7 *(Released Fri, 4 Feb 2011)*
9 Many changes and fixes since the beta1 release. While there were some
10 internal changes, the code has been mostly stabilised for the RC
13 Note: the dumb allocator was removed in this release, as it was not kept
14 up-to-date with the IAllocator protocol changes. It is recommended to
15 use the ``hail`` command from the ganeti-htools package.
17 Note: the 2.4 and up versions of Ganeti are not compatible with the
18 0.2.x branch of ganeti-htools. You need to upgrade to
19 ganeti-htools-0.3.0 (or later).
21 Regressions fixed from 2.3
22 ~~~~~~~~~~~~~~~~~~~~~~~~~~
24 - Fixed the ``gnt-cluster verify-disks`` command
25 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
27 - Fixed disk adoption breakage
28 - Fixed wrong headers in instance listing for field aliases
33 - Fixed corner case in KVM handling of NICs
34 - Fixed many cases of wrong handling of non-vm_capable nodes
35 - Fixed a bug where a missing instance symlink was not possible to
36 recreate with any ``gnt-*`` command (now ``gnt-instance
37 activate-disks`` does it)
38 - Fixed the volume group name as reported by ``gnt-cluster
40 - Increased timeouts for the import-export code, hopefully leading to
41 fewer aborts due network or instance timeouts
42 - Fixed bug in ``gnt-node list-storage``
43 - Fixed bug where not all daemons were started on cluster
44 initialisation, but only at the first watcher run
45 - Fixed many bugs in the OOB implementation
46 - Fixed watcher behaviour in presence of instances with offline
48 - Fixed instance list output for instances running on the wrong node
49 - a few fixes to the cluster-merge tool, but it still cannot merge
50 multi-node groups (currently it is not recommended to use this tool)
56 - Improved network configuration for the KVM hypervisor
57 - Added e1000 as a supported NIC for Xen-HVM
58 - Improved the lvmstrap tool to also be able to use partitions, as
60 - Improved speed of disk wiping (the cluster parameter
61 ``prealloc_wipe_disks``, so that it has a low impact on the total time
63 - Added documentation for the OS parameters
64 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
65 hypervisor is not responding
66 - Added display of blacklisted and hidden OS information in
68 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
69 NIC and node parameters, which might create problems with currently
70 invalid (but undetected) configuration files, but prevents validation
71 failures when unrelated parameters are modified
72 - Changed cluster initialisation to wait for the master daemon to become
74 - Expanded the RAPI interface:
76 - Added config redistribution resource
77 - Added activation/deactivation of instance disks
78 - Added export of console information
80 - Implemented log file reopening on SIGHUP, which allows using
81 logrotate(8) for the Ganeti log files
82 - Added a basic OOB helper script as an example
88 *(Released Fri, 14 Jan 2011)*
93 - Fixed timezone issues when formatting timestamps
94 - Added support for node groups, available via ``gnt-group`` and other
96 - Added out-of-band framework and management, see :doc:`design
97 document <design-oob>`
98 - Removed support for roman numbers from ``gnt-node list`` and
99 ``gnt-instance list``.
100 - Allowed modification of master network interface via ``gnt-cluster
101 modify --master-netdev``
102 - Accept offline secondaries while shutting down instance disks
103 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
104 - Added support for multiple LVM volume groups
105 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
107 - Added commands to list available fields:
109 - ``gnt-node list-fields``
110 - ``gnt-group list-fields``
111 - ``gnt-instance list-fields``
113 - Updated documentation and man pages
118 - Moved ``rapi_users`` file into separate directory, now named
119 ``…/ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
121 - Added new tool for running commands on many machines,
122 ``tools/ganeti-listrunner``
123 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
124 improving the ``gnt-instance console`` output
125 - Allowed customisation of disk index separator at ``configure`` time
126 - Export node group allocation policy to :doc:`iallocator <iallocator>`
127 - Added support for non-partitioned md disks in ``lvmstrap``
128 - Added script to gracefully power off KVM instances
129 - Split ``utils`` module into smaller parts
130 - Changed query operations to return more detailed information, e.g.
