Test Execution

The following test topologies are available in Hawkeye:

  1. Node to Node testing - allows execution of tests between one node (probe) and another node (probe);
  2. Mesh - will allow execution of tests on a full mesh of probes;
  3. Real Services – will allow real tests (HTTP,FTP,ICMP,etc) test from probes to servers located in internet.
Node to Node Test Configuration

The node to node test execution will allow the following selections:

  • Probe from: probe where the test is executed FROM (can be network side);
  • Probe to: probe where the test is executed TO (can be access side);
  • Test type: the type of test that will be executed;
  • Test duration: the duration of the test time while the active traffic will be generated;
  • Test options: depending on the test types, some options would be available to select from (e.g. bit rate, packet size, number of concurrent pairs, etc…).

The direction of the test will be from Node 1 to Node 2. When test names include downstream, the traffic will be generated from Node 1 to Node 2. When the name includes upstream the traffic will be generated from Node 2 to Node 1.

Tests can only be executed between 2 manual probes or 2 automatic probes, you can not have 1 automatic and 1 manual probe into the same test schedule.

Mesh Test Configuration

By selecting Test Execution Mesh option from the Test Execution main menu, the user will have the option to configure tests for mesh topologies available. To see how to create a mesh, check Probe Management section.

The mesh test execution will allow the following selections:

  • Mesh Name: select the mesh under test previously configured;
  • Test type: the type of test that will be executed;
  • Test duration: the duration of the test time while the active traffic will be generated.

Not all test types available for Node to Node topologies are available for Mesh topologies.

Starting a test on a mesh will add all available paths in the mesh (see mesh creation and configuration section).

Real Services Test Configuration

To configure a real service test scenario, select the Test Execution Real Service option located under the Test execution menu.

The real services test execution will allow the following selections:

  • Probe from: probe where the test is executed FROM
  • Test type: the type of test that will be executed
  • Destination Server: the real server where the test is executed TO
  • Test options: depending on the test types, some options are available to select from or configure to (for example, DNS server, packet size, and ping interval).

Real Services test are available only for hardware probes type.

Test Queue Management

The status of a test schedule can be one of the following values:

  • Running: the tests are under execution. This will be displayed into black color.
  • Queued: waiting for execution – the test resources aren’t available (busy), so the test is waiting in queue.
  • Waiting next execution: the test is scheduled and the schedule is active, but the next execution time is not available yet.
  • Schedule on hold: the test is scheduled but the schedule has been placed on hold (see putting a schedule on hold section).
  • For mesh tests: queuing for completion. This status is displayed for a mesh test when the mesh is partially completed and the test is split into different test runs.

The following priority in the queue applies:

  • The later next test execution takes priority;
  • One shot tests are executed in priority to scheduled tests;
  • The mesh schedules that need to be completed are executed in priority. The next execution time for a mesh is not changed until last run completion;

Node to Node tests are executed in priority to Mesh tests.

Test Scheduling Configuration

The test scheduling options are available for all types of test configurations. The functionality is the same all configurations.

Test interval [minutes]: defines the time interval between tests. This interval is calculated on best effort basis, as the tests in the queue might be too numerous to allow all intervals to be respected. In case the interval is too small for tests to be executed, the queue will automatically adjust to best possible interval to execute the test.

If the test Interval is set to 0, then the test will be executed only one time. This will be displayed as one shot test. In this case, the test will be displayed into test execution table while waiting for execution (Queuing) or running. Once the test is completed, it will be removed from the test execution table and the results will be available into the test results lists.

Start schedule : defines the date when the schedule shall start.

  • If left blank, the current date will be applied;
  • If Interval is set to 0 (one shot test), then this will define when to start the test (in batch mode);
  • To select the day and time, change time (24 hours selection), and then select it.

End schedule : defines the date when schedule must end.

  • If left blank, the schedule will last forever;
  • If Interval is set to 0 (one shot test), then this end schedule will not be considered;

To select the day and time, change time (24 hours selection), and then select it.

The test execution list will display only the schedule currently active or active in the future. All schedules not active any more will not be displayed in the list.

Test execution frequency of less than a minute is supported and needs to be enabled in preferences /main section by administrator.

Cancel a Schedule Test

Tests can be canceled by selecting the delete button. This will cancel the test or schedule.

If the test is waiting for execution, in the queue or on hold, the test will be canceled directly.

If the test is canceled while running, first it will finish the current run and then will be canceled.

The scheduled test list is refreshed every minute by default. It might be the case that canceling a test or a schedule that appears queuing of waiting next execution gets the test canceled into a running mode instead. In that case the test execution will refresh and the new status of the test be displayed.

Set a Schedule on Hold

Tests can be put on hold by selecting the pause button. This will put the schedule on hold so wait for next execution. The button will transform in play button, and by selecting this button the test schedule will resume.

If the test is waiting for execution, in the queue or on hold, the test will directly change status.

If the test is running, after the current run ends, the test will be put on hold. A specific status (running then hold) will be displayed.

Interval Optimizer and Enforce Schedule

Interval optimizer takes the start time of a test run and adds the interval for the next run. If the queue delays the start time for the first test or if there are some test run interruptions, the interval optimizer considers them and optimizes the queue for being as close as possible for the run interval. A downturn of this behavior is that the exact run test times are not guaranteed.

Turning on Enforce Schedule forces the test to run at the first run time + test interval. It means that when you run the test, the next test run time is set to first time + n × interval time, where n is set to be the next instance for the schedule.

Example: First time is 12:00 and the interval is 1 hour. If you run the test at 16:12, the next time is set to 12:00 + 5 × (1 hour). Here n=5, that is, the fifth instance of the schedule, starting with 12:00. Therefore, the test is enforced to run at 17:00. This setting allows enforcing the test run schedule at exact times. But if the queue is full, the system delays the test run and does not optimize the queue for best use.