Bugs fixed in SGE 6.0u11 since release 6.0u10
--------------------------------------------
scalability issue with qdel and very large array jobs
loadcheck reports on a hyperthreaded CPU only one processor
CSP mode auto installation: certificates are not copied to submit hosts
qmon crashes when displaying about dialog
sge_schedd deadlock upon schedd_job_info job_list being enabled
Qmon has a ticket number limitation
qmon on secondary screen crashes when "Job Control" is pressed
queue status in reporting file is missing.
"Modify"-Button is activated but should be grayed
reporting(5) man page lacks information about sharelog records
Ability to sort queue instances using each column of the queue instances
table
Queue request -q from sge_request can't be overridden through command line
qdel on a job in deleted state does not output any information
qsub -terse to just output job id
qrsh man page doesn't explain which options don't work with interactive jobs
Numbers in error mail too large
still used usersets/project/calendar/pe/checkpoint can be removed under
certain conditions
remove util/sge_log_tee from distribution
shell_start_mode should be documentated to be only used for batch jobs
reserved usage not explained in sge_conf.5
Discrepancy between load values reported by Gridengine and from the
HP-UX 64 bit env.
auto installation option failed to save the install log
function job_get_id_string() is not MT save and used in qmaster
Broken output of job name with 'qsub -N'
meaningless error in clients when reporting_param flush_time is
incorrectly set
remote installation of execd's need enhancement, rework, cleanup
Scheduler does not write pid file in daemonize phase
invalid memory access in cl_com_get_handle
Qconf usage x man page inconsistency
Project man page contains different attribute names.
the binary check in inst_sge is wrong!
Wrong exit code with qconf -sds
qacct -l h=<hostname> dumps core on darwin and linux itanium
inst_sge -m fails for non-root when USER variable is not set
Extraneous space in qsub's "Invalid month specification." message
Hybrid user/project share-tree is broken for user sharing amongst array
jobs
clear usage operation should implicitely trigger refresh in share-tree
dialogue
UNKNOWN group/owner in accouting(5)
qmon's Complex Configuration Load and Save buttons did not work
memory leak in sge_schedd
Bugs fixed in SGE 6.0u10 since release 6.0u9
--------------------------------------------
qdel of just submitted job
can't start qmaster message appears, but qmaster is started
inst_sge script does not add the master host to the shadow_masters file
wrong qlogin_daemon or rlogin_daemon in host conf doesn't set host and
job into error state
reprioritize field in qmon cluster config missing
hostname resolving should not be case sensitive
drmaa_version() function should return 1.0
DRMAA sessions should be persistent
drmaa_control() causes illegal memory access
getDrmaaImplementation() should return the same string as getDrmSystem()
qconf rejects configuration, when attribute value ends with a space
character
during QMaster installation, creation of local database directory fails
on hp11
the execd auto_install takes too long because of long delays after a
parallel install block
Unable to modify Advanced Settings in Configuration for Host in my
cluster using qmon
qacct -l h=<hostname> dump core
shared library name DT_SONAME not set with libdrmaa.so
Berkeley DB backup failed when using hostname with a fully qualified
domain name
the inst_common.sh is missing $SGE_EXECD_PORT
potential security issues in cull library
inst_sge -help wrongly indicates -bup/-rst works with BDB spooling only
qping segfaults in ssl mode
The sge_share_mon utility does not work with the automatic policy enforcement
the auto uninstall execd needs a ssh deamon when the uninstall is done local
accounting records for slave tasks of pe jobs contain invalid submission time
DRMAA 1.0 interface need complete documentation in man pages
qselect matches wrong resources which have been overridden at lower level
qsh does not work if XAUTHORITY is set in root environment
In certain cases jobs may stay in "t" state for 5 minutes
qmaster logging "scheduler tried to remove a incomplete"
ignore_fqdn is broken for the local configuration
host already exists when modifying cluster settings
segmentation fault with empty string
qmaster performance gets throttled if qsub -sync y is used when many
jobs are in the system
sge_conf(5) setting 'max_u_jobs' broken if BDB spooling is used
job control: sorting by different fields
qmon job control: display wider default columns
load formula does not recognize float as weighting factor
qmaster dies when modifying slots value for queue domain when queuename
is missing
spooledit cannot dump USERSET objects
default xterm path in arch_variables script not correct for darwin
architectures
qsh problems on darwin architecture because of wrong crypto lib
execution daemon can crash on Linux where libnss_ldap.so uses BDB 4.2
shared library
Bugs fixed in SGE 6.0u9 since release 6.0u8
-------------------------------------------
CSP mode is affected by OpenSSL Security Advisory [28th September 2006]
account string does not accept the "|" character
unreasonably long scheduler dispatch times if lots of projects are used
in share tree
unreasonably long scheduler dispatch times if lots of cluster queues are
deployed in large clusters
jobs with negative priority will be rejected by qmaster
the owner of all installed files is a non-root user. This change
affects most packages that require special users or groups by making
them use the specified unprivileged user and group instead.
