[
https://issues.apache.org/jira/browse/HADOOP-19027?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17803659#comment-17803659
]
Steve Loughran commented on HADOOP-19027:
-----------------------------------------
The retry logic in S3AInputStream only seems to retry on the first GET, later
ones is treats as fatal and assumes this is a file version issue.
This is from https://github.com/apache/hadoop/pull/794
{code}
// With S3Guard, the metadatastore gave us metadata for the file in
// open(), so we use a slightly different retry policy, but only on initial
// open. After that, an exception generally means the file has changed
// and there is no point retrying anymore.
Invoker invoker = context.getReadInvoker();
invoker.maybeRetry(streamStatistics.openOperations == 0,
"lazySeek", pathStr, true, ...
{code}
# We want to retry on failures, except for version change events.
# we need to map stream problems (closed, no response) to retryable
> S3A: S3AInputStream doesn't recover from HTTP exceptions
> --------------------------------------------------------
>
> Key: HADOOP-19027
> URL: https://issues.apache.org/jira/browse/HADOOP-19027
> Project: Hadoop Common
> Issue Type: Sub-task
> Components: fs/s3
> Affects Versions: 3.4.0
> Reporter: Steve Loughran
> Assignee: Steve Loughran
> Priority: Major
>
> S3AInputStream doesn't seem to recover from Http exceptions raised through
> HttpClient or through OpenSSL.
> * review the recovery code to make sure it is retrying enough, it looks
> suspiciously like it doesn't
> * detect the relevant openssl, shaded httpclient and unshaded httpclient
> exceptions, map to a standard one and treat as comms error in our retry policy
> This is not the same as the load balancer/proxy returning 443/444 which we
> map to AWSNoResponseException. We can't reuse that as it expects to be
> created from an
> {{software.amazon.awssdk.awscore.exception.AwsServiceException}} exception
> with the relevant fields...changing it could potentially be incompatible.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
