diff --git a/ecs/example/main.tf b/ecs/example/main.tf
index a8213283..c2ec7c81 100644
--- a/ecs/example/main.tf
+++ b/ecs/example/main.tf
@@ -103,6 +103,19 @@ module "dashboard" {
 
   name = "terraform-ecs-example"
   widgets = [
+    module.hosts.widgets.instance_scaling,
+    module.hosts.widgets.instance_states,
+    module.hosts.widgets.memory_utilization,
+    module.hosts.widgets.cpu_utilization,
+    module.hosts.widgets.cpu_credit_balance,
+    module.hosts.widgets.cpu_credit_usage,
+    module.hosts.widgets.root_fs_utilization,
+    module.hosts.widgets.root_fs_free,
+    module.hosts.widgets.fs_io_bytes,
+    module.hosts.widgets.fs_io_ops,
+    module.hosts.widgets.network_bytes,
+    module.hosts.widgets.network_packets,
+
     module.web.widgets.responses,
     module.web.widgets.response_percentages,
     module.web.widgets.response_time,
diff --git a/ecs/host_group/README.md b/ecs/host_group/README.md
index 93dd8430..51ea4b30 100644
--- a/ecs/host_group/README.md
+++ b/ecs/host_group/README.md
@@ -97,3 +97,11 @@ Creates an auto-scaling group of EC2 instances which will join the given ECS clu
 * `id`
 
     Autoscaling group id
+
+* `metrics`
+
+    Cloudwatch metrics, see [metrics.tf](./metrics.tf)
+
+* `widgets`
+
+    Cloudwatch dashboard widgets, see [widgets.tf](./widgets.tf)
diff --git a/ecs/host_group/metrics.tf b/ecs/host_group/metrics.tf
new file mode 100644
index 00000000..75ae74d3
--- /dev/null
+++ b/ecs/host_group/metrics.tf
@@ -0,0 +1,422 @@
+locals {
+  metrics = {
+    # scaling inputs
+    min_instances     = module.metrics_instance_count.out_map.min
+    max_instances     = module.metrics_instance_count.out_map.max
+    desired_instances = module.metrics_instance_count.out_map.desired
+
+    # instance counts by status
+    instances             = module.metrics_instance_count.out_map.total
+    in_service_instances  = module.metrics_instance_count.out_map.in_service
+    pending_instances     = module.metrics_instance_count.out_map.pending
+    standby_instances     = module.metrics_instance_count.out_map.standby
+    terminating_instances = module.metrics_instance_count.out_map.terminating
+
+    # memory utilization
+    memory_utilization         = module.metric_memory_utilization
+    min_memory_utilization     = module.metrics_memory_utilization.out_map.min
+    average_memory_utilization = module.metrics_memory_utilization.out_map.average
+    max_memory_utilization     = module.metrics_memory_utilization.out_map.max
+    swap_utilization           = module.metric_swap_utilization
+    min_swap_utilization       = module.metrics_swap_utilization.out_map.min
+    average_swap_utilization   = module.metrics_swap_utilization.out_map.average
+    max_swap_utilization       = module.metrics_swap_utilization.out_map.max
+
+    # cpu utilization
+    min_cpu_utilization     = module.metrics_cpu_utilization.out_map.min
+    average_cpu_utilization = module.metrics_cpu_utilization.out_map.average
+    max_cpu_utilization     = module.metrics_cpu_utilization.out_map.max
+
+    # cpu credit balance
+    cpu_credit_usage            = module.metric_cpu_credit_usage
+    cpu_credit_balance          = module.metrics_cpu_credit_balance.out_map.balance
+    cpu_surplus_credit_balance  = module.metrics_cpu_credit_balance.out_map.surplus
+    cpu_surplus_credits_charged = module.metrics_cpu_credit_balance.out_map.surplus_charged
+
+    # filesystem utilization
+    min_root_fs_utilization     = module.metrics_root_fs_utilization.out_map.min
+    average_root_fs_utilization = module.metrics_root_fs_utilization.out_map.average
+    max_root_fs_utilization     = module.metrics_root_fs_utilization.out_map.max
+    min_root_fs_free            = module.metrics_root_fs_free.out_map.min
+    average_root_fs_free        = module.metrics_root_fs_free.out_map.average
+    max_root_fs_free            = module.metrics_root_fs_free.out_map.max
+
+    # filesystem reads/writes
+    fs_bytes_read            = module.metrics_io.out_map.fs_bytes_read
