hadoop-common-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Tsz Wo Nicholas Sze (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HADOOP-13103) Group resolution from LDAP may fail on javax.naming.ServiceUnavailableException
Date Thu, 05 May 2016 20:22:13 GMT

    [ https://issues.apache.org/jira/browse/HADOOP-13103?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15273003#comment-15273003
] 

Tsz Wo Nicholas Sze commented on HADOOP-13103:
----------------------------------------------

> ...  The old code had one attempt outside the while loop and then up to 3 more attempts
inside the while loop. With this patch, all attempts are collapsed into a single for loop
with up to 3 iterations. ...

Good observation.  I intentionally make such change so that it can fail fast. For the exception
cases I saw, the first try may fail with various reasons (such as connection timeout).  Then,
the second try always success since it reestablishes the connection.  The third try is almost
redundant but let's give it a try.  It is unlikely that the third try will success after the
first and the second tries have failed.)  The fourth try is just wasting time.

Will update the patch to fix the unit test.

> Group resolution from LDAP may fail on javax.naming.ServiceUnavailableException
> -------------------------------------------------------------------------------
>
>                 Key: HADOOP-13103
>                 URL: https://issues.apache.org/jira/browse/HADOOP-13103
>             Project: Hadoop Common
>          Issue Type: Improvement
>          Components: security
>            Reporter: Tsz Wo Nicholas Sze
>            Assignee: Tsz Wo Nicholas Sze
>            Priority: Minor
>         Attachments: c13103_20160505.patch
>
>
> According to the [javadoc|https://docs.oracle.com/javase/7/docs/api/javax/naming/ServiceUnavailableException.html],
ServiceUnavailableException is thrown when attempting to communicate with a directory or naming
service and that service is not available. It might be unavailable for different reasons.
For example, the server might be too busy to service the request, or the server might not
be registered to service any requests, etc.
> We should retry on it.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

---------------------------------------------------------------------
To unsubscribe, e-mail: common-issues-unsubscribe@hadoop.apache.org
For additional commands, e-mail: common-issues-help@hadoop.apache.org


Mime
View raw message