[Pacemaker] ANNOUNCE: Pacemaker 0.6.4 now available (bug-fix only release)

[Andrew Beekhof]

Testing is complete which means its time for another Pacemaker bugfix- 
only release.

Packages for Pacemaker and it's immediate dependancies can be  
downloaded from the usual location:
   http://software.opensuse.org/download/server:/ha-clustering

The source can be obtained from:
  http://hg.clusterlabs.org/pacemaker/stable-0.6/archive/Pacemaker-0.6.4.tar.bz2

For a breakdown of the available packages and what parts are optional,  
please refer to
  http://clusterlabs.org/mw/Install#Package_List


The next 0.6 release (if needed) will be mid-June.


In other news, I hope to have the 0.7 (unstable) series ready for  
people to try in the next couple of weeks.
I'm really excited about some of the features we've been adding.  Some  
notable changes for 0.7 include:

   + Overhaul of the configuration syntax
   + Better validation of configurations with RelaxNG schemas
   + Admin tools now set options consistently
   + A new tool for testing your cluster before it goes into production
   + A new tool for making, testing and verifying changes _before_  
uploading them to the cluster
   + Ability to automatically upgrade to the latest configuration syntax
   + Adopt libxml2 as the XML processing library (this should be  
faster and reduce the amount or copying that was occurring)
   + Implement XPath based versions of query/delete/replace/modify
     Eg. Find _all_ options telling resources to stop: cibadmin -- 
query --xpath '//nvpair[@name="target-role" and value="Stopped"]'
   + Bug LF:1459 - Allow failures to expire
   + Replace the failure-stickiness mess with a simple 'migration- 
threshold'
   + Bug LF:1438 - Allow the list of all/active resource operations to  
be queried by crm_resource
   + Bug LF:1883 - crm_mon should display fail-count and operation  
history
   + Bug LF:1884, Bug LF:1885 - Implement N:M ordering and colocation  
constraints
   + Implement a client-side operation timeout for CIB connections
   + Implementing the ability for clients to connect to remote  
instances of the CIB
   + Implement a stack-independant version of pingd - no need to add  
ping nodes to ha.cf

regards,
Andrew


Begin lazy paste from the changelog...

   * Update source tarball to revision: 226d8e356924 tip
   * Statistics:
       Changesets:       55
       Diff:             199 files changed, 7103 insertions(+), 12378  
deletions(-)

   * Changes since Pacemaker-0.6.3
   + High: crmd: Bug LF:1881 LF:1882 - Overhaul the logic for  
operation cancelation and deletion
   + High: crmd: Bug LF:1894 - Make sure cancelled recurring  
operations are cleaned out from the CIB
   + High: PE: Bug N:387749 - Colocation with clones causes  
unnecessary clone instance shuffling
   + High: PE: Ensure 'master' monitor actions are cancelled _before_  
we demote the resource
   + High: PE: Fix assert failure leading to core dump - make sure  
variable is properly initialized
   + High: PE: Make sure 'slave' monitoring happens after the resource  
has been demoted
   + High: PE: Prevent failure stickiness underflows (where too many  
failures become a _positive_ preference)
   + Medium: Admin: crm_mon - Only complain if the output file could  
not be opened
   + Medium: Common: filter_action_parameters - enable legacy handling  
only for older versions
   + Medium: PE: Bug N:385265 - The failure stickiness of group  
children is ignored until it reaches -INFINITY
   + Medium: PE: Implement master and clone colocation by exlcuding  
nodes rather than setting one's score to INFINITY (similar to cs:  
756afc42dc51)
   + Medium: TE: Bug LF:1875 - Correctly find actions to cancel when  
their node leaves the cluster