+    fs_reads                 = module.metrics_io.out_map.fs_reads
+    fs_bytes_written         = module.metrics_io.out_map.fs_bytes_written
+    fs_writes                = module.metrics_io.out_map.fs_writes
+    average_fs_bytes_read    = module.metrics_io.out_map.average_fs_bytes_read
+    average_fs_reads         = module.metrics_io.out_map.average_fs_reads
+    average_fs_bytes_written = module.metrics_io.out_map.average_fs_bytes_written
+    average_fs_writes        = module.metrics_io.out_map.average_fs_writes
+
+    # network in/out
+    bytes_received           = module.metrics_io.out_map.bytes_received
+    packets_received         = module.metrics_io.out_map.packets_received
+    bytes_sent               = module.metrics_io.out_map.bytes_sent
+    packets_sent             = module.metrics_io.out_map.packets_sent
+    average_bytes_received   = module.metrics_io.out_map.average_bytes_received
+    average_packets_received = module.metrics_io.out_map.average_packets_received
+    average_bytes_sent       = module.metrics_io.out_map.average_bytes_sent
+    average_packets_sent     = module.metrics_io.out_map.average_packets_sent
+  }
+}
+
+module "cloudwatch_consts" {
+  source = "./../../cloudwatch/consts"
+}
+
+locals {
+  colors = module.cloudwatch_consts.colors
+  group_dimensions = {
+    AutoScalingGroupName = var.create ? aws_autoscaling_group.hosts[0].name : ""
+  }
+}
+
+locals {
+  metrics_instance_count = {
+    min = {
+      name  = "GroupMinSize"
+      label = "Minimum instances"
+      color = local.colors.light_grey
+    }
+    max = {
+      name  = "GroupMaxSize"
+      label = "Maximum instances"
+      color = local.colors.light_grey
+    }
+    desired = {
+      name  = "GroupDesiredCapacity",
+      label = "Desired instances",
+      color = local.colors.grey
+    }
+    total = {
+      name  = "GroupTotalInstances",
+      label = "Total instances",
+      color = local.colors.green
+    }
+    pending = {
+      name  = "GroupPendingInstances"
+      label = "Pending instances"
+      color = local.colors.orange
+    }
+    standby = {
+      name  = "GroupStandbyInstances"
+      label = "Standby instances"
+      color = local.colors.light_green
+    }
+    in_service = {
+      name  = "GroupInServiceInstances"
+      label = "In service instances"
+      color = local.colors.green
+    }
+    terminating = {
+      name  = "GroupTerminatingInstances"
+      label = "Terminating instances"
+      color = local.colors.red
+    }
+  }
+}
+
+module "metrics_instance_count" {
+  source = "./../../cloudwatch/metric/many"
+
+  vars_map = { for k, variant in local.metrics_instance_count : k => {
+    namespace  = "AWS/AutoScaling"
+    dimensions = local.group_dimensions
+    name       = variant.name
+    label      = variant.label
+    color      = variant.color
+    period     = 60
+    stat       = "Average"
+  } }
+}
+
+locals {
+  metrics_utilization_variants = {
+    min     = { stat = "Minimum", color = local.colors.light_orange }
+    average = { stat = "Average", color = local.colors.orange }
+    max     = { stat = "Maximum", color = local.colors.red }
+  }
+}
+
+module "metric_memory_utilization" {
+  source = "./../../cloudwatch/metric"
+
+  namespace  = "System/Linux"
+  dimensions = local.group_dimensions
+  name       = "MemoryUtilization"
+  label      = "Memory utilization"
+  stat       = "Sum"
+  period     = 60
+}
+
+module "metrics_memory_utilization" {
+  source = "./../../cloudwatch/metric/many"
+
+  vars_map = { for k, variant in local.metrics_utilization_variants : k => {
+    namespace  = "System/Linux"
+    dimensions = local.group_dimensions
+    name       = "MemoryUtilization"
+    label      = "${variant.stat} memory utilization"
+    color      = variant.color
+    stat       = variant.stat
+    period     = 60
+  } }
+}
+
+locals {
+  metrics_swap_utilization_variants = {
+    min     = { stat = "Minimum", color = local.colors.light_pink }