(1) Add two new variables PKG_GROUPS_VARS and PKG_USERS_VARS to
unprivileged.mk. These two variables are lists of other bmake
variables that define package-specific users and groups. Packages
that have user-settable variables for users and groups, e.g. apache
and APACHE_{USER,GROUP}, courier-mta and COURIER_{USER,GROUP},
etc., should list these variables in PKG_USERS_VARS and PKG_GROUPS_VARS
so that unprivileged.mk can know to set them to ${UNPRIVILEGED_USER}
and ${UNPRIVILEGED_GROUP}.
(2) Modify packages to use PKG_GROUPS_VARS and PKG_USERS_VARS.
Update to 1.2.8 (formerly in devel/apr1), no longer build from the
httpd distfile.
devel/rapidsvn:
devel/subversion-base:
parallel/ganglia-monitor-core:
security/hydra:
www/apache2:
Use devel/apr0.
www/apache22:
Use devel/apr and devel/apr-util.
INSTALLATION_DIRS, as well as all occurrences of ${PREFIX}/man with
${PREFIX}/${PKGMANDIR}.
Fixes PR 35265, although I did not use the patch provided therein.
packages with the modular Xorg equivalent. Those are falling back
to the old location by default, so this commmit doesn't change
dependencies.
graphics/xpm ==> x11/libXpm
fonts/Xft2 ==> x11/libXft
x11/Xfixes ==> x11/libXfixes
x11/xcursor ==> x11/libXcursor
x11/Xrender ==> x11/libXrender
x11/Xrandr ==> libXrandr
and add a new helper target and script, "show-buildlink3", that outputs
a listing of the buildlink3.mk files included as well as the depth at
which they are included.
For example, "make show-buildlink3" in fonts/Xft2 displays:
zlib
fontconfig
iconv
zlib
freetype2
expat
freetype2
Xrender
renderproto
variables so that the default INSTALL/DEINSTALL scripts from the
pkginstall framework do the right thing. Where possible, move some
post-install directions for package setup into MESSAGE files so that
they may be re-inspected by querying the installed package using
"pkg_info -D ...".
RECOMMENDED is removed. It becomes ABI_DEPENDS.
BUILDLINK_RECOMMENDED.foo becomes BUILDLINK_ABI_DEPENDS.foo.
BUILDLINK_DEPENDS.foo becomes BUILDLINK_API_DEPENDS.foo.
BUILDLINK_DEPENDS does not change.
IGNORE_RECOMMENDED (which defaulted to "no") becomes USE_ABI_DEPENDS
which defaults to "yes".
Added to obsolete.mk checking for IGNORE_RECOMMENDED.
I did not manually go through and fix any aesthetic tab/spacing issues.
I have tested the above patch on DragonFly building and packaging
subversion and pkglint and their many dependencies.
I have also tested USE_ABI_DEPENDS=no on my NetBSD workstation (where I
have used IGNORE_RECOMMENDED for a long time). I have been an active user
of IGNORE_RECOMMENDED since it was available.
As suggested, I removed the documentation sentences suggesting bumping for
"security" issues.
As discussed on tech-pkg.
I will commit to revbump, pkglint, pkg_install, createbuildlink separately.
Note that if you use wip, it will fail! I will commit to pkgsrc-wip
later (within day).
INSTALL/DEINSTALL script creation within pkgsrc.
