In the storage plugin configuration that you have for S3, the connection URL can be like “s3n://accessKeyId:secretKeyId@bucket”. AFAIK, there is nothing stopping you from creating multiple such configurations pointing to the same bucket with different credentials.
These can even be created dynamically with a REST invocation. It’s the “on the fly” part of your question that is your challenge, I think. Maybe you can get by without that, or maybe you can safely automate the above. On 7/13/15, 3:01 PM, "Hafiz Mujadid" <[email protected]> wrote: >paul can you guide me further what are u trying to say? > > > >On Mon, Jul 13, 2015 at 11:46 PM, Paul Mogren <[email protected]> >wrote: > >> Your question seems to be about clients passing credentials to a Drill >> query which are then passed through to resources as they are accessed. I >> don¹t think you are going to find that. You can put the credentials in >>the >> URL in storage plugin configuration instead, if that helps. Storage >> plugins can be dynamically managed, at least. >> >> >> >> >> >> On 7/13/15, 2:19 AM, "Ted Dunning" <[email protected]> wrote: >> >> >On Sun, Jul 12, 2015 at 9:41 PM, Hafiz Mujadid >><[email protected]> >> >wrote: >> > >> >> I successfully connected Drill to S3 by placing access and secret >>keys >> >>in >> >> core-site.xml. >> >> >> >> Is it possible to use Drill with S3 without hardcoding credentials >>into >> >> core-site like defining credentials for multiple users on the fly? >> >> >> > >> >Not sure if there is. >> > >> >But please do make sure that you issue IAM credentials for this purpose >> >that are highly limited in what they can do. >> >> > > >-- >Regards: HAFIZ MUJADID
