hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Anatoli Shein (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HDFS-11767) libhdfs++: SASL events should be scoped closer to usage
Date Wed, 10 May 2017 19:04:04 GMT

    [ https://issues.apache.org/jira/browse/HDFS-11767?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16005230#comment-16005230
] 

Anatoli Shein commented on HDFS-11767:
--------------------------------------

+1. The patch looks good. It is useful for resolving race condition in SaslProtocol.

> libhdfs++: SASL events should be scoped closer to usage
> -------------------------------------------------------
>
>                 Key: HDFS-11767
>                 URL: https://issues.apache.org/jira/browse/HDFS-11767
>             Project: Hadoop HDFS
>          Issue Type: Sub-task
>          Components: hdfs-client
>            Reporter: James Clampffer
>            Assignee: James Clampffer
>         Attachments: HDFS-11767.HDFS-8707.000.patch
>
>
> "SASL Start" and "SASL End" are called in the SaslProtocol constructor and destructor.
 SaslProtocol is held by a shared_ptr so there's a race where the RpcConnection can be torn
down and another can be put in it's place while a callback is still holding the shared_ptr
preventing "SASL End" from being called.
> This can matter a lot because both Cyrus SASL and gsasl aren't thread safe during negotiation.
 These event hooks provide a mechanism to pick up a process wide lock that guards access to
the sasl library.



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

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


Mime
View raw message