[
https://issues.apache.org/jira/browse/HADOOP-13786?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Steve Loughran updated HADOOP-13786:
------------------------------------
Attachment: HADOOP-13786-HADOOP-13345-024.patch
Patch 024
The staging committers have all moved over to working with S3A via the S3A FS;
write operation helper methods have been added/altered as appropriate. This
pulls them in to the instrumentation & metrics, sets things up for s3guard
(TODO). The hardest part of the migration was the mock tests, as they've been
very brittle to changes inside. I've pulled TestMRStagingJob altogether, as its
copy and upgrade to an IT test obsoletes it. The others are now working again.
and all the helper classes cleaned up as they don't have an s3 client to pass
around. Tests need to know that client errors/results are set in the FS, not
specific committers.
I need to think about what to do with mock directories in the commit process. I
Don't want to have the code generating needless fake empty dirs, for the perf
hit, or generating needless delete requests. I think I'd like
# the initial commit process to create the dest dir as a mock empty dir. This
stops race conditions between >1 committer to same test
# individual file commits to do nothing about the parent dir
# end of job commit to explicitly delete the parent dirs if present, if the
commit succeeded.
I'll just add another action in WriteOperationsHelper here.
> Add S3Guard committer for zero-rename commits to consistent S3 endpoints
> ------------------------------------------------------------------------
>
> Key: HADOOP-13786
> URL: https://issues.apache.org/jira/browse/HADOOP-13786
> Project: Hadoop Common
> Issue Type: New Feature
> Components: fs/s3
> Affects Versions: HADOOP-13345
> Reporter: Steve Loughran
> Assignee: Steve Loughran
> Attachments: HADOOP-13786-HADOOP-13345-001.patch,
> HADOOP-13786-HADOOP-13345-002.patch, HADOOP-13786-HADOOP-13345-003.patch,
> HADOOP-13786-HADOOP-13345-004.patch, HADOOP-13786-HADOOP-13345-005.patch,
> HADOOP-13786-HADOOP-13345-006.patch, HADOOP-13786-HADOOP-13345-006.patch,
> HADOOP-13786-HADOOP-13345-007.patch, HADOOP-13786-HADOOP-13345-009.patch,
> HADOOP-13786-HADOOP-13345-010.patch, HADOOP-13786-HADOOP-13345-011.patch,
> HADOOP-13786-HADOOP-13345-012.patch, HADOOP-13786-HADOOP-13345-013.patch,
> HADOOP-13786-HADOOP-13345-015.patch, HADOOP-13786-HADOOP-13345-016.patch,
> HADOOP-13786-HADOOP-13345-017.patch, HADOOP-13786-HADOOP-13345-018.patch,
> HADOOP-13786-HADOOP-13345-019.patch, HADOOP-13786-HADOOP-13345-020.patch,
> HADOOP-13786-HADOOP-13345-021.patch, HADOOP-13786-HADOOP-13345-022.patch,
> HADOOP-13786-HADOOP-13345-023.patch, HADOOP-13786-HADOOP-13345-024.patch,
> s3committer-master.zip
>
>
> A goal of this code is "support O(1) commits to S3 repositories in the
> presence of failures". Implement it, including whatever is needed to
> demonstrate the correctness of the algorithm. (that is, assuming that s3guard
> provides a consistent view of the presence/absence of blobs, show that we can
> commit directly).
> I consider ourselves free to expose the blobstore-ness of the s3 output
> streams (ie. not visible until the close()), if we need to use that to allow
> us to abort commit operations.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
