8 During the 2.1.0 long release candidate cycle, a lot of improvements and
9 changes have accumulated with were released later as 2.1.1.
14 The node evacuate command (``gnt-node evacuate``) was significantly
15 rewritten, and as such the IAllocator protocol was changed - a new
16 request type has been added. This unfortunate change during a stable
17 series is designed to improve performance of node evacuations; on
18 clusters with more than about five nodes and which are well-balanced,
19 evacuation should proceed in parallel for all instances of the node
20 being evacuated. As such, any existing IAllocator scripts need to be
21 updated, otherwise the above command will fail due to the unknown
22 request. The provided "dumb" allocator has not been updated; but the
23 ganeti-htools package supports the new protocol since version 0.2.4.
25 Another important change is increased validation of node and instance
26 names. This might create problems in special cases, if invalid host
29 Also, a new layer of hypervisor parameters has been added, that sits at
30 OS level between the cluster defaults and the instance ones. This allows
31 customisation of virtualization parameters depending on the installed
32 OS. For example instances with OS 'X' may have a different KVM kernel
33 (or any other parameter) than the cluster defaults. This is intended to
34 help managing a multiple OSes on the same cluster, without manual
35 modification of each instance's parameters.
37 A tool for merging clusters, ``cluster-merge``, has been added in the
43 - Improved the int/float conversions that should make the code more
44 robust in face of errors from the node daemons
45 - Fixed the remove node code in case of internal configuration errors
46 - Fixed the node daemon behaviour in face of inconsistent queue
47 directory (e.g. read-only file-system where we can't open the files
49 - Fixed the behaviour of gnt-node modify for master candidate demotion;
50 now it either aborts cleanly or, if given the new “auto_promote”
51 parameter, will automatically promote other nodes as needed
52 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
53 completely prevent Ganeti from working
54 - Fixed bug for instance export when not all disks were successfully
56 - Fixed behaviour of node add when the new node is slow in starting up
58 - Fixed handling of signals in the LUXI client, which should improve
59 behaviour of command-line scripts
60 - Added checks for invalid node/instance names in the configuration (now
61 flagged during cluster verify)
62 - Fixed watcher behaviour for disk activation errors
63 - Fixed two potentially endless loops in http library, which led to the
64 RAPI daemon hanging and consuming 100% CPU in some cases
65 - Fixed bug in RAPI daemon related to hashed passwords
66 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
68 - Enhanced compatibility with non-Debian OSes, but not using absolute
69 path in some commands and allowing customisation of the ssh
70 configuration directory
71 - Fixed possible future issue with new Python versions by abiding to the
72 proper use of ``__slots__`` attribute on classes
73 - Added checks that should prevent directory traversal attacks
74 - Many documentation fixes based on feedback from users
79 - Added an “early_release” more for instance replace disks and node
80 evacuate, where we release locks earlier and thus allow higher
81 parallelism within the cluster
82 - Added watcher hooks, intended to allow the watcher to restart other
83 daemons (e.g. from the ganeti-nbma project), but they can be used of
84 course for any other purpose
85 - Added a compile-time disable for DRBD barriers, to increase
86 performance if the administrator trusts the power supply or the
87 storage system to not lose writes
88 - Added the option of using syslog for logging instead of, or in
89 addition to, Ganeti's own log files
90 - Removed boot restriction for paravirtual NICs for KVM, recent versions
91 can indeed boot from a paravirtual NIC
92 - Added a generic debug level for many operations; while this is not
93 used widely yet, it allows one to pass the debug value all the way to
95 - Enhanced the hooks environment for instance moves (failovers,
96 migrations) where the primary/secondary nodes changed during the
97 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
98 - Enhanced data validations for many user-supplied values; one important
99 item is the restrictions imposed on instance and node names, which
100 might reject some (invalid) host names
101 - Add a configure-time option to disable file-based storage, if it's not
102 needed; this allows greater security separation between the master
103 node and the other nodes from the point of view of the inter-node RPC
105 - Added user notification in interactive tools if job is waiting in the
106 job queue or trying to acquire locks
107 - Added log messages when a job is waiting for locks
108 - Added filtering by node tags in instance operations which admit
109 multiple instances (start, stop, reboot, reinstall)
110 - Added a new tool for cluster mergers, ``cluster-merge``
111 - Parameters from command line which are of the form ``a=b,c=d`` can now
112 use backslash escapes to pass in values which contain commas,
113 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
115 - For KVM, the instance name is the first parameter passed to KVM, so
116 that it's more visible in the process list
122 Ganeti 2.1 brings many improvements with it. Major changes:
