0,0 +1,4 @@
+init_config: null
+instances:
+- built_by: IBNetworkDetect
+  name: ib_network_stats
diff --git a/ansible/roles/monasca_agent_docker/templates/load.yaml.j2 b/ansible/roles/monasca_agent_docker/templates/load.yaml.j2
new file mode 100644
index 0000000..c5ce0ca
--- /dev/null
+++ b/ansible/roles/monasca_agent_docker/templates/load.yaml.j2
@@ -0,0 +1,4 @@
+init_config: null
+instances:
+- built_by: System
+  name: load_stats
diff --git a/ansible/roles/monasca_agent_docker/templates/memory.yaml.j2 b/ansible/roles/monasca_agent_docker/templates/memory.yaml.j2
new file mode 100644
index 0000000..c9854a3
--- /dev/null
+++ b/ansible/roles/monasca_agent_docker/templates/memory.yaml.j2
@@ -0,0 +1,4 @@
+init_config: null
+instances:
+- built_by: System
+  name: memory_stats
diff --git a/ansible/roles/monasca_agent_docker/templates/network.yaml.j2 b/ansible/roles/monasca_agent_docker/templates/network.yaml.j2
new file mode 100644
index 0000000..5188355
--- /dev/null
+++ b/ansible/roles/monasca_agent_docker/templates/network.yaml.j2
@@ -0,0 +1,5 @@
+init_config: null
+instances:
+- built_by: System
+  excluded_interface_re: lo.*|vnet.*|tun.*|ovs.*|br.*|tap.*|qbr.*|qvb.*|qvo.*
+  name: network_stats
diff --git a/ansible/roles/monasca_fluentd/handlers/main.yml b/ansible/roles/monasca_fluentd/handlers/main.yml
new file mode 100644
index 0000000..e2f583c
--- /dev/null
+++ b/ansible/roles/monasca_fluentd/handlers/main.yml
@@ -0,0 +1,6 @@
+---
+- name: Restart Fluentd
+  docker_container:
+    name: fluentd
+    restart: True
+  when: fluentd_config.changed
diff --git a/ansible/roles/monasca_fluentd/tasks/main.yml b/ansible/roles/monasca_fluentd/tasks/main.yml
new file mode 100644
index 0000000..803f653
--- /dev/null
+++ b/ansible/roles/monasca_fluentd/tasks/main.yml
@@ -0,0 +1,39 @@
+#
+# Copyright StackHPC, 2018
+#
+---
+- name: Create Fluentd user
+  user:
+    name: fluent
+    state: present
+
+- name: Create Fluentd config directory
+  file:
+    path: /etc/fluentd
+    state: directory
+    owner: fluent
+    group: fluent
+    mode: 0755
+
+- name: Generate Fluentd config
+  template:
+    src: fluentd.conf.j2
+    dest: /etc/fluentd/fluentd.conf
+    owner: fluent
+    group: fluent
+    mode: 0644
+  register: fluentd_config
+  notify: Restart Fluentd
+
+- name: Start Fluentd container
+  docker_container:
+    name: fluentd
+    image: stackhpc/monasca-fluentd:latest
+    state: started
+    ports:
+      - "24224:24224"
+      - "24224:24224/udp"
+    volumes:
+      - /etc/fluentd/:/fluentd/etc:ro
+    env:
+      FLUENTD_CONF: "fluentd.conf"
diff --git a/ansible/roles/monasca_fluentd/templates/fluentd.conf.j2 b/ansible/roles/monasca_fluentd/templates/fluentd.conf.j2
new file mode 100644
index 0000000..609a003
--- /dev/null
+++ b/ansible/roles/monasca_fluentd/templates/fluentd.conf.j2
@@ -0,0 +1,45 @@
+# Accept logs from Docker Fluentd log driver
+<source>
+  @type forward
+  port 24224
+  bind 0.0.0.0
+</source>
+
+# Add a timestamp dimension to all logs to record the event time. The
+# event time is the time extracted from the log message in all cases
+# where the time_key is set, and the time the record entered fluentd
+# if no time_key is set.
+# logs.
+<filter *.**>
+  @type record_transformer
+  <record>
+    timestamp ${time}
+  </record>
+</filter>
+
+# Docker saves all logs under the 'log' field. The fluentd-monasca
+# plugin assumes that they are saved under the 'message' field. Here
+# we map the 'log' field to the 'message' field for all logs.