If an INSTALL or DEINSTALL script is found in the package directory,
it is automatically used as a template for the pkginstall-generated
scripts. If instead, they should be used simply as the full scripts,
then the package Makefile should set INSTALL_SRC or DEINSTALL_SRC
explicitly, e.g.:
INSTALL_SRC= ${PKGDIR}/INSTALL
DEINSTALL_SRC= # emtpy
As part of the restructuring of the pkginstall framework internals,
we now *always* generate temporary INSTALL or DEINSTALL scripts. By
comparing these temporary scripts with minimal INSTALL/DEINSTALL
scripts formed from only the base templates, we determine whether or
not the INSTALL/DEINSTALL scripts are actually needed by the package
(see the generate-install-scripts target in bsd.pkginstall.mk).
In addition, more variables in the framework have been made private.
The *_EXTRA_TMPL variables have been renamed to *_TEMPLATE, which are
more sensible names given the very few exported variables in this
framework. The only public variables relating to the templates are:
INSTALL_SRC INSTALL_TEMPLATE
DEINSTALL_SRC DEINSTALL_TEMPLATE
HEADER_TEMPLATE
The packages in pkgsrc have been modified to reflect the changes in
the pkginstall framework.
without underscores (REPLACE.*.old, REPLACE.*.new, and REPLACE_FILES.*).
Also convert REPLACE.*.new= ${SH:Q} back to ${SH}, as it should not be quoted
here, if at all.
Ok with rillig.
developer is officially maintaining the package.
The rationale for changing this from "tech-pkg" to "pkgsrc-users" is
that it implies that any user can try to maintain the package (by
submitting patches to the mailing list). Since the folks most likely
to care about the package are the folks that want to use it or are
already using it, this would leverage the energy of users who aren't
developers.
Change pkg to use tk instead of tk83. I found no compelling reason for
it to still be dependant on the old version of tk.
Change the HOMEPAGE, as the old one no longer works.
Tested install, the pkg seems to install stuff properly now.. I believe
the PR should now be addressed fully.
backslashes anymore. A single backslash is enough. Changed the
definition in all affected packages. For those that are not caught, an
additional check is placed into bsd.pkginstall.mk.
file's sole purpose was to provide a dependency on pkg-config and set
some environment variables. Instead, turn pkg-config into a "tool"
in the tools framework, where the pkg-config wrapper automatically
adds PKG_CONFIG_LIBDIR to the environment before invoking the real
pkg-config.
For all package Makefiles that included pkg-config/buildlink3.mk, remove
that inclusion and replace it with USE_TOOLS+=pkg-config.
around at either build-time or at run-time is:
USE_TOOLS+= perl # build-time
USE_TOOLS+= perl:run # run-time
Also remove some places where perl5/buildlink3.mk was being included
by a package Makefile, but all that the package wanted was the Perl
executable.
Several changes are involved since they are all interrelated. These
changes affect about 1000 files.
The first major change is rewriting bsd.builtin.mk as well as all of
the builtin.mk files to follow the new example in bsd.builtin.mk.
The loop to include all of the builtin.mk files needed by the package
is moved from bsd.builtin.mk and into bsd.buildlink3.mk. bsd.builtin.mk
is now included by each of the individual builtin.mk files and provides
some common logic for all of the builtin.mk files. Currently, this
includes the computation for whether the native or pkgsrc version of
the package is preferred. This causes USE_BUILTIN.* to be correctly
set when one builtin.mk file includes another.
The second major change is teach the builtin.mk files to consider
files under ${LOCALBASE} to be from pkgsrc-controlled packages. Most
of the builtin.mk files test for the presence of built-in software by
checking for the existence of certain files, e.g. <pthread.h>, and we
now assume that if that file is under ${LOCALBASE}, then it must be
from pkgsrc. This modification is a nod toward LOCALBASE=/usr. The
exceptions to this new check are the X11 distribution packages, which
are handled specially as noted below.
The third major change is providing builtin.mk and version.mk files
for each of the X11 distribution packages in pkgsrc. The builtin.mk
file can detect whether the native X11 distribution is the same as
the one provided by pkgsrc, and the version.mk file computes the
version of the X11 distribution package, whether it's built-in or not.
