You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@flink.apache.org by "Ufuk Celebi (JIRA)" <ji...@apache.org> on 2017/03/08 09:41:38 UTC

[jira] [Created] (FLINK-5993) Add section about Azure deployment

Ufuk Celebi created FLINK-5993:
----------------------------------

             Summary: Add section about Azure deployment 
                 Key: FLINK-5993
                 URL: https://issues.apache.org/jira/browse/FLINK-5993
             Project: Flink
          Issue Type: Improvement
          Components: Documentation
            Reporter: Ufuk Celebi


Running Flink on Azure can lead to unexpected problems. From the user mailing list:
{quote}
For anyone seeing this thread in the future, we managed to solve the issue. The problem was in the Azure storage SDK.
Flink is using Hadoop 2.7, so we used version 2.7.3 of the Hadoop-azure package. This package uses version 2.0.0 of the azure-storage package, dated from 2014. It has several bugs that were since fixed, specifically one where the socket timeout was infinite. We updated this package to version 5.0.0 and everything is working smoothly now.
{quote}
(http://apache-flink-user-mailing-list-archive.2336050.n4.nabble.com/Flink-checkpointing-gets-stuck-td11776.html)

At least this caveat should be covered.



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)