Azure Managed Cache Service Performance Counters

Important

Microsoft recommends all new developments use Azure Redis Cache. For current documentation and guidance on choosing an Azure Cache offering, see Which Azure Cache offering is right for me?

This topic provides a description of the performance counters available for monitoring Azure Managed Cache Service applications.

Cache Performance Counters

Microsoft Azure Cache provides performance monitor counters that allow you to monitor and troubleshoot Managed Cache Service applications. There are separate performance counters for cache clients and the Managed Cache Service.

• Cache Service Performance Counters

  • Configure Cache Service Performance Counters

  • Cache Service Performance Counters Reference

• Cache Client Performance Counters

  • Azure Caching:Client

  • Azure Caching:Client Host

Cache Service Performance Counters

Managed Cache Service provides performance counters to help you monitor the health and performance of your cache, and to help with capacity planning for cache scaling. These performance counters can be viewed and configured in the Management Portal on the Monitor tab for Cache.

• Configure Cache Service Performance Counters

• Cache Service Performance Counters Reference

Configure Cache Service Performance Counters

To view and configure Managed Cache Service performance counters, log into the Management Portal and navigate to the Monitor tab for Cache.

You can choose up to twelve metrics for the metrics table, and plot any six of those metrics on the chart by selecting the check boxes by their table headers.

To add or remove performance counters, click Add Metrics.

Cache Service Performance Counters Reference

The following table lists the performance counters for Managed Cache Service.

Note

Performance counters that provide information from the start of the cache service are reset if the cache service is stopped, such as when the cache offering is changed.

The GetAndLock method is considered a read operation. If the object is returned to the client, counters that measure reads and data sent are updated. If the object is already locked, the Cache Miss % count is incremented. In the same way, the PutAndUnlock is a write operation.

Methods that change the state of objects in the cache, such as Unlock and ResetTimeout, are not included in the read or write counts. Instead, these methods are reflected in the client request count.

Performance Counter	Description
Bandwidth Used %	The percentage of bandwidth capacity used. The cache is hosted on a dedicated infrastructure which provides incoming and outgoing bandwidth to the system. The bandwidth utilization reflects how this bandwidth is being used. If your system as a lot of large objects, you may see the bandwidth utilization being high.
Bytes Received /sec	The number of bytes of data received per second by the cache from clients.
Bytes Sent /sec	The number of bytes of data sent per second by the cache to clients.
Cache Miss %	The percentage of unsuccessful cache requests to the total number of requests since the start of the cache service.
Compute Used %	The percentage of the processing used. The cache service utilizes compute resources to service all cache requests. For high throughput applications, the compute resources made available to your dedicated infrastructure can get exhausted. Hence it is important for such applications to monitor the Compute Used % to ensure that it isn’t more than 70-80%, which will ensure that your cache remains responsive and data access continues with low latency.
Data Size in Bytes	The total size of cached data in the cache, not including cache overhead.
Evicted Objects	The number of evicted objects since the start of the cache service.
Expired Objects	The number of expired objects since the start of the cache service.
Memory Used %	The percentage of memory capacity used, including the overhead to memory provisioned for the cache service. The allocated amount of memory for cache is used for storing the user data and the metadata about these objects that the cache needs to store for serving requests. This counter reflects how much of the usable memory of cache is already utilized. Note that you may not see an exact match on the amount of data you put in vs. the utilization, as it will vary depending on your object size and use of features such as tags, regions and notifications.
Object Count	The number of objects stored in cache.
Primary Data Size in Bytes	The size of primary memory data usage of the cache, not including cache overhead, and not including secondary copies of data for High Availability.
Read Requests	The number of read requests (Bulk Get, Get and Enumeration) received from all clients since start of the cache service.
Read Requests /sec	The number of read requests (Bulk Get, Get and Enumeration) received per second from all clients since start of cache service.
Write Requests	The number of write requests (Put, Add and Lock) since start of the cache service.
Write Requests /sec	The number of write requests per second since start of cache service. Put, Add and Lock methods are included with write operations.

Cache Client Performance Counters

There are two categories of cache client performance counters: Azure Caching:Client and Azure Caching:Client Host. Azure Caching:Client has counters that provide information about a single cache client. Azure Caching:Client Host has the same counters as Azure Caching:Client, and provides aggregated information about all of the cache clients on a machine.

• Azure Caching:Client

• Azure Caching:Client Host

Note

To view cache client performance counters in the , select counters from the Azure Caching:Client Host category.

Azure Caching:Client

The Azure Caching:Client performance counter category provides counters that pertain to a single cache client.

