jackrabbit-oak-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Amit Jain (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (OAK-7887) [S3DataStore] HTTP 409 returned S3 in init() can cause init() failure
Date Mon, 12 Nov 2018 06:15:00 GMT

    [ https://issues.apache.org/jira/browse/OAK-7887?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16683260#comment-16683260

Amit Jain commented on OAK-7887:

{quote}I wonder if a better place to do this would be in the test code. When the tests start
we could use similar logic to make sure a bucket is created for the test. Then we don't burden
the production code with this logic, but instead allow a simple try and fail if the bucket
can't be created the first time. The odds of this error happening in production seem extremely
Yes, my thinking as well. But do you know what tests caused this? IIRC, we create unique bucket
names. Does this occur even then? or there are places we have missed creating the unique buckets?

> [S3DataStore] HTTP 409 returned S3 in init() can cause init() failure
> ---------------------------------------------------------------------
>                 Key: OAK-7887
>                 URL: https://issues.apache.org/jira/browse/OAK-7887
>             Project: Jackrabbit Oak
>          Issue Type: Bug
>          Components: blob-cloud
>            Reporter: Matt Ryan
>            Assignee: Matt Ryan
>            Priority: Minor
>         Attachments: OAK-7887.patch.1
> When running S3DataStore tests on Jenkins CI, about 10% of the tests failed for the same
reason:  a 409 error code returned from Amazon during the {{init()}} method of {{S3Backend}}. 
I suspect this is being caused by the call to create the bucket if it doesn't already exist.
> To fix this we would need to implement some retry logic in the {{init()}} call.

This message was sent by Atlassian JIRA

View raw message