131 whether an information is unavailable due to an offline node. To use
132 this new functionality, the LUXI call ``Query`` must be used. Field
133 information is now stored by the master daemon and can be retrieved
134 using ``QueryFields``. Instances, nodes and groups can also be queried
135 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
136 exposed via RAPI). The following commands make use of this
137 infrastructure change:
140 - ``gnt-group list-fields``
142 - ``gnt-node list-fields``
143 - ``gnt-instance list``
144 - ``gnt-instance list-fields``
145 - ``gnt-debug locks``
150 - New RAPI resources (see :doc:`rapi`):
154 - ``/2/groups/[group_name]``
155 - ``/2/groups/[group_name]/assign-nodes``
156 - ``/2/groups/[group_name]/modify``
157 - ``/2/groups/[group_name]/rename``
158 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
162 - Implemented ``no_install`` for instance creation
163 - Implemented OS parameters for instance reinstallation, allowing
164 use of special settings on reinstallation (e.g. for preserving data)
169 - Added IPv6 support in import/export
170 - Pause DRBD synchronization while wiping disks on instance creation
171 - Updated unittests and QA scripts
172 - Improved network parameters passed to KVM
173 - Converted man pages from docbook to reStructuredText
179 *(Released Mon, 20 Dec 2010)*
181 Released version 2.3.1~rc1 without any changes.
187 *(Released Wed, 1 Dec 2010)*
189 - impexpd: Disable OpenSSL compression in socat if possible (backport
190 from master, commit e90739d625b, see :doc:`installation guide
191 <install-quick>` for details)
192 - Changed unittest coverage report to exclude test scripts
193 - Added script to check version format
199 *(Released Wed, 1 Dec 2010)*
201 Released version 2.3.0~rc1 without any changes.
207 *(Released Fri, 19 Nov 2010)*
209 A number of bugfixes and documentation updates:
211 - Update ganeti-os-interface documentation
212 - Fixed a bug related to duplicate MACs or similar items which should be
214 - Fix breakage in OS state modify
215 - Reinstall instance: disallow offline secondaries (fixes bug related to
216 OS changing but reinstall failing)
217 - plus all the other fixes between 2.2.1 and 2.2.2
223 *(Released Tue, 2 Nov 2010)*
225 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
226 - Fixed master failover race with watcher
227 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
229 - Accept previously stopped instance for export with instance removal
230 - Simplify and extend the environment variables for instance OS scripts
231 - Added new node flags, ``master_capable`` and ``vm_capable``
232 - Added optional instance disk wiping prior during allocation. This is a
233 cluster-wide option and can be set/modified using
234 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
235 - Added IPv6 support, see :doc:`design document <design-2.3>` and
237 - Added a new watcher option (``--ignore-pause``)
238 - Added option to ignore offline node on instance start/stop
239 (``--ignore-offline``)
240 - Allow overriding OS parameters with ``gnt-instance reinstall``
241 - Added ability to change node's secondary IP address using ``gnt-node
243 - Implemented privilege separation for all daemons except
244 ``ganeti-noded``, see ``configure`` options
245 - Complain if an instance's disk is marked faulty in ``gnt-cluster
247 - Implemented job priorities (see ``ganeti(7)`` manpage)
248 - Ignore failures while shutting down instances during failover from
250 - Exit daemon's bootstrap process only once daemon is ready
251 - Export more information via ``LUInstanceQuery``/remote API
252 - Improved documentation, QA and unittests
253 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
254 restart if the file was created or changed
255 - Added LUXI protocol version sent with each request and response,
256 allowing detection of server/client mismatches
257 - Moved the Python scripts among gnt-* and ganeti-* into modules
258 - Moved all code related to setting up SSH to an external script,
260 - Infrastructure changes for node group support in future versions
266 *(Released Fri, 19 Nov 2010)*
268 A few small bugs fixed, and some improvements to the build system:
270 - Fix documentation regarding conversion to drbd
271 - Fix validation of parameters in cluster modify (``gnt-cluster modify
273 - Fix error handling in node modify with multiple changes
274 - Allow remote imports without checked names
280 *(Released Tue, 19 Oct 2010)*
282 - Disable SSL session ID cache in RPC client
288 *(Released Thu, 14 Oct 2010)*
290 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
291 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
293 - Fix problems with interaction between Curl and Python's HTTP server,
294 resulting in increased speed in many RPC calls
295 - Improve our release script to prevent breakage with older aclocal and
302 *(Released Thu, 7 Oct 2010)*
304 - Fixed issue 125, replace hardcoded “xenvg” in ``gnt-cluster`` with
305 value retrieved from master
306 - Added support for blacklisted or hidden OS definitions
307 - Added simple lock monitor (accessible via (``gnt-debug locks``)
308 - Added support for -mem-path in KVM hypervisor abstraction layer
