Re: [Duplicity-talk] Azure in cygwin

[Marcin Okraszewski]

Hi,
I have updated the Azure backend to the latest version Storage SDK. It
is attached. Following the convention I made it backward compatible,
but unfortunately I wasn't able to test it with the old version.

I have also extended the functionality to set the uploaded blob
storage tier (hot, cool or archive). You can set it by environment
variables, eg. AZURE_DATA_TIER='Archive' and
AZURE_MANIFEST_TIER='Cool' . Leaving those unset works as previously,
without setting any tier. Unfortunately I had to introduce two tier
settings as backup process for some reason reads latest manifest file
from the backend even if local cache is in sync, and is not possible
to download archived blobs in that process (it works similar to AWS
Glacier). By the way I wonder why it downloads the latest manifest
when it has it in local cache.

Do you want to include it into duplicity maybe? Should I create some
pull request?

Thanks,
Marcin


On Fri, Dec 28, 2018 at 12:14 AM Marcin Okraszewski <address@hidden> wrote:
>
> Hi,
> I wanted to use Azure with duplicity on cygwin. Unfortunately I keep
> getting following error:
>
> BackendException: Azure backend requires Microsoft Azure Storage SDK
> for Python (https://pypi.python.org/pypi/azure-storage/).
> Exception: cannot import name BlobService
>
> I tried with duplicity 0.7.11 (that comes from cygwin) and manual
> compilation of 0.7.18.2. I have installed Python azure-storage-blob
> (1.4) as https://github.com/Azure/azure-storage-python suggest as well
> as tried with legacy azure-storage (0.36).
>
> Can I make it work somehow? Should I use some specific old module version?
>
> Thanks you,
> Marcin

azurebackend.py
Description: Text Data