[Libvirt-cim] [PATCH] [TEST] Update 32_start_reboot.py to use cim_() functions

[Deepti B Kalakeri]

Kaitlin Rupert wrote:
>>> @@ -606,7 +610,7 @@
>>> def cim_reboot(self, server, req_time=const.TIME, poll_time=30): 
>>> return self.cim_state_change(server, const.CIM_REBOOT, - req_time, 
>>> poll_time) + req_time, poll_time, const.CIM_ENABLE) 
>> The test case gives a false positive. The following are the log 
>> messages which means the tc is suppose to fail.
>>
>> Tue, 06 Jan 2009 01:10:14:TEST LOG:INFO - ====32_start_reboot.py Log====
>> Tue, 06 Jan 2009 01:10:17:TEST LOG:ERROR - Unable to check guest state
>> Tue, 06 Jan 2009 01:10:17:TEST LOG:ERROR - Exception: Wrong guest 
>> instance
>> Tue, 06 Jan 2009 01:10:18:TEST LOG:ERROR - Unable to check guest state
>> Tue, 06 Jan 2009 01:10:18:TEST LOG:ERROR - Exception: EnabledState is 
>> 9, expected 2.
>>
>> But, the test case gives a PASS message on the console.
>>
>
> When you reboot a guest, the will be paused briefly before being 
> restarted. When cim_state_change(), it calls check_guest_state() in a 
> loop to verify the guest reaches the appropriate state.
>
> check_guest_state() prints an error message if the state is not in the 
> expected state.
>
> I could change these from error messages to info messages. But 
> check_guest_state() would still need to return an error in this case 
> (even though the test doesn't report an error because the guest really 
> is rebooted in this case).
I get your point here, but still it left me confused as to why I am 
getting the Error Messages.
It would be good to change the Error messages to Info messages whenever 
appropriate.
We can use the timeout period to decide whether it is time to print an 
error message in case the guest info is not found or the states are not 
yet properly.
Any thoughts ??

Thanks and Regards,
Deepti.