309 - Allow overriding instance parameters in tool for inter-cluster
310 instance moves (``tools/move-instance``)
311 - Improved opcode summaries (e.g. in ``gnt-job list``)
312 - Improve consistency of OS listing by sorting it
313 - Documentation updates
319 *(Released Fri, 8 Oct 2010)*
321 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
327 *(Released Mon, 4 Oct 2010)*
329 - Fixed regression in ``gnt-instance rename``
335 *(Released Wed, 22 Sep 2010)*
337 - Fixed OS_VARIANT variable for OS scripts
338 - Fixed cluster tag operations via RAPI
339 - Made ``setup-ssh`` exit with non-zero code if an error occurred
340 - Disabled RAPI CA checks in watcher
346 *(Released Mon, 23 Aug 2010)*
348 - Support DRBD versions of the format "a.b.c.d"
350 - Re-introduce support for usage from multiple threads in RAPI client
351 - Instance renames and modify via RAPI
352 - Work around race condition between processing and archival in job
354 - Mark opcodes following failed one as failed, too
355 - Job field ``lock_status`` was removed due to difficulties making it
356 work with the changed job queue in Ganeti 2.2; a better way to monitor
357 locks is expected for a later 2.2.x release
358 - Fixed dry-run behaviour with many commands
359 - Support ``ssh-agent`` again when adding nodes
360 - Many additional bugfixes
366 *(Released Fri, 30 Jul 2010)*
368 Important change: the internal RPC mechanism between Ganeti nodes has
369 changed from using a home-grown http library (based on the Python base
370 libraries) to use the PycURL library. This requires that PycURL is
371 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
372 against GnuTLS by default. cURL's support for GnuTLS had known issues
373 before cURL 7.21.0 and we recommend using the latest cURL release or
374 linking against OpenSSL. Most other distributions already link PycURL
375 and cURL against OpenSSL. The command::
377 python -c 'import pycurl; print pycurl.version'
379 can be used to determine the libraries PycURL and cURL are linked
382 Other significant changes:
384 - Rewrote much of the internals of the job queue, in order to achieve
385 better parallelism; this decouples job query operations from the job
386 processing, and it should allow much nicer behaviour of the master
387 daemon under load, and it also has uncovered some long-standing bugs
388 related to the job serialisation (now fixed)
389 - Added a default iallocator setting to the cluster parameters,
390 eliminating the need to always pass nodes or an iallocator for
391 operations that require selection of new node(s)
392 - Added experimental support for the LXC virtualization method
393 - Added support for OS parameters, which allows the installation of
394 instances to pass parameter to OS scripts in order to customise the
396 - Added a hypervisor parameter controlling the migration type (live or
397 non-live), since hypervisors have various levels of reliability; this
398 has renamed the 'live' parameter to 'mode'
399 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
400 logical volumes, meaning that cluster verify will ignore them and not
401 flag their presence as errors
402 - The watcher will now reset the error count for failed instances after
403 8 hours, thus allowing self-healing if the problem that caused the
404 instances to be down/fail to start has cleared in the meantime
405 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
406 check for, and warn, if the drbd module parameter ``usermode_helper``
407 is not consistent with the cluster-wide setting; this is needed to
408 make diagnose easier of failed drbd creations
409 - Started adding base IPv6 support, but this is not yet
410 enabled/available for use
411 - Rename operations (cluster, instance) will now return the new name,
412 which is especially useful if a short name was passed in
413 - Added support for instance migration in RAPI
414 - Added a tool to pre-configure nodes for the SSH setup, before joining
415 them to the cluster; this will allow in the future a simplified model
416 for node joining (but not yet fully enabled in 2.2); this needs the
417 paramiko python library
418 - Fixed handling of name-resolving errors
419 - Fixed consistency of job results on the error path
420 - Fixed master-failover race condition when executed multiple times in
422 - Fixed many bugs related to the job queue (mostly introduced during the
423 2.2 development cycle, so not all are impacting 2.1)
424 - Fixed instance migration with missing disk symlinks
425 - Fixed handling of unknown jobs in ``gnt-job archive``
426 - And many other small fixes/improvements
430 - Enhanced both the unittest and the QA coverage
431 - Switched the opcode validation to a generic model, and extended the
432 validation to all opcode parameters
433 - Changed more parts of the code that write shell scripts to use the
435 - Switched the master daemon to use the asyncore library for the Luxi
442 *(Released Thu, 17 Jun 2010)*
444 - Added tool (``move-instance``) and infrastructure to move instances
445 between separate clusters (see :doc:`separate documentation
446 <move-instance>` and :doc:`design document <design-2.2>`)
447 - Added per-request RPC timeout
448 - RAPI now requires a Content-Type header for requests with a body (e.g.