+    average = { stat = "Average", color = local.colors.pink }
+    max     = { stat = "Maximum", color = local.colors.purple }
+  }
+}
+
+module "metric_swap_utilization" {
+  source = "./../../cloudwatch/metric"
+
+  namespace  = "System/Linux"
+  dimensions = local.group_dimensions
+  name       = "SwapUtilization"
+  label      = "Swap utilization"
+  stat       = "Sum"
+  period     = 60
+}
+
+module "metrics_swap_utilization" {
+  source = "./../../cloudwatch/metric/many"
+
+  vars_map = { for k, variant in local.metrics_swap_utilization_variants : k => {
+    namespace  = "System/Linux"
+    dimensions = local.group_dimensions
+    name       = "SwapUtilization"
+    label      = "${variant.stat} swap utilization"
+    color      = variant.color
+    stat       = variant.stat
+    period     = 60
+  } }
+}
+
+module "metrics_cpu_utilization" {
+  source = "./../../cloudwatch/metric/many"
+
+  vars_map = { for k, variant in local.metrics_utilization_variants : k => {
+    namespace  = "AWS/EC2"
+    dimensions = local.group_dimensions
+    name       = "CPUUtilization"
+    label      = "${variant.stat} CPU utilization"
+    color      = variant.color
+    stat       = variant.stat
+    period     = 60
+  } }
+}
+
+module "metric_cpu_credit_usage" {
+  source = "./../../cloudwatch/metric"
+
+  namespace  = "AWS/EC2"
+  dimensions = local.group_dimensions
+  name       = "CPUCreditUsage"
+  label      = "CPU credit usage"
+  stat       = "Sum"
+  period     = 300
+}
+
+locals {
+  metrics_cpu_credit_balance = {
+    balance = {
+      name  = "CPUCreditBalance"
+      label = "CPU credit balance"
+      color = local.colors.green
+    }
+    surplus = {
+      name  = "CPUSurplusCreditBalance"
+      label = "CPU surplus credit balance"
+      color = local.colors.orange
+    }
+    surplus_charged = {
+      name  = "CPUSurplusCreditsCharged"
+      label = "CPU surplus credits charged"
+      color = local.colors.red
+    }
+  }
+}
+
+module "metrics_cpu_credit_balance" {
+  source = "./../../cloudwatch/metric/many"
+
+  vars_map = { for k, variant in local.metrics_cpu_credit_balance : k => {
+    namespace  = "AWS/EC2"
+    dimensions = local.group_dimensions
+    name       = variant.name
+    label      = variant.label
+    color      = variant.color
+    stat       = "Average"
+    period     = 300
+  } }
+}
+
+locals {
+  group_root_fs_dimensions = merge(local.group_dimensions, {
+    MountPath  = "/"
+    Filesystem = "/dev/nvme0n1p1"
+  })
+}
+
+module "metrics_root_fs_utilization" {
+  source = "./../../cloudwatch/metric/many"
+
+  vars_map = { for k, variant in local.metrics_utilization_variants : k => {
+    namespace  = "System/Linux"
+    dimensions = local.group_root_fs_dimensions
+    name       = "DiskSpaceUtilization"
+    label      = "${variant.stat} root filesystem utilization"
+    color      = variant.color
+    stat       = variant.stat
+    period     = 60
+  } }
+}
+
+locals {
+  metrics_fs_free_variants = {
+    min     = { stat = "Minimum", color = local.colors.red }
+    average = { stat = "Average", color = local.colors.orange }
+    max     = { stat = "Maximum", color = local.colors.light_orange }
+  }
+}
+
+module "metrics_root_fs_free" {
+  source = "./../../cloudwatch/metric/many"
+
+  vars_map = { for k, variant in local.metrics_fs_free_variants : k => {
+    namespace  = "System/Linux"
+    dimensions = local.group_root_fs_dimensions
+    name       = "DiskSpaceAvailable"
+    label      = "${variant.stat} root filesystem free space"
+    color      = variant.color
+    stat       = variant.stat
+    period     = 60
+  } }
+}
+
+locals {
+  io_colors = {
+    read  = local.colors.green
+    write = local.colors.orange
+  }
+
+  io_metrics = {
+    fs_bytes_read = {
+      name  = "EBSReadBytes"
+      label = "Bytes read"
+      color = local.io_colors.read
+      stat  = "Sum"
+    }
+    average_fs_bytes_read = {
+      name  = "EBSReadBytes"
