[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: [rhel6-branch] Do not traceback on mpath errors caused by faulty hardware.



>  exception.py            |    8 ++++++++
>  installinterfacebase.py |   11 +++++++++++
>  storage/devices.py      |    2 +-
>  storage/errors.py       |    5 ++++-
>  4 files changed, 24 insertions(+), 2 deletions(-)

I like this approach much better.  In general I do like the idea of
trying to track down hardware failures and report them instead of
generating tracebacks that people will file as bugs.  I'm just
pessimistic as to how successful we are going to be.

> diff --git a/installinterfacebase.py b/installinterfacebase.py
> index 71cb9cb..6d4dd11 100644
> --- a/installinterfacebase.py
> +++ b/installinterfacebase.py
> @@ -97,3 +97,14 @@ class InstallInterfaceBase(object):
>                                               custom_icon=icon,
>                                               custom_buttons=buttons,
>                                               expanded=True)
> +
> +    def hardwareError(self, exception):
> +        text=_("The installation was stopped due to what seems to be a problem "
> +               "with your hardware, the exact error message is:\n\n%s.\n\n "
> +               "The installer will now terminate.") % str(exception)

Little grammar nitpick here.  The first sentence should be broken up
into two, as follows:

   "The installation was stopped due to what seems to be a problem with
    your hardware.  The exact error message is:"

> diff --git a/storage/errors.py b/storage/errors.py
> index 1d1fea0..cec075c 100644
> --- a/storage/errors.py
> +++ b/storage/errors.py
> @@ -21,7 +21,10 @@
>  #
>  
>  class StorageError(Exception):
> -    pass
> +    def __init__(self, *args, **kwargs):
> +        hardware_fault = kwargs.pop("hardware_fault", False)
> +        super(StorageError, self).__init__(*args, **kwargs)
> +        self.hardware_fault = hardware_fault
>  
>  # Device
>  class DeviceError(StorageError):

Another little nitpick.  Is there a problem with keeping hardware_fault
in the kwargs when you call the superclass's __init__?  Alternately, is
there a problem with setting self.hardware_fault before you call the
same?

- Chris


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]