Parameters Effecting Performance

Hawkeye Timeout for Receiving Statistics

In the Test Engine section of Preferences the parameter Statistics receive timeout (seconds) defines the max amount of time the Hawkeye server will wait for an endpoint to send the test results once a test has completed.

Test Timeouts

In the Test Controller section of Preferences there are many parameters defining max test durations. After a timer expires Hawkeye server ends the test and cleans up.

In the Advanced Options section, the parameter Maximum time processing Node to Node and Mesh tests specifies how many seconds after test duration is over before the Hawkeye server ends the test and cleans up.

There are two parameters for Real Services. In the Test Controller section the parameter Real Test Max duration defines how long the endpoint processes the test, after which time the endpoint terminates the test. In the Advanced Options section, the parameter Maximum time processing Real Service tests specifies how many seconds after test duration is over before the Hawkeye server ends the test and cleans up.

Auto Refresh of Web Pages

The user needs to be aware that the autorefresh of web pages impacts CPU usage and must not be too frequent. There are three autorefresh timers, one for the main dashboard, one for the map display, and one for other web pages such as floorplans and test execution pages.

There are two parameters in the Main section of preferences.

The parameter Dashboard autorefresh enables or disables the automatic refresh of the main dashboard statistics. The second parameter is Dashboard autorefresh interval, which defines how often the statistics for the main dashboard are re-calculated and displayed. We recommend you to keep this above 30 seconds.

The parameter Refresh Execution List Timer specifies how many seconds between updating the test execution web pages. It is recommended to keep this value above 10 seconds for system performance.

The parameter Map update frequency in seconds impacts how frequently the maps are re-calculated. It is recommended to keep this above 30 seconds. The parameter Refresh list timer value impacts other web pages that support autorefresh. We recommend you to keep this above 10 seconds.

Max Concurrent Tests in GUI

There is a parameter Max Concurrent Tests in GUI in the Main section of Preferences that defines how many tests are allowed to be running concurrently. This parameter, when set correctly, will ensure no issues with tests failing to run because the number of licensed tests has been reached. This parameter effectively manages the maximum number of simultaneous tests and will queue up those tests that need to run. This also indirectly allows the user to define how much CPU processing is allocated to processing test results.

Map Refresh

The frequency for updating maps can be CPU intensive. There are two parameters under Preferences in Administration that defines this. First the parameter Map Report must be enabled or disabled. If disabled no processing of maps to be displayed will take place. In the GUI section of preferences the parameter Map update Frequency in seconds defines how frequent the processing or refreshing of maps will be performed. As this impacts the CPU we advise not setting this value to less than 30 seconds.

Probe Healthcheck Polling Interval

In the Advanced Options section, the parameter Test Controller: Polling interval defines how often the test controller process will poll each endpoint to check its status. We recommend you to keep this above 4 seconds. This impacts Probe healthcheck feature and reporting alarms for endpoint down (email/SNMP traps).

Log Retention Minutes

In the Logs section, there are two parameters used to define the test log retention time on the Hawkeye server. The Log Retention minutes specifies how many minutes to retain Hawkeye test logs in the directory /opt/hawkeye/logs/Hawkeye. The parameter Test Engine Log retention specifies how many minutes to retain the endpoint test logs in /opt/hawkeye/logs/Chariot.

Automatic Endpoint Retry of Tests

Hawkeye has a feature to automatically retry a test that fails to execute. This is useful in unstable networks. However if a Customer has limited licenses and has carefully scheduled a number of tests to maximize the available licenses enabling this feature may result in increased licenses being used at once.

Under Administration in Preferences for the Test Engine there is a parameter allow_pair_reinit_run which is used to enable disable the automatic retry of failed tests. If enabled and a test fails to start due to connection issues, the test will retry in pair_reinit_retry_interval seconds. The test will be tried a further pair_reinit_max_run number of times. The retry interval after the first attempt is defined by pair_reinit_retry_interval_run seconds.