Chip, I had done the manual testing for the feature covering the tests cases.
Creation: ========= 1. creating of healthcheck with default api params, with optional params. 2. Validation of the params ( checks of response time values) 3. Verified if the LB has the capability of healthcheck then create else throw the exception and do clean up 4. Verified created monitor in NetScaler device has created the monitor with all the specified params. 5. Verified created monitor is bind to the services in the LB. 6. Addition of new VM to the LB rules, should create new services and if monitor is available then monitor is bind to the services 7. Deletion of VM's from LB rules should unbind the monitor from services and services will be delete from LB 8. create LB rule of TCP protocol and assign instances, create lb healthcheck on lb rule. A TCP monitor is created and it will binded to all the services. 9. create LzB rule of HTTP protocol and assign instances, create lb healthcheck on lb rule. HTTP monitor is created and it will binded to all the services. 10. create LB rule of HTTP protocol and assign instances, create lb healthcheck on lb rule with pingpath and other params, HTTP monitor is created and it will binded to all the services and its properties pingpath value will be present. 11. create healthcheck for LB where already healthcheck is present, verified error is thrown. 12. Create Healthcheck on LB where LB provider don't have the healthcheck capability, Exception is throw informing the LB provider don't have the capability. Deletion: ======== 1. Deletion of healthcheck, will unbind the monitor from services and deletes the monitor 2. Deletion of healthcheck with invalide id, verified error is thrown. 3. Deletion of LB rule will delete the monitor associated to the rule in cloudstack db and in the Netscaler device. 4. Deletion of healthcheck of type TCP protocol, verified healthcheck is deleted. 5. Deletion of healthcheck of type HTTP protocol, verified healthcheck is deleted. 6. Deletion of Healthcheck will reset the state of the services of the LB in the in lb_vm_map to null Listing LB Healthchecks: =================== 1. listing of lbhealth checks with lbrule id a. if healthcheck is present then a response with healthcheck values will be shown b. Empy list will be shown. HealthCheckManager: ================== 1. Verified healthacheckManager is updating the state of the service in lb_vm_map table with the value from NetScaler. 2. Verified healthacheckManager don't update the service state for the LB rule where LB Provider don't have the HealthCheck capability. 3. Modify the config interval ( default 600sec) to 60(1 min) verified the state of the service is updated in the table every 1 min. Thanks Rajesh Battala > -----Original Message----- > From: Chip Childers [mailto:chip.child...@sungard.com] > Sent: Tuesday, March 19, 2013 12:08 AM > To: cloudstack-dev@incubator.apache.org > Cc: Rajesh Battala; Brian Federle; Animesh Chaturvedi; Haroon Abdelrahman > Subject: Re: [Merged] - CloudStack -664:AWS Health Check feature into Master > > On Mon, Mar 18, 2013 at 06:35:33PM +0000, Pranav Saxena wrote: > > Hi Chip , > > > > I guess there was a discussion about the reason behind not having unit tests > for this feature now - (http://mail-archives.apache.org/mod_mbox/incubator- > cloudstack- > dev/201301.mbox/%3C97F4356AEA71904482CD192135C038F9011CC0D20E7C > @BANPMAILBOX01.citrite.net%3E ) . Not sure , if they hold good now as well > otherwise Rajesh would need to add them . > > > > Thanks, > > Pranav > > Indeed we did discuss this. It had been over a month and a half, so I was > assuming that the comment about Rajesh working on some tests would have > meant that they were being worked on... > > Any reason to not at least have marvin tests?
