What would you recommend that we return from the "asadmin get" command,
which is currently returning the time in milliseconds. "0"? "-1"?
Not return the attribute at all?
Hans Hrasna wrote on 10/16/09 13:40:
> If no sample was ever taken the value should be null, since any
> numerical value would represent an actual time and falsely insinuate
> that a measurement was taken at that time.
>
> Bill Shannon wrote:
>> And what should its value be if no sample was ever taken?
>>
>>
>> Hans Hrasna wrote on 10/16/09 12:48:
>>> Hi,
>>>
>>> lastSampleTime is the time that the measurement was taken. For
>>> example, if measurements are being taken every hour and being stored
>>> in a variable or table, then the lastSampleTime should reflect the
>>> time the data was sampled and stored. Or for example if measurements
>>> are being taken every time an event occurs, like a connection, then
>>> the lastSampleTime is the time of the last event.
>>>
>>> -Hans
>>>
>>>
>>> Sankar Neelakandan wrote:
>>>> May be Hans can tell us what he meant by this in JSR 77 spec.
>>>>
>>>> JSR77.6.4.1.5
>>>> getLastSampleTime long getLastSampleTime()
>>>> Returns the time the most recent measurment was taken represented as
>>>> a long,whose value is the number of milliseconds since January 1,
>>>> 1970, 00:00:00.
>>>>
>>>>
>>>>> And if, after that 10th hit, I do "get" 10 times, every one of them
>>>>> will
>>>>> return exactly the same data, until the next hit of the web page?
>>>> > hmm..actually I just tried this and the lastSampleTime gets
>>>> updated every time you do "get" with the current time. That doesn't
>>>> seem right to me.
>>>>
>>>> What you are seeing is correct because stats is a continuously
>>>> changing entity and something else (being get command/GUI etc.,) who
>>>> is interested in the entity can take a measurement at a desired time
>>>> specified by lastSampleTime.
>>>>
>>>>>>
>>>>>> The confusion is over whether "get" is *causing* the sampling, or is
>>>>>> just returning the sampled data that was caused by something else.
>>>>> the latter - returns the sampled data that was caused by something
>>>>> else
>>>> The example with processingtime shows the process of "how it is
>>>> calculated rather than how it is sampled."
>>>>
>>>>>
>>>
>>
>