Service Bus for Windows Server Performance Counters

 

Applies To: Service Bus for Windows Server 1.1

Service Bus for Windows Server implements a set of performance counters that enable you to monitor the state of the server. The Azure Service Bus Server Management Pack evaluates performance counters to determine the health of the server.

The Service Bus performance counters are organized into three groups that reflect the architecture of the product.

To access the Service Bus performance counters, open the Performance Monitor. In the Perfomance pane, expand Monitoring Tools, and then click Performance Monitor. In the main pane click the green plus (+) icon. In the Add Counters window, scroll to the Service Bus performance counter groups.

To open the Performance Monitor on Windows Server 2008 R2, click Start, then Administrative Tools, then Performance Monitor. To open the Performance Monitor application on Windows Server 2012, type Performance Monitor in the Start screen, then click the Performance Monitor icon.

Service Bus for Windows Server Message Container

The following performance counters are scoped to a message container. For each message container, the Service Bus for Windows Server maintains one copy of each counter. The current value of these counters is available on the node on which the container is loaded.

Name	Description
Incoming Messages/sec	Number of incoming messages per second.
Outgoing Messages/sec	Number of outgoing messages per second.
Bytes Per Incoming Message	Average size of incoming message in bytes.
Bytes Per Outgoing Message	Average size of outgoing message in bytes.
Average Incoming Message Batch Size	Average batch size for incoming messages.
Average Outgoing Message Batch Size	Average batch size for outgoing messages.
Send Message Latency	Average time to send a message to the Service Bus.
Receive Message Latency	Average time to receive a message from the Service Bus.
Complete Message Latency	Average time to complete a message.
Active Soap Connections	Number of active SOAP connections.
Sql Operations/sec	Number of SQL operations per second.
Sql Operation Latency	Average processing time of a SQL operation. The system should be considered unhealthy if this counter value exceeds 500 milliseconds.
Sql Exceptions/sec	Number of SQL exceptions received per second. The system should be considered unhealthy if this counter value exceeds 100.
Operation Timeouts/sec	Number of messaging operations that timed out per second. The system should be considered unhealthy if this counter value exceeds 10.
Storage Throttle Level	Current storage throttle level. This throttle limits the number of SQL operations. Level 0: unthrottled, Level 5: fully throttled. The system should be considered unhealthy if this counter value exceeds 2.
Storage Throttle Limit	Maximum number of messaging operations permitted in the current 10-second interval.
Loaded Queue Count	Number of queues loaded.
Loaded Topic Count	Number of topics loaded.
Loaded Subscription Count	Number of subscriptions loaded.
Entities Loaded/sec	Rate at which entities are loaded.
Entities Unloaded/sec	Rate at which entities are unloaded.
Entities Aborted/sec	Rate at which entities are aborted. The system should be considered unhealthy if this counter value exceeds 10.
Available Sessions Count	Number of sessions with messages that are available for consumption.
Scaled Receive Sessions Requested/sec	Number of incoming scaled receive session requests per second.
Scaled Receive Sessions Handed Out/sec	Number of scaled receive sessions that are handed out per second.
PrefetchReceivePerSecond	Rate of prefetch operations.
FailedPrefetchPerSecond	Rate of failed prefetch operations.
Prefetch Receive/sec	Number of prefetch operations per second.
Failed Prefetch/sec	Number of failed prefetch operations per second.
Broker Current Transaction Count	Number of in-flight user transactions.
Broker Transactions Committed/sec	Rate at which user transactions are committed.
Broker Transactions Aborted/sec	Rate at which user transactions are aborted.
Broker Transaction Latency	Average time it takes to process a user transaction.
Batch Scheduler Backlog Queue Size	The size of the batch scheduler’s connection pool backlog queue.
Batch Scheduler In-Flight Operation Count	The current serving connection pool size of the batch scheduler.
Batch Scheduler In-Flight Operation Batch Count	The number of operations within a batch.
Request Scheduler Queue Size	Number of work items enqueued in the request scheduler queue.
Truncation Job Backlog Size	Number of SQL truncation jobs that are queued for execution.
Pump Messages/sec	Rate of messages that are pumped from topics to subscriptions.
Failed Pump Messages/sec	Rate of messages that are pumped from topics to subscriptions.
Failed Publish Requests/sec	Rate of failed requests to send a message or set a session state.
Failed Consumer Requests/sec	Rate of failed requests to receive or delete a message or get a session state.
Management Requests/sec	Rate of management requests.
Send Batches/sec	Number of incoming batches of send operations per second.
Receive Batches/sec	Number of incoming batches of receive operations per second.
Complete Batches/sec	Number of incoming batches of complete operations per second.
Cache Total Requests	Total number of message requests for the cache log store.
Cache Hits	Number of messages retrieved from the cache log store.
Cache Message Count	Current number of messages in the cache log store.
Cache Incoming Messages/sec	Number of messages added to the cache log store per second.
Cache Outgoing Messages/sec	Number of messages removed from the cache log store per second.
Cache Prefetch Rate	Number of messages pre-fetched per second.

Service Bus for Windows Server Messaging Broker Node

The following performance counters are scoped to the Service Bus for Windows Server broker process. Each node in the farm maintains one copy of these counters.

Name	Description
Cpu Throttle Level	Current CPU throttle level. This throttle limits the number of messaging operations. Level 0: unthrottled, Level 5: fully throttled. The system should be considered unhealthy if this counter value exceeds 1.
Cpu Throttle Limit	Maximum number of messaging operations permitted in the current 10-second interval.
Memory Throttle Level	Current memory throttle limit. This throttle limits the size of the messaging cache. Level 0: max cache size, Level 5: Minimum cache size. The system should be considered unhealthy if this counter value exceeds 2.

Service Bus for Windows Server Messaging Gateway Node

The following performance counters are scoped to the Service Bus for Windows Server gateway process. Each node in the farm maintains one copy of these counters.

Name	Description
Connections Established	Current number of client connections.
Connections Established/sec	Rate at which client connections are created.
Connections Disconnected/sec	Rate at which client connections are closed.
Pending Operations	Number of in-flight brokered messaging operations requested by clients.
Pending Operations/sec	Number of brokered messaging operations requested by clients per second.
Succeeded Operations/sec	The number of brokered messaging operations that executed successfully per second.
Failed Operations/sec	Total number of failed brokered messaging operations per second.
Non-waiting Request Latency	The number of milliseconds to execute the most recent non-waiting brokered messaging request.
Sql Operation Latency	Average processing time of a SQL operation. The system should be considered unhealthy if this counter value exceeds 100 milliseconds.
Sql Operations/sec	Number of SQL operations per second.
Sql Exceptions/sec	Number of SQL exceptions received per second. The system should be considered unhealthy if this counter value exceeds 100.
Total Cache Misses	Total number of times an endpoint is not found in the cache.