124 - Added infrastructure to ease automated disk repairs
125 - Added new daemon to export configuration data in a cheaper way than
127 - Instance NICs can now be routed instead of being associated with a
129 - Improved job locking logic to reduce impact of jobs acquiring multiple
130 locks waiting for other long-running jobs
132 In-depth implementation details can be found in the Ganeti 2.1 design
138 - Added chroot hypervisor
139 - Added more options to xen-hvm hypervisor (``kernel_path`` and
141 - Added more options to xen-pvm hypervisor (``use_bootloader``,
142 ``bootloader_path`` and ``bootloader_args``)
143 - Added the ``use_localtime`` option for the xen-hvm and kvm
144 hypervisors, and the default value for this has changed to false (in
145 2.0 xen-hvm always enabled it)
146 - Added luxi call to submit multiple jobs in one go
147 - Added cluster initialization option to not modify ``/etc/hosts``
149 - Added network interface parameters
150 - Added dry run mode to some LUs
151 - Added RAPI resources:
153 - ``/2/instances/[instance_name]/info``
154 - ``/2/instances/[instance_name]/replace-disks``
155 - ``/2/nodes/[node_name]/evacuate``
156 - ``/2/nodes/[node_name]/migrate``
157 - ``/2/nodes/[node_name]/role``
158 - ``/2/nodes/[node_name]/storage``
159 - ``/2/nodes/[node_name]/storage/modify``
160 - ``/2/nodes/[node_name]/storage/repair``
162 - Added OpCodes to evacuate or migrate all instances on a node
163 - Added new command to list storage elements on nodes (``gnt-node
164 list-storage``) and modify them (``gnt-node modify-storage``)
165 - Added new ssconf files with master candidate IP address
166 (``ssconf_master_candidates_ips``), node primary IP address
167 (``ssconf_node_primary_ips``) and node secondary IP address
168 (``ssconf_node_secondary_ips``)
169 - Added ``ganeti-confd`` and a client library to query the Ganeti
170 configuration via UDP
171 - Added ability to run hooks after cluster initialization and before
173 - Added automatic mode for disk replace (``gnt-instance replace-disks
175 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
176 after catastrophic data-loss
177 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
179 - Added ``gnt-instance move`` command to move instances
180 - Added ``gnt-cluster watcher`` command to control watcher
181 - Added ``gnt-node powercycle`` command to powercycle nodes
182 - Added new job status field ``lock_status``
183 - Added parseable error codes to cluster verification (``gnt-cluster
184 verify --error-codes``) and made output less verbose (use
185 ``--verbose`` to restore previous behaviour)
186 - Added UUIDs to the main config entities (cluster, nodes, instances)
187 - Added support for OS variants
188 - Added support for hashed passwords in the Ganeti remote API users file
190 - Added option to specify maximum timeout on instance shutdown
191 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
192 - Added new helper script to start and stop Ganeti daemons
193 (``daemon-util``), with the intent to reduce the work necessary to
194 adjust Ganeti for non-Debian distributions and to start/stop daemons
196 - Added more unittests
197 - Fixed critical bug in ganeti-masterd startup
198 - Removed the configure-time ``kvm-migration-port`` parameter, this is
199 now customisable at the cluster level for both the KVM and Xen
200 hypervisors using the new ``migration_port`` parameter
201 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
202 reinstalling an instance
203 - Allowed ``@`` in tag names
204 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
205 - Many documentation updates
206 - Distribute hypervisor files on ``gnt-cluster redist-conf``
207 - ``gnt-instance reinstall`` can now reinstall multiple instances
208 - Updated many command line parameters
209 - Introduced new OS API version 15
210 - No longer support a default hypervisor
211 - Treat virtual LVs as inexistent
212 - Improved job locking logic to reduce lock contention
213 - Match instance and node names case insensitively
214 - Reimplemented bash completion script to be more complete
221 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
222 - Fix a string formatting bug
223 - Improve safety of the code in some error paths
224 - Improve data validation in the master of values returned from nodes
230 - Fix security issue due to missing validation of iallocator names; this
231 allows local and remote execution of arbitrary executables
232 - Fix failure of gnt-node list during instance removal
233 - Ship the RAPI documentation in the archive
239 - Fixed many wrong messages
240 - Fixed a few bugs related to the locking library
241 - Fixed MAC checking at instance creation time
242 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
243 - Fixed a few issues related to signal handling in both daemons and
245 - Fixed the example startup script provided
246 - Fixed insserv dependencies in the example startup script (patch from
248 - Fixed handling of drained nodes in the iallocator framework
249 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
251 - Fixed error related to invalid job IDs in job polling
252 - Fixed job/opcode persistence on unclean master shutdown
253 - Fixed handling of partial job processing after unclean master
255 - Fixed error reporting from LUs, previously all errors were converted
256 into execution errors
257 - Fixed error reporting from burnin
258 - Decreased significantly the memory usage of the job queue
259 - Optimised slightly multi-job submission
260 - Optimised slightly opcode loading
