Github user JoshRosen commented on the pull request:
https://github.com/apache/spark/pull/1120#issuecomment-54852214
I think that `boto` will read its credentials from those environment
variables [or through configuration
files](http://boto.readthedocs.org/en/latest/boto_config_tut.html#details), so
those environment variables aren't necessarily the only way to set these keys.
What will happen if I use those? Do we want to allow users to use separate
launch vs. S3 credentials, or should we always copy the AWS credentials used to
launch the cluster, irrespective of whether they were supplied via environment
variables, command-line flags, or boto configs?
Also, just to be paranoid, is there any security concern with implicitly
copying the launch credentials to the cluster? Users who read from S3 still
end up embedding credentials into their programs / environment variables /
launch commands on the server, so I don't whether this is a big reduction in
security (except now users who don't read from S3 will end up with these
credentials on the server).
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
