You are viewing a plain text version of this content. The canonical link for it is here.
Posted to notifications@accumulo.apache.org by GitBox <gi...@apache.org> on 2018/12/31 19:03:11 UTC
[GitHub] ctubbsii commented on issue #862: Fix TokenFile in MapReduce
ctubbsii commented on issue #862: Fix TokenFile in MapReduce
URL: https://github.com/apache/accumulo/pull/862#issuecomment-450680895
> I agree. I think the only way to fix this in the old API is to revert how we serialize the client properties.
I think that's best.
> How do you propose handling the token file in the new API?
In the new API, we do not need to have a "Token File" feature at all. We already support serializing credentials into a client properties file. We just need to make sure that if a user sets the client properties file (vs. setting a token directly in the fluent API), we do not immediately read that file and store its contents as properties in the job. We should handle it either by putting the file in the distributed cache, or simply serialize the path to the file around so we can read it from HDFS directly in the input/output formats. This would basically mean protecting the contents of the client properties file in the new API the same way we treated the token file in the old API.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
With regards,
Apache Git Services