- Added the 'env' command. This prints a small shell snippet that can be
imported into the current shell. The printed code defines a set of
convenience global variables and functions to work with the source tree.
- Use netbsd-*.gz files to install the kernel rather than kern-*.tgz sets.
This lets sysupgrade work in conjunction with the "releasekernel" target
of build.sh, as this generates the former instead of the latter.
- Preserve timestamps of the kernel when backing it up as onetbsd.
- Fix some tests to not rely on the current state of the system (the cache
directory and the autodetection of KERNEL).
- Add support to specify per-machine targets by prefixing the targets in
BUILD_TARGETS by the machine they correspond to. Useful, for example,
to build different kernels depending on the machine type.
OpenXenManager is a graphical interface to manage XenServer / Xen
Cloud Platform (XCP) hosts through the network. OpenXenManager is
an open-source multiplatform clone of XenCenter (Citrix).
Some functions are not working properly with Citrix XenServer 6.
But boot and suhtdown are works.
version's bug fix really fixes the bug.
All:
- conf/runrmt_android has been enhanced to better deal with the
"runrmt -r file-to-test" calling variant.
- configure now ignores a stramge directory that is created on Mac OS X
when compiling with cc -g and that is in conflict with other rules.
- Introduce a deoendency for man page subdiretories to allow syning
parallel make calls.
- Try to support Debian with FreeBSD kernel
Libschily:
- getperm.c now correctly supports umask +w
- Libschily is now linked against $(LIB_INTL) when a shared libschily
is created.
Libscg:
- struct scsi_inquiry is now using a union around the vendor ID strings in
order to avoid incorrect buffer overflow warnings from GCC-4.x
Libmdigest:
- Cygwin is broken, when using #pragma weak, so we cannot create
weak symbols on Cygwin for sha2.c
Cdrecord:
- struct scsi_inquiry is now using a union around the vendor ID strings in
order to avoid incorrect buffer overflow warnings from GCC-4.x
Readcd:
- Cygwin is broken, when using #pragma weak, so we cannot create
weak symbols on Cygwin for sha2.c
Mkisofs (Maintained/enhanced by J
- While introducing the new option -modification-date, we did accidently remove the
code to set up the creation date in te PVD. Now creation date again contains
correct data.
Originally from wip with netbsd-6 patches sketched by yours truly and
finished/tested by Edgar Fuss.
The Quota module provides access to file system quotas. The
quotactl system call or ioctl is used to query or set quotas
on the local host, or queries are submitted via RPC to a
remote host. Mount tables can be parsed with getmntent and
paths can be translated to device files (or whatever the
actual quotactl implementations needs as argument) of the
according file system.
Apparently depending on the compiler and version of lex used, one could
encounter two lex problems:
1) 'yy_fatal_error' defined but not used
2) 'yy_scan_bytes' compares signed and unsigned (char and int)
This patch resolves the problem.
sysutils/grub2 currently has no support for DragonFly, and there is
significant divergence from FreeBSD. It will require some tested
patches to hostdisk.c and getroot.c at least.
Collection.
The Marionette Collective aka. mcollective is a framework to build server
orchestration or parallel job execution systems.
Mcollective's primary use is to programmatically execute actions on clusters of
servers. In this regard it operates in the same space as tools like Func,
Fabric or Capistrano.
By not relying on central inventories and tools like SSH, it's not simply a
fancy SSH "for loop". MCollective uses modern tools like Publish Subscribe
Middleware and modern philosophies like real time discovery of network
resources using meta data and not hostnames. Delivering a very scalable and
very fast parallel execution environment.
The focus is on catering to the needs of enterprises and large deploys.
Pluggable Authentication, Authorization and Auditing capabilities sets it apart
from other tools in this space.
* Convert to use FIND_PREFIX mechanism instead of LOCALBASE
* Set LICENSE
Changelog:
VERSION 1.3.1 (Aug 31 2008)
------------------------------------------------------------------------------
- Fix help message mixup in lines between -c and -t.
- Add more specific error messages for not currently implemented potential
per backup point options, like cmd_preexec.
- Allow named pipe as logfile - suggested by Patrice Levesque.
- Include rsnapshot-copy script written by Matt McCutchen.