261 - Backported the multi-job submit framework from the development
262 branch; multi-instance start and stop should be faster
263 - Added script to clean archived jobs after 21 days; this will reduce
264 the size of the queue directory
265 - Added some extra checks in disk size tracking
266 - Added an example ethers hook script
267 - Added a cluster parameter that prevents Ganeti from modifying of
269 - Added more node information to RAPI responses
270 - Added a ``gnt-job watch`` command that allows following the ouput of a
272 - Added a bind-address option to ganeti-rapi
273 - Added more checks to the configuration verify
274 - Enhanced the burnin script such that some operations can be retried
276 - Converted instance reinstall to multi-instance model
282 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
283 to allow using the pre-2.0.2 behaviour in activation, if any existing
284 instances have mismatched disk sizes in the configuration
285 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
286 any configuration mismatches for disk sizes
287 - Added ``gnt-master cluste-failover --no-voting`` to allow master
288 failover to work on two-node clusters
289 - Fixed the ``--net`` option of ``gnt-backup import``, which was
291 - Fixed detection of OS script errors in ``gnt-backup export``
292 - Fixed exit code of ``gnt-backup export``
298 - Added experimental support for stripped logical volumes; this should
299 enhance performance but comes with a higher complexity in the block
300 device handling; stripping is only enabled when passing
301 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
302 affected even in the non-stripped mode
303 - Improved resiliency against transient failures at the end of DRBD
304 resyncs, and in general of DRBD resync checks
305 - Fixed a couple of issues with exports and snapshot errors
306 - Fixed a couple of issues in instance listing
307 - Added display of the disk size in ``gnt-instance info``
308 - Fixed checking for valid OSes in instance creation
309 - Fixed handling of the "vcpus" parameter in instance listing and in
310 general of invalid parameters
311 - Fixed http server library, and thus RAPI, to handle invalid
312 username/password combinations correctly; this means that now they
313 report unauthorized for queries too, not only for modifications,
314 allowing earlier detect of configuration problems
315 - Added a new "role" node list field, equivalent to the master/master
316 candidate/drained/offline flags combinations
317 - Fixed cluster modify and changes of candidate pool size
318 - Fixed cluster verify error messages for wrong files on regular nodes
319 - Fixed a couple of issues with node demotion from master candidate role
320 - Fixed node readd issues
321 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
322 - Added a new ``--no-voting`` option for masterfailover to fix failover
323 on two-nodes clusters when the former master node is unreachable
324 - Added instance reinstall over RAPI
330 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
331 allow re-adding the start in single-user option (regression from 1.2)
332 - the watcher writes the instance status to a file, to allow monitoring
333 to report the instance status (from the master) based on cached
334 results of the watcher's queries; while this can get stale if the
335 watcher is being locked due to other work on the cluster, this is
337 - the watcher now also restarts the node daemon and the rapi daemon if
339 - fixed the watcher to handle full and drained queue cases
340 - hooks export more instance data in the environment, which helps if
341 hook scripts need to take action based on the instance's properties
342 (no longer need to query back into ganeti)
343 - instance failovers when the instance is stopped do not check for free
344 RAM, so that failing over a stopped instance is possible in low memory
346 - rapi uses queries for tags instead of jobs (for less job traffic), and
347 for cluster tags it won't talk to masterd at all but read them from
349 - a couple of error handling fixes in RAPI
350 - drbd handling: improved the error handling of inconsistent disks after
351 resync to reduce the frequency of "there are some degraded disks for
352 this instance" messages
353 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
354 error handling path called a wrong function name)
360 - no changes from rc5
363 Version 2.0 release candidate 5
364 -------------------------------
366 - fix a couple of bugs (validation, argument checks)
367 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
368 - some small improvements to RAPI and IAllocator
369 - make watcher automatically start the master daemon if down
372 Version 2.0 release candidate 4
373 -------------------------------
375 - change the OS list to not require locks; this helps with big clusters
376 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
377 volume group is broken
378 - ``gnt-instance info``, without any arguments, doesn't run for all
379 instances anymore; either pass ``--all`` or pass the desired
380 instances; this helps against mistakes on big clusters where listing
381 the information for all instances takes a long time
382 - miscellaneous doc and man pages fixes
385 Version 2.0 release candidate 3
386 -------------------------------
388 - Change the internal locking model of some ``gnt-node`` commands, in
389 order to reduce contention (and blocking of master daemon) when
390 batching many creation/reinstall jobs
