Red Hat Enterprise MRG Messaging and Grid Version 1.1.1

Red Hat Enterprise MRG Messaging and Grid Update 1.1.1 is now available for
Red Hat Enterprise Linux 5.
This new release includes updated amqp, cumin, python-qpid, qpid-java,
qpidc, rhm and rhm-docs packages that fix bugs and add enhancements.

Red Hat Enterprise MRG (Messaging, Real-time and Grid) is a real-time IT
infrastructure for enterprise computing. MRG Messaging implements the
Advanced Message Queuing Protocol (AMQP) standard, adding persistence
options, kernel optimizations, and operating system services.

This update moves Red Hat Enterprise MRG to version 1.1.1 and includes
numerous bug fixes and enhancements.

Messaging enhancements include:

* The C++, JMS, and Ruby clients, and the qpidd broker now support full
SASL using the Cyrus library. This includes GSSAPI for authentication and
SASL negotiated security layers.

* The C++ and Java clients can now use heartbeats sent by the broker to
detect a failed connection/broker more rapidly.

* A new max-session-rate option has been added to qpidd to enable users to
set the maximum publication rate per session.

* A disaster recovery feature was added. Queue changes can now be
replicated from a data center to a passive replica at another site.

* Changes can now be made atomically for an incoming stream of messages
when a queue is bound with a given key in a direct exchange.

* New C++ API for interaction with the Qpid Management Framework

* The throughput and stability of the fault tolerant cluster have been
improved. The cluster now supports all standard broker features with the
exception of distributed transactions.

* Support has been added for Realtime threads in the JMS client.

Messaging bug fixes include:

* Several bugs relating to the stability of the persistence plugin for
qpidd were found and fixed.

* The qpid-config tool now correctly displays the LVQ_NO_BROWSE option for
queues on which that has been configured.

* A bug caused the broker to fail when a message without the required
header was sent to an LVQ. This has been corrected by ensuring that these
messages are handled as FIFO.

* A bug stopping the publication of events by a QMF agent was fixed.

Grid enhancements include:

* The number of virtual CPUs and specific MAC addresses to use for a
virtual job can now be selected using the vm_vcpus parameter.

* Enhanced the displayable name for jobs routed to EC2 reflecting
information about the routing.

* Remote configuration tools have been enhanced to allow listing of
configured nodes and explicit indication of default values during
configuration.

Grid bug fixes include:

* Periodic expressions are not evaluated by the JobRouter for managed
jobs. This provides the ability to stop routing jobs with systematic errors.

* The man and --help pages have been corrected to include a more accurate
description of the -name option arguments.

* Vanilla jobs routed to EC2 are now removed in the proper order during
completion.

* Statistics for jobs routed to EC2 are now reflected in the condor_schedd.

* A particular configuration was causing the condor_startd to stop working.
This error was resolved by causing the daemon to check for a NULL condition.

* Fixed access to exception data in JobRouter hooks for EC2 routing.

* Fixed memory leak in Condor's SOAP interface. Leak resulted in loss of
buffered data on each connection.

* Condor logs, lock files and run files are more tightly integrated with
Red Hat Enterprise Linux standards. Log files once found in ~condor/log/
are now available from /var/log/condor, and condor_master's pid file is
in /var/run/condor.

* Condor's init script was improved to remove an arbitrary 5 second delay
to allow for condor_master to shutdown. A delay is now only present if
the condor_master's shutdown is delayed.

Other bug fixes, feature enhancements and changes included with this update
are noted in the Red Hat Enterprise MRG 1.1 Release Notes, available at the
location noted in the References section below.

All Red Hat Enterprise MRG users are advised to install this updated
release which fixes these bugs, adds these enhancements and introduces
these changes.

Before applying this update, make sure that all previously-released
errata relevant to your system have been applied.

