[
https://issues.apache.org/jira/browse/HADOOP-14749?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16120177#comment-16120177
]
Steve Loughran commented on HADOOP-14749:
-----------------------------------------
+ feedback from [~ehiggs]
{code}
+ // with a metadata store, the object entries need tup be updated,
Grammar/spelling.
+ * This will always be non-null, but may be bound to the
If something will be not null, maybe use @NotNull. I don’t see any uses of it
yet in the Hadoop codebase, so maybe someone decided against using it.
+ if (status == DirectoryStatus.DOES_NOT_EXIST
+ || status == DirectoryStatus.EXISTS_AND_IS_DIRECTORY_ON_S3_ONLY) {
I think this indents the || one too many. checkstyle should pick it up.
+ // TODO s3guard: retry on file not found exception
Other places you are normalizing spelling to use capital S and capital G (even
in comments) and the nature of this patch is nit fixes... :)
+ * Generally, callers should use {@link #initialize(FileSystem)}
+ * with an initialized S3 file system.
A wise man once said “Object Stores are not File Systems”. So do we want “with
an initialized {@link S3AFileSystem} ? or “initialized S3 FileSystem” so it
includes S3 and S3N (which will will be removed soon).
+ * Without a filesystem to act as a reference point, the configuration itself
file system or filesystem. cf previous comment.
+#### Errpr `"DynamoDB table TABLE does not exist in region REGION;
auto-creation is turned off"`
Error (spelling).
+
+### Warning About Concurrent Tests
+
+You must not run S3A and S3N tests in parallel on the same bucket. This is
+especially true when S3Guard is enabled. S3Guard requires that all clients
+that are modifying the bucket have S3Guard enabled, so having S3N
+integration tests running in parallel with S3A tests will cause strange
+failures.
So if someone adds to the bucket using s3cmd in production what will happen?
This seems like a severe limitation that can effect of ephemeral mounts for
Provided Storage where a purpose is to async repl between s3 and hdfs.
+The two S3Guard scale testse are `ITestDynamoDBMetadataStoreScale` and
tests (spelling)
{code}
> review s3guard docs & code prior to merge
> -----------------------------------------
>
> Key: HADOOP-14749
> URL: https://issues.apache.org/jira/browse/HADOOP-14749
> Project: Hadoop Common
> Issue Type: Sub-task
> Components: documentation, fs/s3
> Affects Versions: HADOOP-13345
> Reporter: Steve Loughran
> Assignee: Steve Loughran
> Attachments: HADOOP-14749-HADOOP-13345-001.patch,
> HADOOP-14749-HADOOP-13345-002.patch
>
> Original Estimate: 24h
> Remaining Estimate: 24h
>
> Pre-merge cleanup while it's still easy to do
> * Read through all the docs, tune
> * Diff the trunk/branch files to see if we can reduce the delta (and hence
> the changes)
> * Review the new tests
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
---------------------------------------------------------------------
To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: common-issues-h...@hadoop.apache.org