391 - Fixes to Xen soft reboot
392 - No longer build documentation at build time, instead distribute it in
393 the archive, in order to reduce the need for the whole docbook/rst
397 Version 2.0 release candidate 2
398 -------------------------------
400 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
401 - Fix watcher startup sequence, improves the behaviour of busy clusters
402 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
403 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
404 bind address and other places
405 - Some documentation fixes and updates
408 Version 2.0 release candidate 1
409 -------------------------------
411 - More documentation updates, now all docs should be more-or-less
413 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
415 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
417 - Fix an issue related to $libdir/run/ganeti and cluster creation
423 - Xen PVM and KVM have switched the default value for the instance root
424 disk to the first partition on the first drive, instead of the whole
425 drive; this means that the OS installation scripts must be changed
427 - Man pages have been updated
428 - RAPI has been switched by default to HTTPS, and the exported functions
429 should all work correctly
430 - RAPI v1 has been removed
431 - Many improvements to the KVM hypervisor
432 - Block device errors are now better reported
433 - Many other bugfixes and small improvements
439 - Version 2 is a general rewrite of the code and therefore the
440 differences are too many to list, see the design document for 2.0 in
441 the ``doc/`` subdirectory for more details
442 - In this beta version there is not yet a migration path from 1.2 (there
443 will be one in the final 2.0 release)
444 - A few significant changes are:
446 - all commands are executed by a daemon (``ganeti-masterd``) and the
447 various ``gnt-*`` commands are just front-ends to it
448 - all the commands are entered into, and executed from a job queue,
449 see the ``gnt-job(8)`` manpage
450 - the RAPI daemon supports read-write operations, secured by basic
451 HTTP authentication on top of HTTPS
452 - DRBD version 0.7 support has been removed, DRBD 8 is the only
453 supported version (when migrating from Ganeti 1.2 to 2.0, you need
454 to migrate to DRBD 8 first while still running Ganeti 1.2)
455 - DRBD devices are using statically allocated minor numbers, which
456 will be assigned to existing instances during the migration process
457 - there is support for both Xen PVM and Xen HVM instances running on
459 - KVM virtualization is supported too
460 - file-based storage has been implemented, which means that it is
461 possible to run the cluster without LVM and DRBD storage, for
462 example using a shared filesystem exported from shared storage (and
463 still have live migration)
469 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
470 - Reuse the old instance mac address by default on instance import, if
471 the instance name is the same.
472 - Handle situations in which the node info rpc returns incomplete
474 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
475 - Improved version of batcher:
478 - instance mac address support
479 - support for HVM clusters/instances
481 - Add an option to show the number of cpu sockets and nodes in
483 - Support OSes that handle more than one version of the OS api (but do
484 not change the current API in any other way)
485 - Fix ``gnt-node migrate``
486 - ``gnt-debug`` man page
487 - Fixes various more typos and small issues
488 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
494 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
495 type of disk exported to fully virtualized instances.
496 - provide access to the serial console of HVM instances
497 - instance auto_balance flag, set by default. If turned off it will
498 avoid warnings on cluster verify if there is not enough memory to fail
499 over an instance. in the future it will prevent automatically failing
500 it over when we will support that.
501 - batcher tool for instance creation, see ``tools/README.batcher``
502 - ``gnt-instance reinstall --select-os`` to interactively select a new
503 operating system when reinstalling an instance.
504 - when changing the memory amount on instance modify a check has been
505 added that the instance will be able to start. also warnings are
506 emitted if the instance will not be able to fail over, if auto_balance
508 - documentation fixes
509 - sync fields between ``gnt-instance list/modify/add/import``
510 - fix a race condition in drbd when the sync speed was set after giving
511 the device a remote peer.
517 - note: the allowed size and number of tags per object were reduced
518 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
519 - fixed twisted 8.x compatibility
520 - fixed ``gnt-instance replace-disks`` with iallocator
521 - add TCP keepalives on twisted connections to detect restarted nodes
522 - disk increase support, see ``gnt-instance grow-disk``
523 - implement bulk node/instance query for RAPI
524 - add tags in node/instance listing (optional)
525 - experimental migration (and live migration) support, read the man page
526 for ``gnt-instance migrate``
527 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
528 has some small improvements in handling its state file
534 - Experimental readonly, REST-based remote API implementation;
535 automatically started on master node, TCP port 5080, if enabled by
536 ``--enable-rapi`` parameter to configure script.