+<filter *.**>
+  @type record_transformer
+  enable_ruby true
+  <record>
+    message ${record["log"]}
+  </record>
+  remove_keys log
+</filter>
+
+<match *.**>
+    type copy
+    <store>
+       @type monasca
+       keystone_url {{ monasca_fluentd_keystone_uri }}
+       monasca_log_api {{ monasca_fluentd_log_api_uri }}
+       monasca_log_api_version v3.0
+       username {{ monasca_fluentd_username }}
+       password {{ monasca_fluentd_password }}
+       domain_id {{ monasca_fluentd_project_domain_id }}
+       project_name {{ monasca_fluentd_project_name }}
+    </store>
+</match>
+
diff --git a/ansible/roles/prometheus-docker-node/tasks/main.yml b/ansible/roles/prometheus-docker-node/tasks/main.yml
new file mode 100644
index 0000000..ff6d255
--- /dev/null
+++ b/ansible/roles/prometheus-docker-node/tasks/main.yml
@@ -0,0 +1,23 @@
+---
+- name: Add Prometheus node-exporter
+  docker_container:
+    name: prom-node-exporter
+    image: prom/node-exporter
+    network_mode: host  # TODO should access host fs too
+    ports:
+      - 9100:9100
+
+- name: Add cAdvisor
+  docker_container:
+    name: cAdvisor
+    image: google/cadvisor:latest
+    privileged: yes
+    detach: yes
+    ports:
+      - 8080:8080
+    volumes:
+      - /:/rootfs:ro
+      - /var/run:/var/run:rw
+      - /sys:/sys:ro
+      - /var/lib/docker/:/var/lib/docker:ro
+      - /dev/disk/:/dev/disk:ro
diff --git a/ansible/roles/prometheus-server/defaults/main.yml b/ansible/roles/prometheus-server/defaults/main.yml
new file mode 100644
index 0000000..72c8632
--- /dev/null
+++ b/ansible/roles/prometheus-server/defaults/main.yml
@@ -0,0 +1,3 @@
+---
+
+prometheus_home: /home/centos/prometheus  # TODO - better default?
diff --git a/ansible/roles/prometheus-server/files/grafana_dashboard.json b/ansible/roles/prometheus-server/files/grafana_dashboard.json
new file mode 100644
index 0000000..5d16730
--- /dev/null
+++ b/ansible/roles/prometheus-server/files/grafana_dashboard.json
@@ -0,0 +1,406 @@
+{
+  "annotations": {
+    "list": []
+  },
+  "editable": true,
+  "gnetId": null,
+  "graphTooltip": 1,
+  "hideControls": false,
+  "id": 1,
+  "links": [],
+  "refresh": false,
+  "rows": [
+    {
+      "collapse": false,
+      "height": "250px",
+      "panels": [
+        {
+          "aliasColors": {},
+          "bars": false,
+          "dashLength": 10,
+          "dashes": false,
+          "datasource": null,
+          "fill": 1,
+          "id": 1,
+          "legend": {
+            "avg": false,
+            "current": false,
+            "max": false,
+            "min": false,
+            "show": true,
+            "total": false,
+            "values": false
+          },
+          "lines": true,
+          "linewidth": 1,
+          "links": [],
+          "nullPointMode": "null",
+          "percentage": false,
+          "pointradius": 5,
+          "points": false,
+          "renderer": "flot",
+          "seriesOverrides": [],
+          "spaceLength": 10,
+          "span": 6,
+          "stack": false,
+          "steppedLine": false,
+          "targets": [
+            {
+              "expr": "irate(node_network_receive_bytes{job='node',device='eno1'}[5m])",
+              "format": "time_series",
+              "hide": true,
+              "intervalFactor": 2,
+              "legendFormat": "{{instance}} - {{device}}",
+              "refId": "A",
+              "step": 2
+            },
+            {
+              "expr": "irate(node_infiniband_port_data_received_bytes{job='node',device=\"mlx5_0\"}[5m])",
+              "format": "time_series",
+              "hide": true,
+              "intervalFactor": 2,
+              "legendFormat": "{{instance}} - {{device}}",
+              "refId": "C",
+              "step": 2
+            },
+            {
+              "expr": "irate(node_infiniband_port_data_transmitted_bytes{job='node',device=\"mlx5_0\"}[5m])",
+              "format": "time_series",
+              "intervalFactor": 2,
+              "legendFormat": "{{instance}} - {{device}}",
+              "refId": "D",
+              "step": 2
+            },
+            {
+              "expr": "irate(node_network_transmit_bytes{job='node',device='eno1'}[5m])",
+              "format": "time_series",
+              "intervalFactor": 2,
+              "legendFormat": "{{instance}} - {{device}}",
+              "refId": "E",