- Allow `backticks` in include_conf.
- Apply fix-sync_first-rollback.diff patch from Matt McCutchen (02 Apr 2008).
- hopefully fix bug with link_dest not being used on second and later backups
when you have link_dest 1 and sync_first 1. (Ignore $sync_dir_was_present)
- Patch from Adam Greenbaum to allow passing of ssh arguments correctly.
- David Grant added rsync_numtries to rsnapshot.conf.
- Applied Ben Low's Linux LVM patch.
- Added stop_on_stale_lockfile, thanks to Henning Moll.
- Michael Ashley's fix for filter/space problems on the rsync command line.
- Remove trailing whitespace from command names in rsnapshot.conf.
- Warn about extra spaces between tab and argument.
- Added multi-line config options, thanks to Dieter Bloms.
- The 'interval' config option is now called 'retain'.
- chdir to avoid an obscure bug in some versions of GNU rm.
- Changed use_lazy_deletes option to use _delete.$$ directories.
- Added note about -H and hard links to docs for rsync_short_args.
- Include rsnapshot-diff.1 in rpm.
- Fix bug with rsnapshot sync the first time (when .sync does not exist)
trying to copy hourly.0 to .sync, even if hourly.0 doesn't exist.
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=413439
VERSION 1.3.0 (Oct 21 2006)
------------------------------------------------------------------------------
- Add files rsnapreport.pl Upgrading_from_1.1 rsnapshot-HOWTO.en.html to rpm.
- fix a bug with removing interval.delete (with use_lazy_deletes).
- rsnapshot detects the global ssh_args option being set multiple times,
and gives a config error suggesting the per-backup ssh_args instead.
- Move Upgrading section of INSTALL to docs/Upgrading_from_1.1.
- Incorporate patch from Alan Batie to fix bugs with include_conf.
- check for quoting on $rsync_long_args rather than just splitting on space.
- Change rsnapshot.conf default to have lockfile enabled.
- Check for stale lock files (PID not currently running).
- explicitly add mode 0644 to sysopen() to create lockfile
- give warning if write() or close() return 0 when writing/closing lockfile
- Make prev_max_interval "not present (yet), nothing to copy" message
when rotating interval.0 more visible (level 3 instead of level 4).
- Add man page for rsnapshot-diff (generated from perldoc).
- Updates to rsnapshot man pages (via perldoc).
- Use =item in AUTHORS section
- move mis-placed =back
- document stale lock file detection
- strongly recommend using a lock file
- add a chmod to example backup script
- change crontab example for weekly from 4 "weekly"s per month to 1 per week
- expand on why higher intervals are done first via cron.
- Change space to tab after #cmd_rsnapshot_diff in rsnapshot.conf.default.in.
- In rpm patch, set cmd_rsnapshot_diff to /usr/bin/rsnapshot-diff.
- Set the test scripts t/*.t.in to run from the source directory
and create directory t/support/files/a if necessary.
- Trim comments in README about gnu cp versions > 5.3 since rsnapshot
strips trailing slashes for gnu cp with rsnapshot 1.2.9 and later.
- Add pointers to HOWTO and utils/rsnapreport.pl in README.
- More examples (eg: timestamp backup_script) in rsnapshot.conf.default.in.
- Change "Perhaps this is not GNU cp?" error message.
changes:
- fixed crash in scsi layer for FreeBSD
- fixed double free when reading mode page 1 fails
- fixed DMA misalignment when reading mode page 1
- changed #includes for glibc for compatibility with 2.32
- dropped Darwin/Mac OS X support
- removed dependency on installed md5sum command
- changed libpng test in configure to utilize pkg-config
- fixed some minor flaws for NetBSD
1) The 4th argument of the sysctl function requires type size_t, not int,
which matters for LP64.
2) DragonFly's kinfo structure changed dramatically a few years ago,
breaking this package. Repatched.
3) sysctl can return pid of -1 for pure kernel processes which must be
skipped to avoid a segfault.
to use 64bit file offsets where available for seek()ing etc. However, the
easy case (sizeof(off_t)>4) is handled incorrectly: mt_size_t is set to
size_t - maybe a copy and paste from the fall-back-to-32bit case.