449 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
450 :rfc:`2616` (HTTP/1.1), section 7.2.1)
451 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
453 - Implemented initial support for running Ganeti daemons as separate
454 users, see configure-time flags ``--with-user-prefix`` and
455 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
457 - Instances can be removed after export (``gnt-backup export
459 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
460 key (instead of 1024 bit)
461 - Added new cluster configuration file for cluster domain secret
462 - Import/export now use SSL instead of SSH
463 - Added support for showing estimated time when exporting an instance,
464 see the ``ganeti-os-interface(7)`` manpage and look for
471 *(Released Tue, 16 Nov 2010)*
473 Some more bugfixes. Unless critical bugs occur, this will be the last
476 - Fix case of MAC special-values
477 - Fix mac checker regex
478 - backend: Fix typo causing “out of range” error
479 - Add missing --units in gnt-instance list man page
485 *(Released Tue, 24 Aug 2010)*
488 - Don't ignore secondary node silently on non-mirrored disk templates
490 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
491 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
492 - Properly document the usb_mouse parameter
493 - Fix path in ganeti-rapi(8) (issue 116)
494 - Adjust error message when the ganeti user's .ssh directory is
496 - Add same-node-check when changing the disk template to drbd
502 *(Released Fri, 16 Jul 2010)*
505 - Add an option to only select some reboot types during qa/burnin.
506 (on some hypervisors consequent reboots are not supported)
507 - Fix infrequent race condition in master failover. Sometimes the old
508 master ip address would be still detected as up for a short time
509 after it was removed, causing failover to fail.
510 - Decrease mlockall warnings when the ctypes module is missing. On
511 Python 2.4 we support running even if no ctypes module is installed,
512 but we were too verbose about this issue.
513 - Fix building on old distributions, on which man doesn't have a
515 - Fix RAPI not to ignore the MAC address on instance creation
516 - Implement the old instance creation format in the RAPI client.
522 *(Released Thu, 01 Jul 2010)*
524 A small bugfix release:
525 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
526 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
527 a non-existing option
528 - Fix instance create: the --force-variant option was ignored
529 - Improve pylint 0.21 compatibility and warnings with Python 2.6
530 - Fix modify node storage with non-FQDN arguments
531 - Fix RAPI client to authenticate under Python 2.6 when used
532 for more than 5 requests needing authentication
533 - Fix gnt-instance modify -t (storage) giving a wrong error message
534 when converting a non-shutdown drbd instance to plain
540 *(Released Fri, 18 Jun 2010)*
542 A small bugfix release:
544 - Fix live migration of KVM instances started with older Ganeti
545 versions which had fewer hypervisor parameters
546 - Fix gnt-instance grow-disk on down instances
547 - Fix an error-reporting bug during instance migration
548 - Better checking of the ``--net`` and ``--disk`` values, to avoid
549 silently ignoring broken ones
550 - Fix an RPC error reporting bug affecting, for example, RAPI client
552 - Fix bug triggered by different API version os-es on different nodes
553 - Fix a bug in instance startup with custom hvparams: OS level
554 parameters would fail to be applied.
555 - Fix the RAPI client under Python 2.6 (but more work is needed to
556 make it work completely well with OpenSSL)
557 - Fix handling of errors when resolving names from DNS
563 *(Released Thu, 3 Jun 2010)*
565 A medium sized development cycle. Some new features, and some
566 fixes/small improvements/cleanups.
571 The node deamon now tries to mlock itself into memory, unless the
572 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
573 its logs, and falls back to console logging. This allows emergency
574 features such as ``gnt-node powercycle`` to work even in the event of a
575 broken node disk (tested offlining the disk hosting the node's
576 filesystem and dropping its memory caches; don't try this at home)
578 KVM: add vhost-net acceleration support. It can be tested with a new
579 enough version of the kernel and of qemu-kvm.
581 KVM: Add instance chrooting feature. If you use privilege dropping for
582 your VMs you can also now force them to chroot to an empty directory,
583 before starting the emulated guest.
