The HA team is cooperating with the Yardstick team on developing multiple HA test cases. These test cases are continious enriched and developed. Below are detaied info about the test cases.
Detailed Test Cases Discription
Yardstick Test Case Description TC019?Control Node Openstack Service High Availability
Yardstick Test Case Description TC025?OpenStack Controller Node abnormally shutdown High Availability
Yardstick Test Case Description TC045?Control Node Openstack Service High Availability - Neutron Server
Yardstick Test Case Description TC046?Control Node Openstack Service High Availability - Keystone
Yardstick Test Case Description TC047?Control Node Openstack Service High Availability - Glance Api
Yardstick Test Case Description TC048?Control Node Openstack Service High Availability - Cinder Api
Yardstick Test Case Description TC049?Control Node Openstack Service High Availability - Swift Proxy
Yardstick Test Case Description TC050?OpenStack Controller Node Network High Availability
Yardstick Test Case Description TC051?OpenStack Controller Node CPU Overload High Availability
Yardstick Test Case Description TC052?OpenStack Controller Node Disk I/O Block High Availability
Yardstick Test Case Description TC053?OpenStack Controller Load Balance Service High Availability
Yardstick Test Case Description TC054?OpenStack Virtual IP High Availability
General test case information for Dovetail
- Provide a high level description of the test area, main features being tested, and answer why it is in scope for Dovetail Danube. If there is a subset that is in scope, describe the subset.
The High availability test cases focus on the availability testing for OPNFV platform. Currently most of test cases are focusing on Openstack HA, including multiple services in OpenStack, such as nova, neutron and etc.. There are also test cases on I/O, and CPU. These testcases are released mostly in Colorado. And the HA testing framework is reshaped in Danube. We see these test cases as basic HA testing for OPNFV, making sure the platform meet the Carrier-Grade availability requirement. The test cases will extend in future releases and more HA aspects will be taken into consideration.
- Describe how the features are tested, and justify why it is appropriate methodology for compliance testing. If modification or enhancement are needed, please describe the work needed in Dovetail.
The test cases mainly include an attacker and a monitor. The attacker is responsible for creating the failure. Monitor is responsible for monitoring the certain service under test and see if it could recover within SLA. After testing, there is a certain recovery process that help the whole system recover to the initial state.
- Describe how the automated test cases are implemented, and why it is appropriate for compliance testing. If modification or enhancement are needed, please describe the work needed in Dovetail
Currently the HA testcases are included in Yardstick.
- Describe test results, pass/fail criteria
The test result is pass if the monitor finds out the service under testing recovers within SLA.
- Describe the conditions that the SUT must be in for the test cases to run. Are the test cases executable in a limited subset of OPNFV Scenarios only? If so, which subset have the pre-conditions met to pass?
The SUT should follow the HA scenarios in the OPNFV releases.
- Provide a link of pass test run results / examples, if any.
- Any other important information or comments