This type is used at least in init.c, when detecting media size and comparing
to the FAT geometry, consequently failing and erroring out with the message
"Big disks not supported on this architecture."
The patch does handle the (e.g. NetBSD) case of 64bit off_t the same as
the case where a off64_t is available (and the other 64bit off_t-equivalent
cases); namely using off_t as mt_size_t.
Thanks to riastradh@ for pointing out where the bug in llong.h was.
Upstream changes:
-----------------
2012-07-06: released Fabric 1.4.3
2012-07-06: released Fabric 1.3.8
[Bug] #671: reject_unknown_hosts sometimes resulted in a password prompt
instead of an abort. This has been fixed. Thanks to Roy Smith for the report.
[Bug] #659: Update docs to reflect that local currently honors env.path. Thanks
to @floledermann for the catch.
[Bug] #652: Show available commands when aborting on invalid command names.
[Support] #651: Added note about nesting with statements on Python 2.6+. Thanks
to Jens Rantil for the patch.
[Bug] #649: Don?t swallow non-abort-driven exceptions in parallel mode. Fabric
correctly printed such exceptions, and returned them from execute, but did not
actually cause the child or parent processes to halt with a nonzero status.
This has been fixed. execute now also honors env.warn_only so users may still
opt to call it by hand and inspect the returned exceptions, instead of
encountering a hard stop. Thanks to Matt Robenolt for the catch.
[Support] #645: Update Sphinx docs to work well when run out of a source
tarball as opposed to a Git checkout. Thanks again to @Arfrever for the catch.
[Support] #640: (also #644) Update packaging manifest so sdist tarballs include
all necessary test & doc files. Thanks to Mike Gilbert and @Arfrever for catch
& patch.
[Support] #634: Clarified that lcd does no special handling re: the user?s
current working directory, and thus relative paths given to it will be relative
to os.getcwd(). Thanks to @techtonik for the catch.
sysupgrade is a script to automate NetBSD system upgrades. sysupgrade
works by first fetching distribution sets from a specified site or local
directory, then by upgrading the system using such distribution sets and
later by ensuring that the system configuration is up to date. All the
process is controlled by a configuration file, and the defaults should
suit the most common NetBSD upgrades.
sysupgrade can be used to perform upgrades across different system major
and/or minor versions, and it can also be used to track a stable or
development branch from the CVS repository.
sysbuild is the perfect companion to sysupgrade in those cases where you
want to roll your own binaries: both utilities share a very similar
command-line and configuration interface, and the default configuration
files provide examples on how to integrate one with the other.
A few notes about the import:
Right after I submitted sysbuild, I was pointed at etcmanage and its
scripts to build and upgrade NetBSD. I am sending this anyway because
1) it matches sysbuild's behavior closely, 2) it has a detailed manual
page, 3) it has tests... and, well, 4) I had already written most of it
at that time and didn't want to throw it away!
The config and utils modules in this import are a duplicate of the code
in sysbuild, with a few tweaks. This is really bad and the code should
be deduplicated somehow. I'm not sure what the best way of doing so is
and can only think about introducing a common base package with the
shared code (which brings its own problems).
I have tested this to upgrade both -current and 6.0_BETA2 to newer
snapshots, both from local and remote release files.
this is a step of two major branches, to much to list here
pkgsrc notes:
-I've left out our NETBSD_ATOMIC_OPS patch because it is very invasive
and gcc intrinsics are hopefully good enough
-A workaround for non-standard behavior of vsnprintf has been added
upstream, for HPUX as the comment says. We had a patch for __hpux||__sgi
which is removed now - the sgi case should be checked.
Changes in release 3.3.1
* amrecover
o New 'settransalte' command.
* amanda.conf
o new '$s' substitution in autolabel.
o new 'max-dle-by-volume' global option.
o new 'eject-volume' global option.
* diskfile
o new 'includefile' option.
* amcheck
o new '--client-verbose' argument.
* All programs have a new '--version' argument.
* ampgsql can take property from server configuration.
* amgetconf can print one property of a section.
* s3 device
o new S3_SERVER_SIDE_ENCRYPTION property.
o use multiple thread to speedup deleting a volume.
* Many bug fix.