This update is available via Red Hat Network. Details on how to use
the Red Hat Network to apply this update are available at
http://kbase.redhat.com/faq/docs/DOC-11259

441698 - Feature: Support async queue replication
446784 - Missing display of route info for links
452128 - Feature: message throttling
461761 - Make binding changes atomic
461932 - Support heartbeats
464645 - [RFE] Allow virtual jobs to assign virtual cpus.
464647 - [RFE] Add the ability to select a MAC address for a virtual job started in condor.
467878 - Cluster to support message TTL
470286 - Schedd stats graph drawing off right hand side
470615 - Job Details: QDate
471337 - command line tools talk about hostname, they really mean name
471906 - condor init script writes pidfile to /var/lib/condor
472599 - missing slot_type_x config causes dump in startd
472607 - Slot Type definitions only allow lowercase
472900 - Apply thread abstraction patch to the java rpms
472937 - TPL recoverTplStore() failed: jexception 0x0b01 txn_map::get_tdata_list() threw JERR_MAP_NOTFOUND: Key not found in map. (xid=rhm-tid0x2aac2c9b0ee0) (MessageStoreImpl.cpp:1079)
473088 - Cluster does not handle flow-to-disk correctly.
474366 - qpidd+store jcntl::check_owi() threw JERR_JCNTL_OWIMISMATCH on journal recovery
474405 - condor-low-latency daemon init script should change to reflect new package name
474813 - [RFE] customize EC2E routed CMD
475209 - Cluster members hang during start-up of a cluster.
475851 - Cluster passing python test suite.
475865 - Periodic* race in JobRouter (and elsewhere)
475999 - Mint not shutdown by initscript
476544 - cumin - 'APPLICATION ERROR' after cancelled hold job
476716 - Feature: GSSAPI support in ruby client
476759 - QMF Agent stops publishing after running for a time
476895 - When routed job is removed, job router reports confusing message
476984 - update EC2E source job stats from hooks
476996 - JobRouter's status hook output not saved to Schedd
477622 - Feature request java client heart beat (IG Index)
478874 - Feature: Target cluster fail-over time
478882 - Feature: Async DR replication throughput
479033 - C++ API for QMF
479330 - cluster crashes without mcast-max workaround
480051 - qpid-config doesn't display LVQ_NO_BROWSE option
480295 - The broker crashes when a standard message is sent to a LVQ
480841 - When an EC2e job fails it continues to be in the job queue indefinitely.
480990 - init script's stop function always sleeps for 5 seconds
481582 - [RFE] No way to list nodes and their configurations
481583 - Collector Name allowed to be empty via condor_configure_node
481584 - [RFE] Be explicit about default options to condor_configure_node
481771 - Slot count (and idle) not accurate
481772 - "Slots on" graph states are Activities not States
482954 - SOAP memory leak (cursoap not free'd properly)
483125 - sporadic cluster_test errors and failures
484220 - JMS receivenowait does not return expected message
484407 - clustered qpidd broker crashes when trying to start up if management is enabled (==default)
485068 - Consumer with no credit can cause messages not to be seen by other consumers on that queue
485422 - clustered qpidd with persistance - error Error decoding events: framing-error: Framing version unsupported (qpid/framing/AMQFrame.cpp:113)
485682 - Replace use of gethostbyname() with getaddrinfo()/getnameinfo()
485952 - "Edit Attributes" page submit and cancel buttons don't leave the page
486377 - cumin-admin touches cumin.log improperly
486596 - cumin.conf not marked a %config file in cumin.spec
487286 - Ring queue fails with resource-limit-exceeded exception
487355 - Deadlock when trying to close connection
487500 - LVQ does not persist durable messages
487519 - Unable to run low-latency jobs needing access to files only available to owner
488806 - SubscriptionManager::get() closes dispatch queue unexpectedly
489000 - Message body test in carod fails when body is not set
489001 - carod rejects non-Vanilla work
489003 - carod does not trim spaces from attribute names
489005 - Durable subscriptions don't work if clientid has underscores in it
489006 - Cannot distinguish between completion and other termination of AMQP submitted work
489037 - Python type mapping interop issue with Java and C++
489661 - Deadlock between ManagementBroker::userLock and LinkRegistry::lock
489874 - concurrency bug in carod, release() of un-acquire()d lock
489880 - execute directory missing under carod, handle_get_work
490115 - rhm-docs-0.5.752937-1.el5: c++ client examples cannot compile because of Makefile issues
490506 - Joining cluster that has recovered durable messages fails

http://www.redhat.com/security/updates/classification/#none
http://www.redhat.com/docs/en-US/Red_Hat_Enterprise_MRG/