Merge lp://qastaging/~doanac/lava-dispatcher/config-warnings into lp://qastaging/lava-dispatcher

Andy Doan <email address hidden> writes:

> Andy Doan has proposed merging lp:~doanac/lava-dispatcher/config-warnings into lp:lava-dispatcher.
>
> Requested reviews:
> Linaro Validation Team (linaro-validation)
>
> For more details, see:
> https://code.launchpad.net/~doanac/lava-dispatcher/config-warnings/+merge/137431
>
> Not a huge fan of this, but it quiets the warnings printed out when a
> device like fastmodel exists that uses our boot options support.

I'm not really a fan of considering extra entries in the config as
errors for a couple of reasons:

1) I do this in my panda config to help readability:

device = /dev/serial/by-path/pci-0000:00:1a.0-usb-0:1.1.3:1.0-port0
connection_command = sg dialout "cu -l %(device)s -s 115200"

2) We're going to want to add things like "which alsa device do we
record from when capturing audio from this device" to the device config
and I don't think it's reasonable to add all these values to the config
as well.

> This fix becomes more important very soon, because the scheduler will
> be asking the dispatcher for its known devices periodically which
> could cause our log files to be filled with warning messages.

That said, +1 on the proposal unless I've convinced you to change
approach :-)

Cheers,
mwh

On 12/02/2012 02:01 PM, Michael Hudson-Doyle wrote:
> Andy Doan <email address hidden> writes:
>
>> Andy Doan has proposed merging lp:~doanac/lava-dispatcher/config-warnings into lp:lava-dispatcher.
>>
>> Requested reviews:
>> Linaro Validation Team (linaro-validation)
>>
>> For more details, see:
>> https://code.launchpad.net/~doanac/lava-dispatcher/config-warnings/+merge/137431
>>
>> Not a huge fan of this, but it quiets the warnings printed out when a
>> device like fastmodel exists that uses our boot options support.
>
> I'm not really a fan of considering extra entries in the config as
> errors for a couple of reasons:
>
> 1) I do this in my panda config to help readability:
>
> device = /dev/serial/by-path/pci-0000:00:1a.0-usb-0:1.1.3:1.0-port0
> connection_command = sg dialout "cu -l %(device)s -s 115200"
>
> 2) We're going to want to add things like "which alsa device do we
> record from when capturing audio from this device" to the device config
> and I don't think it's reasonable to add all these values to the config
> as well.
>
>> This fix becomes more important very soon, because the scheduler will
>> be asking the dispatcher for its known devices periodically which
>> could cause our log files to be filled with warning messages.
>
> That said, +1 on the proposal unless I've convinced you to change
> approach :-)

I think I'll keep my change because it does silence something. However,
we might consider changing the log.warning of the configglue report to
be a log.debug.