Changes in release 3.3.0
* The default auth is changed to "bsdtcp", if you are using the default bsd
then you must add it to your configuration.
o in amanda.conf
o in amanda-client.conf
o in dumptype/disklist
o in xinetd (if no '-auth' argument to amandad)
* amdump trap crtl-c, it still send the report and do cleanup if you do
one crtl-c, do it more than once to abort the run.
* s3 device
o use multiple threads to speedup the transfer
o can connect to eucalytus.
o new NB_THREADS_BACKUP property
o new NB_THREADS_RECOVERY property
o new S3_HOST property
o new S3_SERVICE_PATH property
o new S3_SUBDOMAIN property
* chg-aggregate: new changer that use other changer sequentially.
* meta-volume
* Add meta label in tapelist file
* chg-disk:
o support for removable disk
o new NUM-SLOT property
o new AUTO-CREATE-SLOT property
o new REMOVABLE property
o new MOUNT property
o new UMOUNT property
o new UMOUNT-LOCKFILE property
o new UMOUNT-IDLE property
* new taperscan algorithm:
o oldest: this algorithm try to run through the volumes in the oldest
order
o lexical: this algorithm try to run through the volumes in the natural
order
* Change in amanda.conf
o new meta-autolabel option
o autolabel can include org, config, barcode, meta in the label
o new client-name option in appication and script.
* application and script in amanda-client.conf can be used to set default
properties for application or script.
* amlabel
o The label argument is no longer required, an autolabel can be
generated
o new --meta option
o new --barcode option
o new --assign option
* amgtar, amstar: The path must be specified, it will not works with a
device.
* amrecover: decompression and decryption are now done on the client if
compression/encryption was done on the client.
* amtape: inventory print the current slot
* amanda.conf:
o autoflush have value "no|yes|all"
o script have single-execution setting.
o Add pre-amcheck, post-amcheck, pre-estimate, post-estimate, pre-backup
and post-backup to execute-on of script.
o Add taperscan and interactivity section.
o add 'server' value in recovery-limit.
o add dump-limit in a dumptype.
* amanda-client.conf
o add amdump-server setting.
* script are searched in $APPLICATION_DIR, $CONFIG_DIR/<conf>/application
and $CONFIG_DIR/application
* amservice
o add -s argument
o is also installed on client
* new amdumpd server service, if enable, it allow client to start a backup
of itself.
* new amdump_client program, it is use on client to start a backup of itself
* implement restore command amzfs-sendrecv, it can be use with amrecover.
Changes in release 3.2.2
* Do not restore the NUL padding bytes, some program fail with them.
* Fix driver doing nothing if taper crash early.
* Fix taperalog *FIT no going to second volume.
* Fix amrestore '-p' not going to next file.
* flush even if flush threshold are not met if it save tape space.
* fix crash in amtrmidx due to memory leak.
* amsamba use 'Use smbclient -TF' for restore.
Changes in release 3.2.1
* barcode are added to the tapelist file.
* Faster 'amadmin find', improve speed on many programs.
* device-output-buffer-size default to a minimum of 4*block_size.
* ssh auth use the client-port
* Bug fixed:
o "Can't opendata output stream: Connection refused".
o Better handling of dump to tape.
o Corrupted 'amdump' log file, amstatus not showing correct state.
o Execute subprocess with the config overwrite.
o tape-device allow to set LEOM.
o Crash in robot changer.
o Script output property are not sent to application.
Changes in release 3.2.0
* Support for multiple simultaneous writes to storage devices
o Can write to all available drives in parallel.
o Works only with the new changer API.
o Useful if two or more volumes are used in a single run.
o dump direct to tape can be scheduled any time during the run,
previously they were done sequencially after all dump to holding disk.
o enabled with the new 'taper-parallel-write' configuration option.
* Support for LEOM in storage devices
o allows splitting without partial parts, saving space
o much more efficient, since parts need not be cached on disk in most
cases
* new, simpler splitting commands in the tapetype section:
part-size, part-cache-type, part-cache-dir, part-cache-max-size; see
amanda.conf(5) for details
* Amanda server configuration file changes (amanda.conf)
o new configuration keyword:
+ taper-parallel-write -- How many drive amanda can write to
in parallel.
o deprecated configuration keywords:
+ amrecover_do_fsf
+ amrecover_check_label
* the CONFIG parameter to amidxtaped is now required; this means that
versions of amrecover older than 2.4.3 will be unable to recover from
servers running Amanda-3.2 and later.
