[
https://issues.apache.org/jira/browse/HADOOP-13761?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16340137#comment-16340137
]
Aaron Fabbri commented on HADOOP-13761:
---------------------------------------
Attached v1 RFC patch that just adds a test that attempts to reproduce the
ProvisionedThroughputExceededException that [[email protected]] reported
before. Look for testBatchedWriteExceedsProvisioned(). I moved the DDB scale
test to the .s3guard package for better access to testing functions in DDB
metastore.
I have not been able to reproduce the SDK failing to do its own retries. I'm
wondering if [[email protected]] would have more luck getting it to
reproduce.. Maybe there are differences in DDB service in the region you use
(I'm in us-west-2). IIRC the SDK version is still the same as when you posted
that stack trace?
> S3Guard: implement retries for DDB failures and throttling; translate
> exceptions
> --------------------------------------------------------------------------------
>
> Key: HADOOP-13761
> URL: https://issues.apache.org/jira/browse/HADOOP-13761
> Project: Hadoop Common
> Issue Type: Sub-task
> Components: fs/s3
> Affects Versions: 3.0.0-beta1
> Reporter: Aaron Fabbri
> Assignee: Aaron Fabbri
> Priority: Blocker
> Attachments: HADOOP-13761.001.patch
>
>
> Following the S3AFileSystem integration patch in HADOOP-13651, we need to add
> retry logic.
> In HADOOP-13651, I added TODO comments in most of the places retry loops are
> needed, including:
> - open(path). If MetadataStore reflects recent create/move of file path, but
> we fail to read it from S3, retry.
> - delete(path). If deleteObject() on S3 fails, but MetadataStore shows the
> file exists, retry.
> - rename(src,dest). If source path is not visible in S3 yet, retry.
> - listFiles(). Skip for now. Not currently implemented in S3Guard. I will
> create a separate JIRA for this as it will likely require interface changes
> (i.e. prefix or subtree scan).
> We may miss some cases initially and we should do failure injection testing
> to make sure we're covered. Failure injection tests can be a separate JIRA
> to make this easier to review.
> We also need basic configuration parameters around retry policy. There
> should be a way to specify maximum retry duration, as some applications would
> prefer to receive an error eventually, than waiting indefinitely. We should
> also be keeping statistics when inconsistency is detected and we enter a
> retry loop.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
