accumulo-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From els...@apache.org
Subject svn commit: r1729529 - /accumulo/site/trunk/content/release_notes/1.6.5.mdtext
Date Wed, 10 Feb 2016 05:34:32 GMT
Author: elserj
Date: Wed Feb 10 05:34:32 2016
New Revision: 1729529

URL: http://svn.apache.org/viewvc?rev=1729529&view=rev
Log:
Some more release note contents for 1.6.5.

Modified:
    accumulo/site/trunk/content/release_notes/1.6.5.mdtext

Modified: accumulo/site/trunk/content/release_notes/1.6.5.mdtext
URL: http://svn.apache.org/viewvc/accumulo/site/trunk/content/release_notes/1.6.5.mdtext?rev=1729529&r1=1729528&r2=1729529&view=diff
==============================================================================
--- accumulo/site/trunk/content/release_notes/1.6.5.mdtext (original)
+++ accumulo/site/trunk/content/release_notes/1.6.5.mdtext Wed Feb 10 05:34:32 2016
@@ -54,13 +54,33 @@ ACCUMULO_CLIENT_CONF_PATH, and improved
 ACCUMULO_CLIENT_CONF_PATH was set to a directory containing client.conf.
 ([ACCUMULO-4026][ACCUMULO-4026],[ACCUMULO-4027][ACCUMULO-4027])
 
+### Transient ZooKeeper disconnect causes FATE threads to exit
+
+ZooKeeper clients are expected to handle the situation where they become
+disconnected from the ZooKeeper server and must wait to be reconnected
+before continuing ZooKeeper operations.
+
+The dedicated threads running inside the Accumulo Master process for FATE
+actions had the potential unexpectedly exit in this disconnected state.
+This caused a scenario where all future FATE-based operations would
+be blocked until the Accumulo Master process was restarted. ([ACCUMULO-4060][ACCUMULO-4060])
+
+### Incorrect management of certain Thrift RPCs
+
+Accumulo relies on Apache Thrift to implement remote procedure calls between
+Accumulo services. Accumulo's use of Thrift uncovered an unfortunate situation
+where this special RPC (a "oneway" call) would leave unwanted data on the underlying
+Thrift connection. After this extra was left on connection, all subsequent RPCs
+re-using that connection would fail with "out of sequence reponse" error messages.
+Accumulo would be left in a bad state until the mishandled connections were released
+or Accumulo services were restarted. ([ACCUMULO-4065][ACCUMULO-4065])
+
 ## Other Notable Changes
 
  * [ACCUMULO-3509][ACCUMULO-3509] Fixed some lock contention in TabletServer, preventing
resource cleanup
  * [ACCUMULO-3734][ACCUMULO-3734] Fixed quote-escaping bug in VisibilityConstraint
  * [ACCUMULO-4025][ACCUMULO-4025] Fixed cleanup of bulk load fate transactions
- * [ACCUMULO-4065][ACCUMULO-4065] Worked around buggy thrift behavior with oneway methods
- * [ACCUMULO-4070][ACCUMULO-4070] Fixed Kerberos renewal
+ * [ACCUMULO-4070][ACCUMULO-4070] Fixed Kerberos ticket renewal for all Accumulo services
  * [ACCUMULO-4098][ACCUMULO-4098],[ACCUMULO-4113][ACCUMULO-4113] Fixed widespread misuse
of ByteBuffer
 
 ## Testing
@@ -99,6 +119,7 @@ HDFS High-Availability instances, forcin
 [ACCUMULO-4025]: https://issues.apache.org/jira/browse/ACCUMULO-4025
 [ACCUMULO-4026]: https://issues.apache.org/jira/browse/ACCUMULO-4026
 [ACCUMULO-4027]: https://issues.apache.org/jira/browse/ACCUMULO-4027
+[ACCUMULO-4060]: https://issues.apache.org/jira/browse/ACCUMULO-4060
 [ACCUMULO-4065]: https://issues.apache.org/jira/browse/ACCUMULO-4065
 [ACCUMULO-4066]: https://issues.apache.org/jira/browse/ACCUMULO-4066
 [ACCUMULO-4070]: https://issues.apache.org/jira/browse/ACCUMULO-4070



Mime
View raw message