The fourth major change is that the buildlink3.mk files for X11 packages
that install parts which are part of X11 distribution packages, e.g.
Xpm, Xcursor, etc., now use imake to query the X11 distribution for
whether the software is already provided by the X11 distribution.
This is more accurate than grepping for a symbol name in the imake
config files. Using imake required sprinkling various builtin-imake.mk
helper files into pkgsrc directories. These files are used as input
to imake since imake can't use stdin for that purpose.
The fifth major change is in how packages note that they use X11.
Instead of setting USE_X11, package Makefiles should now include
x11.buildlink3.mk instead. This causes the X11 package buildlink3
and builtin logic to be executed at the correct place for buildlink3.mk
and builtin.mk files that previously set USE_X11, and fixes packages
that relied on buildlink3.mk files to implicitly note that X11 is
needed. Package buildlink3.mk should also include x11.buildlink3.mk
when linking against the package libraries requires also linking
against the X11 libraries. Where it was obvious, redundant inclusions
of x11.buildlink3.mk have been removed.
is broken anyway.
Remove makedepend tarball and corresponding master site, it does not
exist any longer.
Remove FLAGS=-m486, newer gcc does not like it.
Replace USE_PERL5 with USE_TOOLS+=perl.
Fixes hang bug with ssh under linux, adds a number of speed improvements, now
uses autoconf to build rather than handwritten makefiles. Greatly enhanced
stability.
And always is defined as share/examples/rc.d
which was the default before.
This rc.d scripts are not automatically added to PLISTs now also.
So add to each corresponding PLIST as required.
This was discussed on tech-pkg in late January and late April.
Todo: remove the RCD_SCRIPTS_EXAMPLEDIR uses in MESSAGES and elsewhere
and remove the RCD_SCRIPTS_EXAMPLEDIR itself.
Changes (mostly bugfixes)::
* srclib/libmetrics/freebsd/metrics.c (1.6): Many bug fixes and
cleanups: - Make cpu_state act like get_netbw and get new values
only if called more than 1/2 second from the last value update.
This causing obviously weird results from the CPU metrics on
sparc64 (where the counters seem to be very course) and bogus,
but more subtlety broken results on other architectures. This
has always been broken. - Implement cpu_intr_func (one line!)
- Make the logic for handling bad returns from sysctl make sense.
It should never be triggered in most cases, but at least this
way it won't return bogus values when it happens. - Prefer
sysctlbyname() to sysctl(). It's much easier to read. - Reduce
the use of pointless temporary variables. - Comment/white space
fixes, include more comments of metrics we are unlikely to
actually implement and comments on other rather bogus metrics,
mostly memory related ones.
* lib/libgmond.c (1.17): Set the default time for
tcp_accept_channels to be -1 (blocking io)
* srclib/libmetrics/linux/metrics.c (1.5): Fixed a bug in
pkts_in/out bytes_in/out on for some Linux 2.6.x kernels
http://bugzilla.ganglia.info/cgi-bin/bugzilla/show_bug.cgi?id=21
* gmond/: g25_config.c (1.3), gmond.c (1.102): Patched two bugs in
gmond. The first bug causes gmond to occasionally stop reporting
occasionally when there is a network failure. The second bug in
gmond relates to the host mask being set to 24 instead of 32 when
converting old gmond.conf configuration files.
* srclib/libmetrics/freebsd/metrics.c (1.5): Fix a number of bugs
of varying severity: - makenetvfslist had some nasty uninitilized
variable bugs under FreeBSD 4.x, fix those. - general
reorganization and logic clarity improvements in
makenetvfslist. - Make machine_type_func, os_name_func, and
os_release_func and correct their error handling code to
actually do something useful (not that it should ever be
triggered).
* srclib/libmetrics/freebsd/metrics.c (1.4): - Fix a memory leak in
find_disk_space() as reported by Glen Beane. - Overhaul
makenetvfslist() a bit to fix a leak in low memory situations,
reduce duplicated code, and streamline error handling. - Fix a few
compiler warnings.