* The new dumptype and global parameters 'recovery-limit' can be used to
limit which hosts may recover from a particular DLE. See amanda.conf(5)
and amanda-auth(7) for more information.
* Several old changers have been removed - these changers will work with
Amanda for the forseeable future, but are no longer included in the
distribution.
o chg-chio
o chg-chs
o chg-iomega
o chg-juke
o chg-mcutil
o chg-mtx
o chg-null (use the new "chg-null:")
o chg-rait (use the new "chg-rait:{dev1,dev2}")
o chg-rth
o chg-scsi-chio
o chg-scsi
* Amdump change:
o new '--no-taper' option to start the run in degraded mode
* Amvault is much improved, but still experimental:
o supports assembling split parts on the source volume and re-splitting
them on the destination
o supports filtering dumps with the same syntax as amfetchdump
o a --fulls-only option skips all incremental dumps on the source
o an --export option tries to move tertiary volumes to import/export
slots when completed
o command-line syntax has changed incompatibly; see manpage or
'amvault --help'
* Rewritten and improved:
o amoverview
o amcheckdump
Changes in release 3.1.0
* Deprecated old changers; see amanda-changers(7) for replacements. These
changers are still available in the distribution, but will be removed in a
future release.
o chg-null
o chg-zd-mtx
o chg-rait
o chg-disk
o chg-multi
* Although chg-zd-mtx is still supported, we recommend that all users upgrade
to the more efficient chg-robot. See contrib/convert-zd-mtx-to-robot.sh
for a useful conversion script.
* Amanda server configuration file changes (amanda.conf)
o deprecated configuration keywords:
+ label_new_tapes
o keywords deprecated in 2.6.1:
+ rawtapedev
+ tapebufs
+ file-pad
o new configuration keyword
+ autolabel -- replace label_new_tapes
+ columnspec -- can specify a precision.
+ order -- in script, to specify script order execution.
+ client_port -- in dumptype, to specify which port to connect on
the client.
+ estimate -- in dumptype, can specify multiple estimate method.
o accept 'define' keyword for defining an holdingdisk
* Amanda client configuration file changes (amanda-client.conf)
o new configuration keyword
+ debug_days -- how many days to keep debug files.
+ client_port -- use by amrecover, specify which port to connect on
the server.
* Removed compile-time default --with-changer-device: specify a device
explicitly in amanda.conf instead
* amtape behavior has changed:
o 'device' subcommand removed
o 'slot advance' subcommand removed
o 'update' subcommand no longer displays each slot as it is updated, and
is not supported by all changers
o taperscan output has changed
o new 'inventory' subcommand
* amrmtape rewrite
o use long option
+ --changer -- Specify which changer to use
+ --cleanup -- Remove logs and indexes associated with label
+ --dryrun -- do not update the original copies.
+ --erase -- Attempt to erase the data contained on the volume
+ --keep-label -- Do not remove label from the tapelist
+ --quiet -- Opposite of --verbose
+ --verbose -- List backups of hosts and disks that are being
discarded.
* amdevcheck
o new --label option.
o new --properties option.
* Device API
o changed wildcard setting for S3_BUCKET_LOCATION from "" to "*"
o new 'ndmp:' device to write to a tape on an ndmp server.
o new 'dvdrw:' device to write to a dvd drive.
* Application API
o new properties to many applications
o amgtar
+ new ACLS, SELINUX and XATTRS properties
o amsamba
+ Allow '\' in diskname and amandapass.
+ new ALLOW-ANONYMOUS property.
o new applications:
+ ampgsql -- Backup PostgreSQL using continuous WAL archiving
+ amsuntar -- Backup filesytem with the SUN tar
+ amraw -- Backup only one directory entry
* New taper, with DirectTCP support, changed tape and catalog format:
o all dumpfiles are now F_SPLIT_DUMPFILE
o all on-tape dumpfiles have numparts=-1, since it's no longer possible
to calculate this value in advance
o there is no logging or reporting of zero-byte, successful parts (this
may cause gaps in filenums in the catalog)
* Redesigned amreport
o much more natural command-line interface (just run 'amreport $config')
o experimental XML output
* configuration override for dumptype works with inheritance.
