How to integrate Amazon Glacier with Alfresco as a file system

[abhinav]

Hi All,
I am having a requirement to integrate the alfresco with Amazon Glacier file store. I don't want to use "c:\alf_data" i.e. local file system storage.
Is this possible to integrate the alfresco with Amazon Glacier file store ?

Can anyone suggest ?

Thanks,
Abhinav Mishra

[afaust]

Hello,

there recently was a virtual hackathon where people worked on integrating Alfresco with Amazon Glacier. Please see <a href="https://www.youtube.com/watch?v=_UoMv4m02TE">this episode of Tech Talk Live</a> for a summary of the hackathon including the Glacier project.

Short summary - yes, it is possible, but requires diligent development. I don't know how ready / available the prototype of Nathan McMinn et al is.

Regards
Axel

[abhinav]

Thanks for the reply, i have gone through the video.. It seems Nathan is working on plugin development. And he is having some issues in retrieval of data.

[mrogers]

On the glacier web site it says that "Amazon Simple Storage Service (Amazon S3) allows you to seamlessly move data between Amazon S3 and Amazon Glacier using data lifecycle policies. You can also use AWS Import/Export to accelerate moving large amounts of data into Amazon Glacier using portable storage devices for transport."

So I'm reading that to mean you could use the Alfresco S3 store with glacier in the background.

[abhinav]

Hi Mrogers,

Thanks for reply..
Actually my requirement is to use Alfresco Glacier as a file system..

Alfresco S3 store is an another file system. Sorry if i am wrong.

Can it be possible to use Alfresco Glacier as a file system directly.. or it can be done only in combination with S3 ?

[mrogers]

I think you are going to struggle on this one.    Glacier is an archival system not a file system.    And retrieval of content is asynchronous.

For a start (after a skim read of the glacier API)  I think glacier is generating unique identfiers for content.   Rather than having the identifiers given to it.   So there will need to be some way of persisting that mapping.

Then you could be waiting for a long time for an archive job to complete to begin to read content.  We already have some issues with S3 being slow. Glacier is potentially much worse.

Sticking a CachingContentStore in front of your Glacier store will help.

Its going to be possible but I think you have your work cut out.

[abhinav]

Hi Mrogers,


Thanks alot Mrogers.