+              "step": 2
+            }
+          ],
+          "thresholds": [],
+          "timeFrom": null,
+          "timeShift": null,
+          "title": "Network",
+          "tooltip": {
+            "shared": true,
+            "sort": 0,
+            "value_type": "individual"
+          },
+          "type": "graph",
+          "xaxis": {
+            "buckets": null,
+            "mode": "time",
+            "name": null,
+            "show": true,
+            "values": []
+          },
+          "yaxes": [
+            {
+              "format": "Bps",
+              "label": null,
+              "logBase": 1,
+              "max": null,
+              "min": null,
+              "show": true
+            },
+            {
+              "format": "short",
+              "label": null,
+              "logBase": 1,
+              "max": null,
+              "min": null,
+              "show": true
+            }
+          ]
+        },
+        {
+          "aliasColors": {},
+          "bars": false,
+          "dashLength": 10,
+          "dashes": false,
+          "datasource": null,
+          "fill": 1,
+          "id": 4,
+          "legend": {
+            "avg": false,
+            "current": false,
+            "max": false,
+            "min": false,
+            "show": true,
+            "total": false,
+            "values": false
+          },
+          "lines": true,
+          "linewidth": 1,
+          "links": [],
+          "nullPointMode": "null",
+          "percentage": false,
+          "pointradius": 5,
+          "points": false,
+          "renderer": "flot",
+          "seriesOverrides": [],
+          "spaceLength": 10,
+          "span": 6,
+          "stack": false,
+          "steppedLine": false,
+          "targets": [
+            {
+              "expr": "irate(node_disk_io_time_ms{job='node',device!~'^(md\\\\d+$|dm-)'}[5m]) / 1000 * 100",
+              "format": "time_series",
+              "intervalFactor": 2,
+              "legendFormat": "{{instance}} - {{device}}",
+              "refId": "A"
+            }
+          ],
+          "thresholds": [],
+          "timeFrom": null,
+          "timeShift": null,
+          "title": "Disk bandwidth",
+          "tooltip": {
+            "shared": true,
+            "sort": 0,
+            "value_type": "individual"
+          },
+          "type": "graph",
+          "xaxis": {
+            "buckets": null,
+            "mode": "time",
+            "name": null,
+            "show": true,
+            "values": []
+          },
+          "yaxes": [
+            {
+              "format": "percent",
+              "label": null,
+              "logBase": 1,
+              "max": null,
+              "min": null,
+              "show": true
+            },
+            {
+              "format": "short",
+              "label": null,
+              "logBase": 1,
+              "max": null,
+              "min": null,
+              "show": true
+            }
+          ]
+        }
+      ],
+      "repeat": null,
+      "repeatIteration": null,
+      "repeatRowId": null,
+      "showTitle": false,
+      "title": "Dashboard Row",
+      "titleSize": "h6"
+    },
+    {
+      "collapse": false,
+      "height": 250,
+      "panels": [
+        {
+          "aliasColors": {},
+          "bars": false,
+          "dashLength": 10,
+          "dashes": false,
+          "datasource": null,
+          "fill": 1,
+          "id": 2,
+          "legend": {
+            "avg": false,
+            "current": false,
+            "max": false,
+            "min": false,
+            "show": true,
+            "total": false,
+            "values": false
+          },
+          "lines": true,
+          "linewidth": 1,
+          "links": [],
+          "nullPointMode": "null",
+          "percentage": false,
+          "pointradius": 5,
+          "points": false,
+          "renderer": "flot",
+          "seriesOverrides": [],
+          "spaceLength": 10,
+          "span": 6,
+          "stack": false,
+          "steppedLine": false,
+          "targets": [
+            {
+              "expr": "100 - (avg by (instance) (irate(node_cpu{job=\"node\",mode=\"idle\"}[5m])) * 100)",
+              "format": "time_series",
+              "hide": false,
+              "intervalFactor": 2,
+              "legendFormat": "{{ instance }}",