Performance Counter	Description
Average Get Latency / operation Microsecond	The average latency in, microseconds for Get requests irrespective of whether they access the local cache or the cache.
Average Get Latency (Network) / operation Microsecond	The average latency in microseconds for Get requests that access the cache. This excludes calls that access the local cache.
Average Put Latency / operation Microsecond	The average latency in microseconds for Put requests.
Local Cache Hits Percentage	The percentage of Get requests across all cache clients (irrespective of their local cache settings) that are fulfilled by the local cache.
Local Cache Filled Percentage	The percentage of objects stored in local cache in relation to the total capacity of the local cache. Note that that is an aggregation of the local cache properties of all cache clients in the current process.
Read Requests	The number of Read requests (Get and Enumeration) since the start of the process.
Write Requests	The number of Write requests (Put and Add) requests since the start of the process.
Bytes Received / sec	The number of bytes of data received from the cache during the previous second.
Bytes Sent / sec	The number of bytes of data sent to the cache during the previous second.
Current Server Connections	The number of active connections to the cache from this process.
Total Connection Requests Failed	The number of failed connection requests since start of the process.
Current Waiting Requests	The current number of requests which have not yet received a response from the cache.
Total Notifications Received	The number of Caching notifications received since start of the process.
Failure Exceptions	The number of Caching exceptions received since the start of the process.
Failure Exceptions / sec	The rate of Caching exceptions received by the process.
Retry Exceptions	The number of retry operation exceptions received since the start of the process.
Retry Exceptions / sec	The rate of retry operation exceptions received by the process.
Timeout Exceptions	The number of timeout exceptions received since the start of the process.
Timeout Exceptions / sec	The rate of timeout exceptions received by the process.
Network Exceptions	The number of network exceptions received since the start of the process.
Network Exceptions / sec	The rate of network exceptions received by the process.
Server Responses Dropped / sec	The total number of times when the process was unable to receive a response from the cache. This can happen when a client thread experiences a timeout and there are no other threads available to receive the response.
Requests	The number of requests sent from process to the cache since start of the process.
Requests / sec	The rate of requests sent from the process to the cache.
Total Local Cache Hits	The total number of requests that are fulfilled by the local cache since start of the process. This does not differentiate between different cache clients with different local cache settings.
Total Local Cache Objects	The total number of objects in local cache. Note that that is an aggregation of the local cache properties of all cache clients in the current process.

Azure Caching:Client Host

The Azure Caching:Client Host performance counter category provides counters that pertain to all of the cache clients on a machine. To view cache client performance counter data in the , use performance counters from the Azure Caching:Client Host category.

Performance Counter	Description
Average Get Latency / operation Microsecond	The average latency in, microseconds for Get requests irrespective of whether they access the local cache or the cache.
Average Get Latency (Network) / operation Microsecond	The average latency in microseconds for Get requests that access the cache. This excludes calls that access the local cache.
Average Put Latency / operation Microsecond	The average latency in microseconds for Put requests.
Local Cache Hits Percentage	The percentage of Get requests across all cache clients (irrespective of their local cache settings) that are fulfilled by the local cache.
Local Cache Filled Percentage	The percentage of objects stored in local cache in relation to the total capacity of the local cache. Note that that is an aggregation of the local cache properties of all cache clients in the current process.
Read Requests	The number of Read requests (Get and Enumeration) since the start of all client processes on the machine.
Write Requests	The number of Write requests (Put and Add) requests since the start of all client processes on the machine.
Bytes Received / sec	The number of bytes of data received from the cache during the previous second.
Bytes Sent / sec	The number of bytes of data sent to the cache during the previous second.
Current Server Connections	The number of active connections to the cache from all client processes on the machine.
Total Connection Requests Failed	The number of failed connection requests since start of all client processes on the machine.
Current Waiting Requests	The current number of requests which have not yet received a response from the cache.
Total Notifications Received	The number of Caching notifications received since start of all client processes on the machine.
Failure Exceptions	The number of Caching exceptions received since the start of all client processes on the machine.
Failure Exceptions / sec	The rate of Caching exceptions received by all client processes on the machine.
Retry Exceptions	The number of retry operation exceptions received since the start of all client processes on the machine.
Retry Exceptions / sec	The rate of retry operation exceptions received by all client processes on the machine.
Timeout Exceptions	The number of timeout exceptions received since the start of all client processes on the machine.
Timeout Exceptions / sec	The rate of timeout exceptions received by all client processes on the machine.
Network Exceptions	The number of network exceptions received since the start of all client processes on the machine.
Network Exceptions / sec	The rate of network exceptions received by the process.
Server Responses Dropped / sec	The total number of times all clients on the machine were unable to receive a response from the cache. This can happen when a client thread experiences a timeout and there are no other threads available to receive the response.
Requests	The number of requests sent from process to the cache since start of all client processes on the machine.
Requests / sec	The rate of requests sent from all client processes on the machine to the cache.
Total Local Cache Hits	The total number of requests that are fulfilled by the local cache since start of all client processes on the machine. This does not differentiate between different cache clients with different local cache settings.
Total Local Cache Objects	The total number of objects in local cache. Note that that is an aggregation of the local cache properties of all cache clients in the current process.

See Also

Concepts

Monitor Azure Managed Cache Service