Ganglia is a scalable distributed monitoring system for high-performance
computing systems such as clusters and Grids. It is based on a hierarchical
design targeted at federations of clusters. It relies on a multicast-based
listen/announce protocol to monitor state within clusters and uses a tree of
point-to-point connections amongst representative cluster nodes to federate
clusters and aggregate their state. It leverages widely used technologies such
as XML for data representation, XDR for compact, portable data transport, and
RRDtool for data storage and visualization. It uses carefully engineered data
structures and algorithms to achieve very low per-node overheads and high
concurrency. The implementation is robust, has been ported to an extensive set
of operating systems and processor architectures, and is currently in use on
over 500 clusters around the world. It has been used to link clusters across
university campuses and around the world and can scale to handle clusters with
2000 nodes.
http://ganglia.sourceforge.net
framework. The list of changes include:
* Modify compiler.mk so that "c" is always prepended to USE_LANGUAGES,
so we no longer need to say it in package Makefiles. Packages
should now append to USE_LANGUAGES instead of setting it.
* Create mk/compiler/f2c.mk which implements another pseudo-compiler
"f2c" that may be used with any C compiler backend, e.g.
PKGSRC_COMPILER= f2c ccache gcc
* Teach the various "real" compiler files, e.g., sunpro.mk, mipspro.mk,
etc., to use f2c if the native Fortran compiler isn't present.
Packages that use Fortran should now simply include the line:
USE_LANGUAGES+= fortran
in the package Makefile.
under share/examples/rc.d. The variable name already was named
RCD_SCRIPTS_EXAMPLEDIR.
This is from ideas from Greg Woods and others.
Also bumped PKGREVISION for all packages using RCD_SCRIPTS mechanism
(as requested by wiz).
module directory has changed (eg. "darwin-2level" vs.
"darwin-thread-multi-2level").
binary packages of perl modules need to be distinguishable between
being built against threaded perl and unthreaded perl, so bump the
PKGREVISION of all perl module packages and introduce
BUILDLINK_RECOMMENDED for perl as perl>=5.8.5nb5 so the correct
dependencies are registered and the binary packages are distinct.
addresses PR pkg/28619 from H. Todd Fujinaka.
The latest PVM release 3.4.5 is mainly bug fixes and additional
support for varying Linux distributions and some better 64-bit
Linux/AMD support (like it doesn't seg fauly any more :-).
Added strerror() support instead of old crusty global vars.
New architectures were added for AIX5*. Fixed hilarious bug
in pvmtmpnam(), caused crash after 10,000 spawns... :-D
While there's a variety of little tweaks, there are no significant
changes or additions to the basic functionality.
in the process. (More information on tech-pkg.)
Bump PKGREVISION and BUILDLINK_DEPENDS of all packages using libtool and
installing .la files.
Bump PKGREVISION (only) of all packages depending directly on the above
via a buildlink3 include.
All library names listed by *.la files no longer need to be listed
in the PLIST, e.g., instead of:
lib/libfoo.a
lib/libfoo.la
lib/libfoo.so
lib/libfoo.so.0
lib/libfoo.so.0.1
one simply needs:
lib/libfoo.la
and bsd.pkg.mk will automatically ensure that the additional library
names are listed in the installed package +CONTENTS file.
Also make LIBTOOLIZE_PLIST default to "yes".
bulk-builds.
Changes:
- fix a bug in GridResource.init() method.
policyType_ attribute should be instantiated before creating the policy_
object.
- fix a bug in GridSim.gridletSubmit(Gridlet, int).
By default, acknowledgement is false or not required.
- fix a bug in GridResource.processGridletSubmit().
If a Gridlet has finished previously, then a resource needs
to send it back to sender/user.
- renumbered the tags in GridSimTags.
Originally, there are two tags with the same integer value.
which are the full option names used to set rpath directives for the
linker and the compiler, respectively. In places were we are invoking
the linker, use "${LINKER_RPATH_FLAG} <path>", where the space is
inserted in case the flag is a word, e.g. -rpath. The default values
of *_RPATH_FLAG are set by the compiler/*.mk files, depending on the
compiler that you use. They may be overridden on a ${OPSYS}-specific
basis by setting _OPSYS_LINKER_RPATH_FLAG and _OPSYS_COMPILER_RPATH_FLAG,
respectively. Garbage-collect _OPSYS_RPATH_NAME and _COMPILER_LD_FLAG.
for packages such as p5-Parallel-Pvm on some archs since that package
creates a shared object. Should fix lossage seen on amd64 bulk
builds. Both packages which use pvm3 have been tested and still work.
setting the SGE arch. Should fix x86_64 (amd64) as well as others that have
a different MACHINE and MACHINE_ARCH. Noted on the SGE developers list
by Nicolas Joly
The Grid Engine project is an open source community effort
to facilitate the adoption of distributed computing solutions.
