hltd-1.6.0-0

Version installed in cdaq on Jan 27th 2015.
Changes:

• support for running multiple instances of hltd on BU using loopback FS mounted in ramdisk
  description: #58
• resource summary json file in ramdisk contains information aggregated from boxinfo files
  description: #59
• blacklisting by BU hltd (updated each run). Currently only reading from local file /etc/appliance/blacklist, which is a json file containing a vector (e.g. ["host1","host2"])
• support for excluding/including FU from global runs for cloud (still not considered stable interface and subject to change). Integrated with CMSSW support to quit processing a run after a certain lumisection.
• meta rpm support for tribe server configuration
• addition of python-procname library which allows renaming a process from a default "python" name
• improved hltd init script with color highlighting, management of multiple instances if they are present
  (e.g. "/sbin/service hltd restart instance1")
• other bugfixes and improvements, full commit log:
  #60

Known bugs (temporarily patched in production machines):

• IP rotation on central ES connection errors is broken when writing box info files due to wrong placement of a break statement.
• when unable to write box file, FU will try a remount. This code was added for triggering remount when loopback instance mountpoint on BU is recreated, however it caused problems in case of legacy FUs that had data network problem.

Note: jsonMerger rpm v1.2 is provided with release 1.5.3-6.