Home | About | Sematext search-lucene.com search-hadoop.com
clear query|facets|time Search criteria: .   Results from 21 to 30 from 125 (0.069s).
Loading phrases to help you
refine your search...
[HADOOP-10146] Workaround JDK7 Process fd close bug - Hadoop - [issue]
...JDK7's Process output streams have an async fd-close race bug.  This manifests as commands run via o.a.h.u.Shell causing threads to hang, OOM, or cause other bizarre behavior.  The...
http://issues.apache.org/jira/browse/HADOOP-10146    Author: Daryn Sharp, 2014-01-17, 13:28
[HADOOP-9363] AuthenticatedURL will NPE if server closes connection - Hadoop - [issue]
...A NPE occurs if the server unexpectedly closes the connection for an AuthenticatedURL w/o sending a response....
http://issues.apache.org/jira/browse/HADOOP-9363    Author: Daryn Sharp, 2014-01-16, 19:30
[HADOOP-10233] RPC lacks output flow control - Hadoop - [issue]
...The RPC layer has input flow control via the callq, however it lacks any output flow control.  A handler will try to directly send the response.  If the full response is not sent t...
http://issues.apache.org/jira/browse/HADOOP-10233    Author: Daryn Sharp, 2014-01-14, 23:48
[HADOOP-10173] Remove UGI from DIGEST-MD5 SASL server creation - Hadoop - [issue]
...Instantiation of SASL server instances within the readers threads is performed within a UGI.getCurrentUser().doAs.  getCurrentUser is synchronized, and doAs also degrades performance. &...
http://issues.apache.org/jira/browse/HADOOP-10173    Author: Daryn Sharp, 2014-01-03, 14:47
[HADOOP-10172] Cache SASL server factories - Hadoop - [issue]
...Performance for SASL server creation is atrocious.  Sasl.createSaslServer does not cache the provider resolution for the factories.  Factory resolution and server instantiation has...
http://issues.apache.org/jira/browse/HADOOP-10172    Author: Daryn Sharp, 2013-12-20, 15:09
[HADOOP-10174] Run RPC server within the Subject that instantiated it - Hadoop - [issue]
...RPC servers would not require as many doAs blocks if the server threads run within the access control context that instantiates the server....
http://issues.apache.org/jira/browse/HADOOP-10174    Author: Daryn Sharp, 2013-12-17, 19:24
Re: How can I get FSNamesystem of running NameNode in cluster? - Hadoop - [mail # dev]
...Are you adding something internal to the NN?  If not, you cannot get the namesystem instance via a client unless you are using a minicluster object.  Daryn  On Dec 9, 2013, at...
   Author: Daryn Sharp, 2013-12-09, 14:42
[HADOOP-9955] RPC idle connection closing is extremely inefficient - Hadoop - [issue]
...The RPC server listener loops accepting connections, distributing the new connections to socket readers, and then conditionally & periodically performs a scan for idle connections.  ...
http://issues.apache.org/jira/browse/HADOOP-9955    Author: Daryn Sharp, 2013-11-15, 13:28
[HADOOP-10099] Reduce chance for RPC denial of service - Hadoop - [issue]
...A RPC server may accept an unlimited number of connections unless indirectly bounded by a blocking operation in the RPC handler threads.  The NN's namespace locking happens to cause thi...
http://issues.apache.org/jira/browse/HADOOP-10099    Author: Daryn Sharp, 2013-11-14, 19:28
Re: [VOTE] Release Apache Hadoop 2.2.0 - Hadoop - [mail # dev]
...+1  I did some basic testing and it appears to work fine.  Daryn  On Oct 7, 2013, at 2:00 AM, Arun C Murthy wrote:  ...
   Author: Daryn Sharp, 2013-10-07, 15:28
Hadoop (125)
HDFS (93)
MapReduce (41)
issue (104)
mail # dev (10)
mail # general (7)
mail # user (4)
last 7 days (4)
last 30 days (13)
last 90 days (20)
last 6 months (29)
last 9 months (125)
Harsh J (537)
Owen O'Malley (402)
Steve Loughran (357)
Todd Lipcon (233)
Eli Collins (181)
Arun C Murthy (156)
Chris Nauroth (129)
Alejandro Abdelnur (121)
Allen Wittenauer (115)
Nigel Daley (112)
Tom White (110)
Daryn Sharp (108)
Konstantin Shvachko (102)
Ted Yu (95)
Aaron Kimball (93)