Sponsored by Sun Microsystems and hosted by CollabNet, the
Grid Engine project provides enabling distributed resource
management software for wide ranging requirements from compute
farms to grid computing.
Install documentation
undef DEFAULT_HOST so it can run on a box other than lnx4.scri.fsu.edu
Add a MESSAGE giving some hint on what you might to use this
Consider this a drive-by-slight-cleanup-of-breakage
An extensive list of changes since the last release is available online
at http://www-unix.mcs.anl.gov/mpi/mpich/r1_2_5changes.html
(It's difficult to destill the important changes from that file, which
is why I'm not pasting the entire collection in here.)
NetBSD-current, the way mpi-ch test for them and implements them
(using the #pragma weak directive) is currently broken.
Fixes PR pkg/23267 by Kouichirou Hiratsuka.
OpenPBS is a generic network queuing system.
The Portable Batch System (PBS) is a flexible batch queueing and
workload management system originally developed for NASA. It operates
on networked, multi-platform UNIX environments, including
heterogeneous clusters of workstations, supercomputers, and massively
parallel systems. Development of PBS is provided by Altair Grid
Technologies.
of distributed systems:
GridSim is a toolkit for modeling and simulation of parallel and distributed
systems such as Clusters and Grids. GridSim is part of the Gridbus Project
(http://www.gridbus.org/).
The GridSim Toolkit includes:
* GridBroker:
It is Grid scheduler built on top of GridSim for application
scheduling on Grid based on Deadline and Budget Constrained
Scheduling Algorithms.
* GridSim:
It is built on top of SimJava. SimJava is a process based discrete
event simulation package for Java with animation facilities.
* Visual Modeler:
It enables the user to create experiments on different Grid testbeds
and generate the default Grid Broker source codes (in Java).
The Distributed Queueing System is designed as a management tool to
aid in computational resource distribution across a network. DQS
provides architecture transparency for both users and administration
across a heterogeneous environment, allowing for seamless interaction
for multiple architectures. Highly mutable custom site configurations
are possible under DQS. This abilty to customize DQS leads to
effective resource distribution and increased network throughput.
From the packages README:
Version 1.2.5 is primarily a bug-fix release. New features include
Significantly better collective algorithms are available.
Major changes for the globus2 device and the ch_nt device.
Information on previous versions
--------------------------------
Features new or improved in Version 1.2.4 included:
SMP support has been inproved through the use of sched_yield. SMP Systems that
are not oversubscribed can consider configuring with --disable-yeild.
A number of bugs in the ch_p4, ch_p4mpd, and ch_shmem device hae been fixed.
These could cause programs to hang or generate error messages.
Features new or improved in Version 1.2.3 included:
Bug fixes and improved-portability,
particularly in the area of Linux and Mac OS X support. Other new features
include better support for Intel compilers, particularly for IA64 Linux,
better support for various Fortran compilers, and some fixes for rare race
conditions in the ch_p4 device. The documentation on the ch_p4mpd device and
the MPD system has been improved. Version 1.2.3 also includes all of the
source for both the Unix and Windows versions. Windows users should still
use the self-installing version available through the web page. The globus
device now supports version 2 of Globus.
Features new or improved in Version 1.2.2 included:
There are a number of improvements to the ch_p4 device that improve both
performance and reliability, particularly on Linux platforms.
Significant upgrades have been made to the MPD system of daemons that provide
fast startup of MPICH jobs, management of stdio, and a crude parallel debugger
based on gdb.
linda is an simple library implementation of Linda parallel
programming system (http://www.cs.yale.edu/Linda/linda.html). To
write a parallel program with linda, you have to know only 6
functions: spawn, out, in, inp, rd and rdp with linda.