+              "refId": "A",
+              "step": 10
+            }
+          ],
+          "thresholds": [],
+          "timeFrom": null,
+          "timeShift": null,
+          "title": "CPU",
+          "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
+            }
+          ]
+        },
+        {
+          "aliasColors": {},
+          "bars": false,
+          "dashLength": 10,
+          "dashes": false,
+          "datasource": null,
+          "fill": 1,
+          "id": 3,
+          "legend": {
+            "avg": false,
+            "current": false,
+            "max": false,
+            "min": false,
+            "show": true,
+            "total": false,
+            "values": false
+          },
+          "lines": true,
+          "linewidth": 1,
+          "links": [],
+          "nullPointMode": "null",
+          "percentage": false,
+          "pointradius": 5,
+          "points": false,
+          "renderer": "flot",
+          "seriesOverrides": [],
+          "spaceLength": 10,
+          "span": 6,
+          "stack": false,
+          "steppedLine": false,
+          "targets": [
+            {
+              "expr": "node_memory_Active{job=\"node\"}",
+              "format": "time_series",
+              "intervalFactor": 2,
+              "legendFormat": "{{instance}}",
+              "refId": "A",
+              "step": 2
+            }
+          ],
+          "thresholds": [],
+          "timeFrom": null,
+          "timeShift": null,
+          "title": "Memory Usage",
+          "tooltip": {
+            "shared": true,
+            "sort": 0,
+            "value_type": "individual"
+          },
+          "type": "graph",
+          "xaxis": {
+            "buckets": null,
+            "mode": "time",
+            "name": null,
+            "show": true,
+            "values": []
+          },
+          "yaxes": [
+            {
+              "format": "bytes",
+              "label": null,
+              "logBase": 1,
+              "max": null,
+              "min": null,
+              "show": true
+            },
+            {
+              "format": "short",
+              "label": null,
+              "logBase": 1,
+              "max": null,
+              "min": null,
+              "show": true
+            }
+          ]
+        }
+      ],
+      "repeat": null,
+      "repeatIteration": null,
+      "repeatRowId": null,
+      "showTitle": false,
+      "title": "Dashboard Row",
+      "titleSize": "h6"
+    }
+  ],
+  "schemaVersion": 14,
+  "style": "dark",
+  "tags": [],
+  "templating": {
+    "list": []
+  },
+  "time": {
+    "from": "now-30m",
+    "to": "now"
+  },
+  "timepicker": {
+    "refresh_intervals": [
+      "5s",
+      "10s",
+      "30s",
+      "1m",
+      "5m",
+      "15m",
+      "30m",
+      "1h",
+      "2h",
+      "1d"
+    ],
+    "time_options": [
+      "5m",
+      "15m",
+      "1h",
+      "6h",
+      "12h",
+      "24h",
+      "2d",
+      "7d",
+      "30d"
+    ]
+  },
+  "timezone": "",
+  "title": "Spark",
+  "version": 11
+}
diff --git a/ansible/roles/prometheus-server/tasks/main.yml b/ansible/roles/prometheus-server/tasks/main.yml
new file mode 100644
index 0000000..02db899
--- /dev/null
+++ b/ansible/roles/prometheus-server/tasks/main.yml
@@ -0,0 +1,101 @@
+---
+- name: Create prometheus data dir
+  file:
+    path: "{{ prometheus_home }}/data"
+    state: directory
+    mode: 0755
+
+- name: Configure prometheus targets
+  template:
+    src: prometheus.yml  # TODO - add ip addresses properly
+    dest: "{{ prometheus_home }}/prometheus.yml"
+
+- name: Configure prometheus alerts
+  template:
+    src: alerts.rules
+    dest: "{{ prometheus_home }}/alerts.rules"
+
+- name: Add Prometheus server
+  docker_container:
+    name: prometheus
+    image: prom/prometheus
+    ports:
+      - 9090:9090
+    volumes:
+      - "{{ prometheus_home }}/prometheus.yml:/etc/prometheus/prometheus.yml"
+      - "{{ prometheus_home }}/alerts.rules:/etc/prometheus/alerts.rules"
+      - "{{ prometheus_home }}/data:/prometheus prom/prometheus"
+
+- name: Create Grafana dir
+  file:
+    path: "/var/lib/grafana"
+    state: directory
+    mode: 0755
+
+- name: Add Grafana
+  docker_container:
+    name: grafana
+    image: grafana/grafana
+    ports:
+      - 3000:3000  # TODO - add more config, github OAuth?