Changes in release 2.6.1p2
* amtapetype: new -p option
* Bugs fixed
o S3 device driver
o amcheckdump
o file not removed from holding disk
o sendbackup compatibility with a 2.4.2 server
o handle EROFS error from tape device.
o zfs snapshot name us the diskname.
o fix fd allocation in amandad
o crash in amflush
Changes in release 2.6.1p1
* amplot: better output
* Don't include genversion.h in distribution tarballs.
* Bugs fixed
o S3 device driver
o rait device driver
o amstatus
o configure
o application-api
o compilation on some platform
o others small bug
Changes in release 2.6.1
* Amanda server configuration file changes (amanda.conf)
o deprecated configuration keywords:
+ rawtapedev
+ tapebufs
+ file-pad
* Application API: Allow to easily write wrappers around any backup program,
See the 'amanda-applications' man page.
o amgtar: Use GNU tar, it is a lot more configurable than the GNUTAR
program. See 'amgtar' man page.
o amstar: Use star to do a backup, it work only on a partition. See
'amstar' man page.
o amsamba: Use smbclient to backup a cifs share, see amsamba man page.
o amzfs-sendrecv: Do a backup of a ZFS filesystem with 'zfs send'.
* Script API: Allow to run script before and after amanda process, see the
'amanda-scripts' man page.
o amzfs-snapshot: Do a snapshot of a ZFS filesystem, then 'amgtar'
application will backup the snapshot. See 'amzfs-snapshot' man page.
o script-email: Simple script to send email. see 'script-email' man page.
* Changer API v2.0: perl-based changer interface supporting concurrent
use of multiple devices and changers.
o currently operating in "compatibility mode," calling old changer
shell scripts.
o under active development.
* Xfer API: generic library to move and filter data with maximal efficiency
o can read from and write to arbitrary devices, files, etc.
o only used in some applications.
* Amanda archive format: A simple archive format that an application can
use to create backup image.
* 'amarchiver' program to manipulate file in amanda archive format.
* Many improvements to report better error message to user.
* amtape subcommands 'slot prev' and 'slot last' are removed.
* Dozens more perl libraries, with more stable interfaces.
* Many bugs fixed and improvement.
* amgetconf '--client' option to retrieve config from
amanda-client.conf on a client.
* Amanda configuration file changes
o new application-tool section
o new script-tool section
o new device section
o new changer section
Changes in release 2.6.0
* configure --disable-shared doesn't work because perl modules require
shared libraries. Use configure --with-static-binaries to build
statically linked binaries.
* 'amverify' and 'amverifyrun' are deprecated and replaced with the
new, more flexible 'amcheckdump'
* 'amdd' and 'ammt' are deprecated.
* Some Amanda files are now installed in new "amanda/" subdirectories:
libraries are now installed in $libdir/amanda and internal programs
are now installed in $libexecdir/amanda.
* The amandates file, previously at /etc/amandates, is now at
$localstatedir/amanda/amandates. You may want to move your existing
/etc/amandates when you upgrade Amanda.
* New 'amcryptsimple', 'amgpgcrypt' - encryption plugins based on gpg.
* New 'amserverconfig', 'amaddclient' - Initial Amanda configuration tools
these tools make assumptions, please see man page.
* Many bugs fixed and code rewrite/cleanup
* glib is required to compile and run amanda.
* Device API: pluggable interface to storage devices, supporting tapes,
vtapes, RAIT, and Amazon S3
* New perl modules link directly to Amanda, to support writing Amanda
applications in Perl. Perl module are installed by default in the perl
installsitelib directory. It can be changed with
'configure --with-amperldir'.
* New 'local' security driver supports backups of the amanda server
without any network connection or other configuration.
* Almost 200 unit tests are available via 'make installcheck'.
* Amanda configuration file changes
o amanda.conf changes
+ flush-threshold-dumped
+ flush-threshold-scheduled
+ taperflush
+ device_property
+ usetimestamps default to yes
