# you can disable an alarm notification by setting the 'to' line to: silent

# --- Errors ---

# Server-side errors indicate Service Bus infrastructure problems.
# AMBA: ServerErrors > 0 is Sev1

 template: am_service_bus_server_errors
       on: azure_monitor.service_bus.errors
    class: Errors
     type: Messaging
component: Azure Service Bus
   lookup: average -5m unaligned of server
    units: errors/s
    every: 1m
     warn: $this > (($status >= $WARNING)  ? (0) : (1))
     crit: $this > (($status == $CRITICAL) ? (1) : (5))
    delay: down 5m multiplier 1.5 max 1h
  summary: Service Bus server errors on ${label:resource_name}
     info: Server-side error rate on Service Bus namespace ${label:resource_name} \
           in ${label:resource_group} (${label:region})
       to: sysadmin

# Throttled requests mean the namespace has exceeded its messaging unit quotas.
# AMBA: ThrottledRequests > 0 is Sev1

 template: am_service_bus_throttled_requests
       on: azure_monitor.service_bus.errors
    class: Errors
     type: Messaging
component: Azure Service Bus
   lookup: average -5m unaligned of throttled
    units: errors/s
    every: 1m
     warn: $this > (($status >= $WARNING)  ? (0) : (1))
     crit: $this > (($status == $CRITICAL) ? (1) : (5))
    delay: down 5m multiplier 1.5 max 1h
  summary: Service Bus throttled requests on ${label:resource_name}
     info: Rate of throttled requests on Service Bus namespace ${label:resource_name} \
           in ${label:resource_group} (${label:region}). \
           Indicates the namespace is exceeding its messaging unit quotas
       to: sysadmin

# User errors (400-class) at sustained high rate may indicate
# client misconfiguration or malformed messages.

 template: am_service_bus_user_errors
       on: azure_monitor.service_bus.errors
    class: Errors
     type: Messaging
component: Azure Service Bus
   lookup: average -5m unaligned of user
    units: errors/s
    every: 1m
     warn: $this > (($status >= $WARNING) ? (10) : (25))
    delay: down 5m multiplier 1.5 max 1h
  summary: Service Bus user errors on ${label:resource_name}
     info: Rate of user (client-side) errors on Service Bus namespace ${label:resource_name} \
           in ${label:resource_group} (${label:region})
       to: sysadmin

# --- Utilization (Premium tier only) ---

# CPU utilization of Premium namespace messaging units.
# AMBA: NamespaceCpuUsage > 70 is Sev2

 template: am_service_bus_namespace_cpu
       on: azure_monitor.service_bus.namespace_resources
    class: Utilization
     type: Messaging
component: Azure Service Bus
   lookup: average -5m unaligned of cpu
    units: percentage
    every: 1m
     warn: $this != nan AND $this > (($status >= $WARNING)  ? (70) : (80))
     crit: $this != nan AND $this > (($status == $CRITICAL) ? (80) : (95))
    delay: down 5m multiplier 1.5 max 1h
  summary: Service Bus namespace CPU on ${label:resource_name}
     info: CPU utilization of Premium Service Bus namespace ${label:resource_name} \
           in ${label:resource_group} (${label:region}). \
           Only available on Premium tier namespaces
       to: sysadmin

# Memory utilization of Premium namespace messaging units.
# AMBA: NamespaceMemoryUsage > 70 is Sev2

 template: am_service_bus_namespace_memory
       on: azure_monitor.service_bus.namespace_resources
    class: Utilization
     type: Messaging
component: Azure Service Bus
   lookup: average -5m unaligned of memory
    units: percentage
    every: 1m
     warn: $this != nan AND $this > (($status >= $WARNING)  ? (70) : (80))
     crit: $this != nan AND $this > (($status == $CRITICAL) ? (80) : (95))
    delay: down 5m multiplier 1.5 max 1h
  summary: Service Bus namespace memory on ${label:resource_name}
     info: Memory utilization of Premium Service Bus namespace ${label:resource_name} \
           in ${label:resource_group} (${label:region}). \
           Only available on Premium tier namespaces
       to: sysadmin

# --- Latency ---

# Server send latency measures how long Service Bus takes to complete
# send operations. High latency indicates performance degradation.

 template: am_service_bus_send_latency
       on: azure_monitor.service_bus.send_latency
    class: Latency
     type: Messaging
component: Azure Service Bus
   lookup: average -5m unaligned of average
    units: milliseconds
    every: 1m
     warn: $this > (($status >= $WARNING)  ? (500)  : (1000))
     crit: $this > (($status == $CRITICAL) ? (1000) : (3000))
    delay: down 5m multiplier 1.5 max 1h
  summary: Service Bus send latency on ${label:resource_name}
     info: Average server send latency on Service Bus namespace ${label:resource_name} \
           in ${label:resource_group} (${label:region})
       to: sysadmin