+      label = "Average bytes read"
+      color = local.io_colors.read
+      stat  = "Average"
+    }
+    fs_reads = {
+      name  = "EBSReadOps"
+      label = "Reads"
+      color = local.io_colors.read
+      stat  = "Sum"
+    }
+    average_fs_reads = {
+      name  = "EBSReadOps"
+      label = "Average reads"
+      color = local.io_colors.read
+      stat  = "Average"
+    }
+    fs_bytes_written = {
+      name  = "EBSWriteBytes"
+      label = "Bytes written"
+      color = local.io_colors.write
+      stat  = "Sum"
+    }
+    average_fs_bytes_written = {
+      name  = "EBSWriteBytes"
+      label = "Average bytes written"
+      color = local.io_colors.write
+      stat  = "Average"
+    }
+    fs_writes = {
+      name  = "EBSWriteOps"
+      label = "Writes"
+      color = local.io_colors.write
+      stat  = "Sum"
+    }
+    average_fs_writes = {
+      name  = "EBSWriteOps"
+      label = "Average writes"
+      color = local.io_colors.write
+      stat  = "Average"
+    }
+    bytes_received = {
+      name  = "NetworkIn"
+      label = "Received bytes"
+      color = local.io_colors.read
+      stat  = "Sum"
+    }
+    average_bytes_received = {
+      name  = "NetworkIn"
+      label = "Average received bytes"
+      color = local.io_colors.read
+      stat  = "Average"
+    }
+    packets_received = {
+      name  = "NetworkPacketsIn"
+      label = "Received packets"
+      color = local.io_colors.read
+      stat  = "Sum"
+    }
+    average_packets_received = {
+      name  = "NetworkPacketsIn"
+      label = "Average received packets"
+      color = local.io_colors.read
+      stat  = "Average"
+    }
+    bytes_sent = {
+      name  = "NetworkOut"
+      label = "Sent bytes"
+      color = local.io_colors.write
+      stat  = "Sum"
+    }
+    average_bytes_sent = {
+      name  = "NetworkOut"
+      label = "Average sent bytes"
+      color = local.io_colors.write
+      stat  = "Average"
+    }
+    packets_sent = {
+      name  = "NetworkPacketsOut"
+      label = "Sent packets"
+      color = local.io_colors.write
+      stat  = "Sum"
+    }
+    average_packets_sent = {
+      name  = "NetworkPacketsOut"
+      label = "Average sent packets"
+      color = local.io_colors.write
+      stat  = "Average"
+    }
+  }
+}
+
+module "metrics_io" {
+  source = "./../../cloudwatch/metric/many"
+
+  vars_map = { for k, variant in local.io_metrics : k => {
+    namespace  = "AWS/EC2"
+    dimensions = local.group_dimensions
+    name       = variant.name
+    label      = variant.label
+    color      = variant.color
+    stat       = variant.stat
+    period     = 60
+  } }
+}
diff --git a/ecs/host_group/outputs.tf b/ecs/host_group/outputs.tf
index efca9ecd..1a150077 100644
--- a/ecs/host_group/outputs.tf
+++ b/ecs/host_group/outputs.tf
@@ -7,3 +7,13 @@ output "arn" {
   value       = var.create ? aws_autoscaling_group.hosts[0].arn : null
   description = "Autoscaling group ARN"
 }
+
+output "metrics" {
+  description = "Cloudwatch metrics, see [metrics.tf](./metrics.tf)"
+  value       = local.metrics
+}
+
+output "widgets" {
+  description = "Cloudwatch dashboard widgets, see [widgets.tf](./widgets.tf)"
+  value       = local.widgets
+}
diff --git a/ecs/host_group/widgets.tf b/ecs/host_group/widgets.tf
new file mode 100644
index 00000000..9c08ed51
--- /dev/null
+++ b/ecs/host_group/widgets.tf
@@ -0,0 +1,165 @@
+locals {
+  widgets = {
+    instance_scaling    = module.widget_instance_scaling
+    instance_states     = module.widget_instance_states
+    memory_utilization  = module.widget_memory_utilization
+    cpu_utilization     = module.widget_cpu_utilization
+    cpu_credit_balance  = module.widget_cpu_credit_balance
+    cpu_credit_usage    = module.widget_cpu_credit_usage
+    root_fs_utilization = module.widget_root_fs_utilization
+    root_fs_free        = module.widget_root_fs_free
+    fs_io_bytes         = module.widget_fs_io_bytes
+    fs_io_ops           = module.widget_fs_io_ops
