[ 
https://issues.apache.org/jira/browse/HADOOP-14936?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16251267#comment-16251267
 ] 

Steve Loughran commented on HADOOP-14936:
-----------------------------------------

I think we should look through the list of things we consider essential & are 
still on the phase II list & make them the blockers for this. Or simply: plan 
for phase II & Hadoop 3.1 being the out-of-experimentation phase; anything 
which isn't essential for that can be postponed.

# I consider HADOOP-15035 to be a blocker 
# We should be confident that either the DDB format is stable, or that we have 
an upgrade strategy which works
# And we may want to consider auth mode to still be experimental.

S3Guard hasn't been in people's hands as part of an ASF release, and while two 
commercial vendors of hadoop based products have offered S3Guard in some format 
over the summer, we haven't seen that much in terms of stack traces coming 
back. I usually view that as the metric of use. Certainly we didn't get 
anything in the 3.0 beta phase.

This is similar in my mind to the S3A client in general: it shipped in 2.6, but 
it was only once it was out that some of the scale problems surfaced (many 
files, reading-to-EOF on stream close. HADOOP-11571 shows all that surfaced 
before we could really declare S3A ready for use. 

> S3Guard: remove "experimental" from documentation
> -------------------------------------------------
>
>                 Key: HADOOP-14936
>                 URL: https://issues.apache.org/jira/browse/HADOOP-14936
>             Project: Hadoop Common
>          Issue Type: Sub-task
>          Components: fs/s3
>    Affects Versions: 3.0.0-beta1
>            Reporter: Aaron Fabbri
>            Priority: Minor
>
> I think it is time to remove the "experimental feature" designation in the 
> site docs for S3Guard.  Discuss.



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to