You are viewing a plain text version of this content. The canonical link for it is here.
Posted to reviews@spark.apache.org by GitBox <gi...@apache.org> on 2020/11/18 14:20:35 UTC
[GitHub] [spark] steveloughran commented on pull request #30366: [WIP][SPARK-33440][CORE] Separate the calculation for the next renewal date for each delegation token
steveloughran commented on pull request #30366:
URL: https://github.com/apache/spark/pull/30366#issuecomment-729709344
Classic HDFS tokens can be renewed generally up to a week and also revoked. The S3ADT is which we ship with Apache Hadoop are not renewable at all… Which is why I hadn't noticed I was getting this wrong. The Knox DTs are renewable as presumably, are those which HD/Insights supports. I don't know if they are doing the right thing. One recommendation could be to actually print a warning if a token has a an issue date of zero.
As to what makes a good default, the usual use case of tokens collected in Spark application launch using now as the default issue date seems absolutely perfect. It is only in the special case of tokens collected in advance and the token implementation doesn't set issue date that there would be any problem from this.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to 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
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org