+    network_bytes       = module.widget_network_bytes
+    network_packets     = module.widget_network_packets
+  }
+}
+
+module "widget_instance_scaling" {
+  source = "./../../cloudwatch/metric_widget"
+
+  title = "${var.name} group scaling"
+  left_metrics = [
+    local.metrics.min_instances,
+    local.metrics.max_instances,
+    local.metrics.desired_instances,
+    local.metrics.terminating_instances,
+    local.metrics.pending_instances,
+    local.metrics.in_service_instances,
+  ]
+  left_range = [0, null]
+}
+
+module "widget_instance_states" {
+  source = "./../../cloudwatch/metric_widget"
+
+  title   = "${var.name} group scaling"
+  stacked = true
+  left_metrics = [
+    local.metrics.terminating_instances,
+    local.metrics.pending_instances,
+    local.metrics.standby_instances,
+    local.metrics.in_service_instances,
+  ]
+  left_range = [0, null]
+}
+
+module "widget_memory_utilization" {
+  source = "./../../cloudwatch/metric_widget"
+
+  title = "${var.name} group memory utilization"
+  left_metrics = [
+    local.metrics.min_memory_utilization,
+    local.metrics.average_memory_utilization,
+    local.metrics.max_memory_utilization,
+    local.metrics.min_swap_utilization,
+    local.metrics.average_swap_utilization,
+    local.metrics.max_swap_utilization,
+  ]
+  left_range = [0, 100]
+}
+
+module "widget_cpu_utilization" {
+  source = "./../../cloudwatch/metric_widget"
+
+  title = "${var.name} group CPU utilization"
+  left_metrics = [
+    local.metrics.min_cpu_utilization,
+    local.metrics.average_cpu_utilization,
+    local.metrics.max_cpu_utilization,
+  ]
+  left_range = [0, 100]
+}
+
+module "widget_cpu_credit_balance" {
+  source = "./../../cloudwatch/metric_widget"
+
+  title   = "${var.name} group CPU credit balance"
+  stacked = true
+  left_metrics = [
+    local.metrics.cpu_surplus_credits_charged,
+    local.metrics.cpu_surplus_credit_balance,
+    local.metrics.cpu_credit_balance,
+  ]
+  left_range = [0, null]
+}
+
+module "widget_cpu_credit_usage" {
+  source = "./../../cloudwatch/metric_widget"
+
+  title        = "${var.name} group CPU credit usage"
+  left_metrics = [local.metrics.cpu_credit_usage]
+  left_range   = [0, null]
+}
+
+module "widget_root_fs_utilization" {
+  source = "./../../cloudwatch/metric_widget"
+
+  title = "${var.name} group root filesystem utilization"
+  left_metrics = [
+    local.metrics.min_root_fs_utilization,
+    local.metrics.average_root_fs_utilization,
+    local.metrics.max_root_fs_utilization,
+  ]
+  left_range = [0, 100]
+}
+
+module "widget_root_fs_free" {
+  source = "./../../cloudwatch/metric_widget"
+
+  title = "${var.name} group root filesystem free space"
+  left_metrics = [
+    local.metrics.min_root_fs_free,
+    local.metrics.average_root_fs_free,
+    local.metrics.max_root_fs_free,
+  ]
+  left_range = [0, null]
+}
+
+module "widget_fs_io_bytes" {
+  source = "./../../cloudwatch/metric_widget"
+
+  title   = "${var.name} group filesystem IO"
+  stacked = true
+  left_metrics = [
+    local.metrics.average_fs_bytes_written,
+    local.metrics.average_fs_bytes_read,
+  ]
+  left_range = [0, null]
+}
+
+module "widget_fs_io_ops" {
+  source = "./../../cloudwatch/metric_widget"
+
+  title   = "${var.name} group filesystem IO"
+  stacked = true
+  left_metrics = [
+    local.metrics.average_fs_writes,
+    local.metrics.average_fs_reads,
+  ]
+  left_range = [0, null]
+}
+
+module "widget_network_bytes" {
+  source = "./../../cloudwatch/metric_widget"
+
+  title   = "${var.name} group network traffic"
+  stacked = true
+  left_metrics = [
+    local.metrics.average_bytes_sent,
+    local.metrics.average_bytes_received,
+  ]
+  left_range = [0, null]
+}
+
+module "widget_network_packets" {
+  source = "./../../cloudwatch/metric_widget"
+
+  title   = "${var.name} group network traffic"
+  stacked = true
+  left_metrics = [
+    local.metrics.average_packets_sent,
+    local.metrics.average_packets_received,
+  ]
+  left_range = [0, null]
+}