[ 
https://issues.apache.org/jira/browse/IMPALA-7738?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16665504#comment-16665504
 ] 

Joe McDonnell commented on IMPALA-7738:
---------------------------------------

For Open(), we only care about the file handle, and if the call blocks, it 
shouldn't hold up any memory (or it doesn't have to). For Read(), the buffer 
that we are reading into becomes toxic, because if the read ever succeeds, it 
will write to that memory. (Killing a hung thread sounds hard / error prone 
(based on 5 mins looking around), so I'm assuming we're just letting the thread 
run.)

I looked at hdfs.h, and there are no async calls or calls with timeouts.

> Implement timeouts for HDFS calls
> ---------------------------------
>
>                 Key: IMPALA-7738
>                 URL: https://issues.apache.org/jira/browse/IMPALA-7738
>             Project: IMPALA
>          Issue Type: Improvement
>          Components: Backend
>    Affects Versions: Impala 2.7.0, Impala 2.8.0, Impala 2.9.0, Impala 2.10.0, 
> Impala 2.11.0, Impala 3.0, Impala 2.12.0
>            Reporter: Michael Ho
>            Priority: Critical
>
> Currently, there is no timeout with the various HDFS calls (e.g. hdfsOpen(), 
> hdfsRead()) we made in libhdfs.so in either the disk-io-mgr thread or scanner 
> thread context. Various users of Impala have complaint in the past about hung 
> queries which eventually boiled down to stuck hdfs calls. HDFS maintainers 
> have been slow to find the root cause of those hangs. To make this kind of 
> stuck queries problem easier to identify in the future, we should just 
> enforce a timeout in various hdfs calls so the queries will fail when certain 
> HDFS calls take longer than a designated timeout period.
> There may be multiple layers which this timeout can be enforced:
>  * at Impala level, we can have a fixed sized thread pool which handles all 
> hdfs calls. The existing hdfs calls will be a wrapper with a timeout.
>  * at libhdfs.so, enforce a timeout at places in the HDFS client code which 
> may block forever.
> The second option is probably beyond the charter of Apache Impala project.
> cc'ing [~tarmstr...@cloudera.com], [~joemcdonnell]



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-all-unsubscr...@impala.apache.org
For additional commands, e-mail: issues-all-h...@impala.apache.org

Reply via email to