+    volumes:
+      - /var/lib/grafana
+
+- name: check datasource
+  uri:
+    url: "http://{{ ansible_host}}:3000/api/datasources"
+    method: GET
+    user: admin
+    password: admin
+    force_basic_auth: yes
+    return_content: yes
+  register: datasources
+
+- name: Add datasource
+  uri:
+    url: "http://{{ ansible_host}}:3000/api/datasources"
+    method: POST
+    user: admin
+    password: admin
+    force_basic_auth: yes
+    status_code: 200
+    body_format: json
+    body: '{"name": "Prometheus",
+            "type": "prometheus",
+            "access": "proxy",
+            "url": "http://{{ ansible_host }}:9090",
+            "password": "",
+            "user": "",
+            "basicAuth": false,
+            "basicAuthUser": "",
+            "basicAuthPassword": "",
+            "isDefault": true,
+            "jsonData": null }'
+  when: datasources.json == []
+
+- name: check datasource
+  uri:
+    url: "http://{{ ansible_host}}:3000/api/search?query=Spark"
+    method: GET
+    user: admin
+    password: admin
+    force_basic_auth: yes
+    return_content: yes
+  register: dashboards
+
+- debug:
+    msg: "Return value: {{ dashboards.json }}"
+
+- name: Add dashboard
+  uri:
+    url: "http://{{ ansible_host}}:3000/api/dashboards/db"
+    method: POST
+    user: admin
+    password: admin
+    force_basic_auth: yes
+    status_code: 200
+    body_format: json
+    body:
+      dashboard: '{{ lookup("file","grafana_dashboard.json") }}'
+      overwrite: true
+  when: dashboards.json == []
diff --git a/ansible/roles/prometheus-server/templates/alerts.rules b/ansible/roles/prometheus-server/templates/alerts.rules
new file mode 100644
index 0000000..a4f0a33
--- /dev/null
+++ b/ansible/roles/prometheus-server/templates/alerts.rules
@@ -0,0 +1,20 @@
+# {{ ansible_managed }}
+
+{% raw %}
+ALERT InstanceDown
+  IF up{job="node"} == 0
+  FOR 5m
+  LABELS { severity = "page" }
+  ANNOTATIONS {
+    summary = "Instance {{ $labels.instance }} down",
+    description = "{{ $labels.instance }} of job {{ $labels.job }} has been down for more than 5 minutes.",
+  }
+
+ALERT DummyInfiniBandTrafficHigh
+  IF irate(node_infiniband_port_data_received_bytes{job='node'}[5m]) > 1400
+  FOR 30s
+  LABELS { severity = "low" }
+  ANNOTATIONS {
+    summary = "Instance {{ $labels.instance }} high traffic on {{ $labels.device}}"
+  }
+{% endraw %}
diff --git a/ansible/roles/prometheus-server/templates/prometheus.yml b/ansible/roles/prometheus-server/templates/prometheus.yml
new file mode 100644
index 0000000..75538b4
--- /dev/null
+++ b/ansible/roles/prometheus-server/templates/prometheus.yml
@@ -0,0 +1,51 @@
+# {{ ansible_managed }}
+
+global:
+  scrape_interval:     15s
+  evaluation_interval: 15s
+
+  external_labels:
+      monitor: 'sahara-test'
+
+rule_files:
+  # TODO  - "alerts.rules"
+
+scrape_configs:
+  - job_name: 'prometheus'
+
+    static_configs:
+      - targets:
+{% for node in groups['master'] %}
+        - '{{ node }}:9090'
+{% endfor %}
+
+  - job_name: 'node'
+    scrape_interval: 15s
+    static_configs:
+      - targets:
+{% for node in groups['master'] %}
+        - '{{ node }}:9100'
+{% endfor %}
+{% for node in groups['slave'] %}
+        - '{{ node }}:9100'
+{% endfor %}
+
+    params:
+      collect[]:
+        - cpu
+        - meminfo
+        - diskstats
+        - netdev
+        - netstat
+        - infiniband
+
+  - job_name: 'cAdvisor'
+    scrape_interval: 15s
+    static_configs:
+      - targets:
+{% for node in groups['master'] %}
+        - '{{ node }}:8080'
+{% endfor %}
+{% for node in groups['slave'] %}
+        - '{{ node }}:8080'
+{% endfor %}