# --- Saturation ---

# Dead-lettered messages accumulate when messages cannot be processed
# after max delivery attempts or when they expire. Growing dead letter
# queues indicate consumer failures or poison messages.
# AMBA: DeadletteredMessages > 0 is Sev2

 template: am_service_bus_dead_lettered_messages
       on: azure_monitor.service_bus.problem_messages
    class: Errors
     type: Messaging
component: Azure Service Bus
   lookup: average -5m unaligned of dead_lettered
    units: messages
    every: 1m
     warn: $this > (($status >= $WARNING)  ? (0)  : (1))
     crit: $this > (($status == $CRITICAL) ? (10) : (100))
    delay: down 5m multiplier 1.5 max 1h
  summary: Service Bus dead-lettered messages on ${label:resource_name}
     info: Dead-lettered messages in Service Bus namespace ${label:resource_name} \
           in ${label:resource_group} (${label:region}). \
           Messages land in dead letter queue after exceeding max delivery attempts or expiring
       to: sysadmin

# Active message queue depth. Sustained growth means consumers
# are not keeping up with producers.
# AMBA: ActiveMessages > 100 is Sev2

 template: am_service_bus_active_messages
       on: azure_monitor.service_bus.queue_depth
    class: Workload
     type: Messaging
component: Azure Service Bus
   lookup: average -10m unaligned of active
    units: messages
    every: 1m
     warn: $this > (($status >= $WARNING)  ? (5000)  : (10000))
     crit: $this > (($status == $CRITICAL) ? (10000) : (50000))
    delay: down 5m multiplier 1.5 max 1h
  summary: Service Bus queue depth on ${label:resource_name}
     info: Active messages queued in Service Bus namespace ${label:resource_name} \
           in ${label:resource_group} (${label:region}). \
           Sustained growth means consumers are not keeping up with producers
       to: sysadmin

# --- Request Success ---

# Helper: total incoming requests over 5 minutes (no alarm, just a value)

 template: am_service_bus_incoming_requests
       on: azure_monitor.service_bus.requests
    class: Workload
     type: Messaging
component: Azure Service Bus
   lookup: sum -5m unaligned of incoming
    units: requests/s
    every: 1m
     info: Total incoming requests on Service Bus namespace ${label:resource_name} \
           in ${label:resource_group} (${label:region})

# Request success rate. Only fires when there is meaningful traffic
# to avoid false positives during idle periods.

 template: am_service_bus_request_success_rate
       on: azure_monitor.service_bus.requests
    class: Errors
     type: Messaging
component: Azure Service Bus
   lookup: sum -5m unaligned of successful
     calc: ($am_service_bus_incoming_requests > 0) ? ($this * 100 / $am_service_bus_incoming_requests) : (100)
    units: %
    every: 1m
     warn: ($am_service_bus_incoming_requests > 120) ? ($this < (($status >= $WARNING)  ? (99) : (95))) : (0)
     crit: ($am_service_bus_incoming_requests > 120) ? ($this < (($status == $CRITICAL) ? (95) : (80))) : (0)
    delay: down 5m multiplier 1.5 max 1h
  summary: Service Bus request success rate on ${label:resource_name}
     info: Percentage of successful requests on Service Bus namespace ${label:resource_name} \
           in ${label:resource_group} (${label:region})
       to: sysadmin

# --- Message Operations ---

# Abandoned messages indicate consumers receiving but not processing messages.
# Sustained abandonment suggests poison messages or consumer failures.

 template: am_service_bus_abandoned_messages
       on: azure_monitor.service_bus.message_operations
    class: Errors
     type: Messaging
component: Azure Service Bus
   lookup: average -5m unaligned of abandoned
    units: messages/s
    every: 1m
     warn: $this > (($status >= $WARNING) ? (1) : (5))
    delay: down 5m multiplier 1.5 max 1h
  summary: Service Bus abandoned messages on ${label:resource_name}
     info: Rate of abandoned messages on Service Bus namespace ${label:resource_name} \
           in ${label:resource_group} (${label:region}). \
           Consumers are receiving but failing to process messages
       to: sysadmin

# --- Replication (Geo-DR only) ---

# Replication lag count measures how many messages are pending replication
# to the secondary namespace. Only relevant with Geo-DR configured.

 template: am_service_bus_replication_lag
       on: azure_monitor.service_bus.replication_lag
    class: Latency
     type: Messaging
component: Azure Service Bus
   lookup: average -5m unaligned of messages
    units: messages
    every: 1m
     warn: $this != nan AND $this > (($status >= $WARNING)  ? (100)  : (1000))
     crit: $this != nan AND $this > (($status == $CRITICAL) ? (1000) : (10000))
    delay: down 5m multiplier 1.5 max 1h
  summary: Service Bus replication lag on ${label:resource_name}
     info: Messages pending replication to secondary namespace on Service Bus ${label:resource_name} \
           in ${label:resource_group} (${label:region}). \
           Only relevant when Geo-DR is configured
       to: sysadmin

# Replication lag duration measures time behind the primary.

 template: am_service_bus_replication_lag_duration
       on: azure_monitor.service_bus.replication_lag_duration
    class: Latency
     type: Messaging
component: Azure Service Bus
   lookup: average -5m unaligned of duration
    units: seconds
    every: 1m
     warn: $this != nan AND $this > (($status >= $WARNING)  ? (30) : (60))
     crit: $this != nan AND $this > (($status == $CRITICAL) ? (60) : (120))
    delay: down 5m multiplier 1.5 max 1h
  summary: Service Bus replication lag duration on ${label:resource_name}
     info: Replication lag duration to secondary namespace on Service Bus ${label:resource_name} \
           in ${label:resource_group} (${label:region}). \
           Only relevant when Geo-DR is configured
       to: sysadmin