537 - Instance allocator support. Add and import instance accept a
538 ``--iallocator`` parameter, and call that instance allocator to decide
539 which node to use for the instance. The iallocator document describes
540 what's expected from an allocator script.
541 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
542 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
543 node is lost there is still enough memory to fail over the instances
545 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
546 ``gnt-cluster verify``, to check for site-specific compliance. All the
547 hooks will run, and their output, if any, will be displayed. Any
548 failing hook will make the verification return an error value.
549 - ``gnt-cluster verify`` now checks that its peers are reachable on the
550 primary and secondary interfaces
551 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
552 that is still declared as part of the cluster and has failed.
553 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
554 specifying output fields, that just adds the chosen fields to the
556 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
557 export from the filesystem.
558 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
559 have been added. Using them you can enable/disable acpi and pae
560 support, and specify a path for a cd image to be exported to the
561 instance. These parameters as the name suggest only work on HVM
563 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
564 PAE support will be set to the previously hardcoded values, but the
565 (previously hardcoded) path to the CDROM ISO image will be unset and
566 if required, needs to be set manually with ``gnt-instance modify``
568 - The address to which an instance's VNC console is bound is now
569 selectable per-instance, rather than being cluster wide. Of course
570 this only applies to instances controlled via VNC, so currently just
571 applies to HVM clusters.
577 - more tweaks to the disk activation code (especially helpful for DRBD)
578 - change the default ``gnt-instance list`` output format, now there is
579 one combined status field (see the manpage for the exact values this
581 - some more fixes for the mac export to hooks change
582 - make Ganeti not break with DRBD 8.2.x (which changed the version
583 format in ``/proc/drbd``) (issue 24)
584 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
585 template, allowing migration from DRBD0.7+MD to DRBD8
591 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
593 - add command aliases infrastructure and a few aliases
594 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
595 man pages and the ``--help`` option of ``gnt-node
596 list``/``gnt-instance list``
597 - fix ``gnt-backup list`` with down nodes (issue 21)
598 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
599 - fix the dist archive and add a check for including svn/git files in
601 - some developer-related changes: improve the burnin and the QA suite,
602 add an upload script for testing during development
608 - experimental HVM support, read the install document, section
609 "Initializing the cluster"
610 - allow for the PVM hypervisor per-instance kernel and initrd paths
611 - add a new command ``gnt-cluster verify-disks`` which uses a new
612 algorithm to improve the reconnection of the DRBD pairs if the device
613 on the secondary node has gone away
614 - make logical volume code auto-activate LVs at disk activation time
615 - slightly improve the speed of activating disks
616 - allow specification of the MAC address at instance creation time, and
617 changing it later via ``gnt-instance modify``
618 - fix handling of external commands that generate lots of output on
620 - update documentation with regard to minimum version of DRBD8 supported
626 - Log the ``xm create`` output to the node daemon log on failure (to
627 help diagnosing the error)
628 - In debug mode, log all external commands output if failed to the logs
629 - Change parsing of lvm commands to ignore stderr
635 - Another round of updates to the DRBD 8 code to deal with more failures
636 in the replace secondary node operation
637 - Some more logging of failures in disk operations (lvm, drbd)
638 - A few documentation updates
645 - Change configuration file format from Python's Pickle to JSON.
646 Upgrading is possible using the cfgupgrade utility.
647 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
648 faster replace disks and is more stable (DRBD 8 has many improvements
649 compared to DRBD 0.7)
650 - Added command line tags support (see man pages for ``gnt-instance``,
651 ``gnt-node``, ``gnt-cluster``)
652 - Added instance rename support
653 - Added multi-instance startup/shutdown
654 - Added cluster rename support
655 - Added ``gnt-node evacuate`` to simplify some node operations
656 - Added instance reboot operation that can speedup reboot as compared to
658 - Soften the requirement that hostnames are in FQDN format
659 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
661 - Removed dependency on debian's patched fping that uses the
662 non-standard ``-S`` option
663 - Now the OS definitions are searched for in multiple, configurable
664 paths (easier for distros to package)
665 - Some changes to the hooks infrastructure (especially the new
666 post-configuration update hook)
667 - Other small bugfixes
669 .. vim: set textwidth=72 :