Next: 10.4 Development Release Series
Up: 10. Version History and
Previous: 10.2 Upgrading from the
Contents
Index
Subsections
10.3 Stable Release Series 8.4
This is a stable release series of HTCondor.
As usual, only bug fixes (and potentially, ports to new platforms)
will be provided in future 8.4.x releases.
New features will be added in the 8.5.x development series.
The details of each version are described below.
Version 8.4.1
Release Notes:
- HTCondor version 8.4.1 released on October 27, 2015.
New Features:
- Added support to allow an admin to always volume mount
certain directories into docker universe containers running
on a host.
(Ticket #5308).
- Added four policy metaknobs to simplify configuring a policy
to either preempt or hold jobs that use more memory
or CPU cores than provisioned in the slot. See the POLICY
category of metaknobs in section 3.3.1 for
additional information.
(Ticket #5250).
- Added configuration variables and documentation so that we uniformly prefer
<var>_ATTRS over <var>_EXPRS but support both. This includes
STARTD_ATTRS, STARTD_JOB_ATTRS and SUBMIT_ATTRS
which are often used by HTCondor sites which customize the configuration. These
configuration variables are now exclusively for use by HTCondor administrators;
The former default values for these variables have been moved into other configuration
which is reserved for use by HTCondor developers. This is done to prevent administrators
from accidently removing the necessary defaults.
A warning about use of STARTD_EXPRS has been disabled unless
STARTD_ATTRS orSLOT_TYPE_<n>_STARTD_ATTRS is also used, since
the use all three of these at the same time is not supported.
(Ticket #5326).
- When condor_reconfig and condor_restart are run as root
they will check to see if the condor user has read access to all of the
configuration files before sending the command. This is done to prevent aborting the daemons
accidently by sending reconfig after the admin creates a new config file and
forgets to give the condor user read access to that file.
(Ticket #4506).
- Added the -natural sort option to condor_status to sort the slots
in numerical order rather than alphabetical order.
(Ticket #5131).
Bugs Fixed:
- When cgroups are enabled, and CGROUP_MEMORY_POLICY is soft,
HTCondor now also sets the hard limit to the virtual memory limit
of the job, if there is one.
(Ticket #5280).
- If cgroups are enabled, and a job goes over the memory limit, the
cgroup OOM killer fires, and the job is put on hold. HTCondor now
updates the job's memory usage statistics with the most up to date
usage, instead of relying on the previous snapshop.
(Ticket #5341).
- Fixed a bug where the condor_kbdd could not accurately measure
the keyboard idle time. This daemon now works correctly on Linux systems
whose X server support the MIT screen saver extension.
(Ticket #5265).
- Fixed a bug which prevented SOAP submissions.
(Ticket #5260).
- The parameter STARTD_HISTORY is now set to record
the job histories per startd, in the log directory of the
execute machine. These can be read with the condor_history command.
Previously the default was not to record these.
(Ticket #5257).
- The parameter SCHEDD_USE_SLOT_WEIGHT now defaults to
true, so that SLOT_WEIGHT can be used with hierarchical group quotas
and partitionable slots.
(Ticket #5256).
- Fixed bug whereby occasionally the command-line tools would
emit debug messages to stderr with text ``I am: hostname: ...''.
(Ticket #5276).
- Fixed a bug that prevented node retries from working on DAG
nodes that are DAG-level NOOP nodes. (This bug has existed at least since
the 8.2 series.)
(Ticket #5277).
- Fixed a problem when the HTCondor executables were not compiled
with RPATH enabled on Enterprise Linux 6 platforms. RPATH is used
to load Globus and other libraries from the condor-externals RPM.
(Ticket #5294).
- The job attribute JobCurrentStartTransferOutputDate is now
properly reported in the job ad.
(Ticket #5298).
- Fixed configuration parameter NETWORK_HOSTNAME, which was
broken starting with version 8.3.2.
(Ticket #5288).
- Fixed a bug that could cause the Job Router to crash when
invoking a transformation hook.
(Ticket #5224).
- Fixed several memory leaks in the nordugrid_gahp.
(Ticket #5322).
- Improved the batch_gahp to better handle batch systems
that reuse job IDs.
(Ticket #5062).
- When the batch_gahp rejects a request because it is
overloaded, the condor_gridmanager now reduces the rate of requests
and retries the rejected request later.
(Ticket #5253).
- Fixed a bug that could cause a daemon to be in the wrong
privilege state when attempting to act as the user.
(Ticket #4522).
- The condor_had and condor_replication daemons now work properly
when Shared Port is enabled. They still require their own dedicated ports.
(Ticket #5301).
- Fixed a bug that cause condor_mips to report numbers about 40 percent lower than it should
on Linux platforms.
(Ticket #5261).
- Fixed a bug in condor_install that would cause it to configure HTCondor to advertise the public
IP addresses to the collector even when using localhost or 127.0.0.1 for a personal HTCondor.
(Ticket #5286).
- Fixed a bug in condor_q that caused slices in the Queue statement to be treated as part of the
arguments filename when the slice was longer than 8 characters.
(Ticket #5273).
- Added SELinux type enforcment rules to allow the condor_schedd to be
able to access user files in NFS mounted file systems.
(Ticket #5343).
Known Issue:
- TCP_FORWARDING_HOST is disregarded by HTCondor clients
starting in version 8.3.6. This bug will be fixed in version 8.4.2 and 8.5.1.
(Ticket #5339).
Version 8.4.0
Release Notes:
- HTCondor version 8.4.0 released on September 14, 2015.
New Features:
Bugs Fixed:
- Fixed a bug introduced in HTCondor version 8.3.7 that caused the
condor_shared_port daemon to leak file descriptors.
Also made HTCondor work better when some HTCondor daemons
are using shared port, but the condor_master is not.
(Ticket #5259).
- The condor_starter lowers the OOM (out of memory) score of jobs
so the OOM killer is more likely to chose an HTCondor job rather than
an HTCondor daemon or other user process.
(Ticket #5249).
- Job submission fails if X.509 certificates are advertised with EC2
grid universe jobs.
Therefore EC2 grid universe jobs no longer advertise their access keys.
(Ticket #5252).
Next: 10.4 Development Release Series
Up: 10. Version History and
Previous: 10.2 Upgrading from the
Contents
Index