eG Monitoring
 

Measures reported by GFireServerTest

Periodically, administrators should measure the workload on the GemFire Cluster server and evaluate the server's ability to handle the load, so that they can figure out whether the server's size commensurate to its load or not. Administrators can perform this load analysis using the GFireServerTest test.

This test tracks the load on the GemFire Cluster server and reports the number threads currently catering to the user requests. In addition, this test reports the time taken for each thread to cater to the user requests as well as the connections that failed. Using this test administrators can figure out the count of threads that are generating the load on the GemFire Cluster server. Capacity planning and clustering decisions can be taken based on insights provided by this test.

Outputs of the test : One set of results for each server/application that is connected to the target GemFire server being monitored.

The measures made by this test are as follows:

Measurement Description Measurement Unit Interpretation
Maximum_connections Indicates the maximum number of connections allowed to the GemFire server. Number  
Maximum_threads Indicates the maximum number of threads allowed to service user requests. Number  
Ping_time Indicates the maximum time taken by the server to cater to the user requests. MilliSeconds A low value is desired for this measure.
Active_execution Indicates the number of executions i.e., CQs (Continuous Queries) that were active on the GemFire server. Number  
Notify_average_latency Indicates the average time taken by the server to process user requests. Milliseconds A low value is desired for this measure.
Client_connection_thread Indicates the number of threads that are currently catering to the user requests. Number  
Client_VM_connected Indicates the number of virtual machines that are connected to the GemFire server in a cluster setup. Number  
Get_average_latency Indicates the average time taken to retrieve data from the backend servers by the GemFire server to cater to the user requests. Milliseconds  
Get_request_rate Indicates the rate at which data was retrieved from the backend server by the GemFire server upon user requests. Requests/sec  
Put_average_latency Indicates the average time taken by the GemFire server to send the data to the backend servers. Milliseconds  
Put_request_rate Indicates the rate at which data is sent to the backend servers from the GemFire server. Requests/sec  
Waiting_thread_connect Indicates the number of connections that are currently waiting for a thread to start processing the user request. Number A low value is desired for this measure.
Failed_connections Indicates the number of connections that failed to connect to the GemFire server. Number Ideally, the value of this measure should be zero.