585 KVM: Add maximum migration bandwith and maximum downtime tweaking
586 support (requires a new-enough version of qemu-kvm).
588 Cluster verify will now warn if the master node doesn't have the master
591 Add a new (incompatible) instance creation request format to RAPI which
592 supports all parameters (previously only a subset was supported, and it
593 wasn't possible to extend the old format to accomodate all the new
594 features. The old format is still supported, and a client can check for
595 this feature, before using it, by checking for its presence in the
596 ``features`` RAPI resource.
598 Now with ancient latin support. Try it passing the ``--roman`` option to
599 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
600 (requires the python-roman module to be installed, in order to work).
605 As usual many internal code refactorings, documentation updates, and
608 - Lots of improvements and cleanups to the experimental Remote API
609 (RAPI) client library.
610 - A new unit test suite for the core daemon libraries.
611 - A fix to creating missing directories makes sure the umask is not
612 applied anymore. This enforces the same directory permissions
614 - Better handling terminating daemons with ctrl+c (used when running
615 them in debugging mode).
616 - Fix a race condition in live migrating a KVM instance, when stat()
617 on the old proc status file returned EINVAL, which is an unexpected
619 - Fixed manpage checking with newer man and utf-8 charachters. But now
620 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
626 *(Released Fri, 7 May 2010)*
628 Fix a bug which prevented untagged KVM instances from starting.
634 *(Released Fri, 7 May 2010)*
636 Another release with a long development cycle, during which many
637 different features were added.
642 The KVM hypervisor now can run the individual instances as non-root, to
643 reduce the impact of a VM being hijacked due to bugs in the
644 hypervisor. It is possible to run all instances as a single (non-root)
645 user, to manually specify a user for each instance, or to dynamically
646 allocate a user out of a cluster-wide pool to each instance, with the
647 guarantee that no two instances will run under the same user ID on any
650 An experimental RAPI client library, that can be used standalone
651 (without the other Ganeti libraries), is provided in the source tree as
652 ``lib/rapi/client.py``. Note this client might change its interface in
653 the future, as we iterate on its capabilities.
655 A new command, ``gnt-cluster renew-crypto`` has been added to easily
656 replace the cluster's certificates and crypto keys. This might help in
657 case they have been compromised, or have simply expired.
659 A new disk option for instance creation has been added that allows one
660 to "adopt" currently existing logical volumes, with data
661 preservation. This should allow easier migration to Ganeti from
662 unmanaged (or managed via other software) instances.
664 Another disk improvement is the possibility to convert between redundant
665 (DRBD) and plain (LVM) disk configuration for an instance. This should
666 allow better scalability (starting with one node and growing the
667 cluster, or shrinking a two-node cluster to one node).
669 A new feature that could help with automated node failovers has been
670 implemented: if a node sees itself as offline (by querying the master
671 candidates), it will try to shutdown (hard) all instances and any active
672 DRBD devices. This reduces the risk of duplicate instances if an
673 external script automatically failovers the instances on such nodes. To
674 enable this, the cluster parameter ``maintain_node_health`` should be
675 enabled; in the future this option (per the name) will enable other
676 automatic maintenance features.
678 Instance export/import now will reuse the original instance
679 specifications for all parameters; that means exporting an instance,
680 deleting it and the importing it back should give an almost identical
681 instance. Note that the default import behaviour has changed from
682 before, where it created only one NIC; now it recreates the original
685 Cluster verify has added a few new checks: SSL certificates validity,
686 /etc/hosts consistency across the cluster, etc.
691 As usual, many internal changes were done, documentation fixes,
694 - Fixed cluster initialization with disabled cluster storage (regression
696 - File-based storage supports growing the disks
697 - Fixed behaviour of node role changes
698 - Fixed cluster verify for some corner cases, plus a general rewrite of
699 cluster verify to allow future extension with more checks
700 - Fixed log spamming by watcher and node daemon (regression introduced
702 - Fixed possible validation issues when changing the list of enabled
704 - Fixed cleanup of /etc/hosts during node removal
705 - Fixed RAPI response for invalid methods
706 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
707 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
708 from ide disks, etc.)
709 - Allow OS changes without re-installation (to record a changed OS
710 outside of Ganeti, or to allow OS renames)
711 - Allow instance creation without OS installation (useful for example if
712 the OS will be installed manually, or restored from a backup not in
714 - Implemented option to make cluster ``copyfile`` use the replication
716 - Added list of enabled hypervisors to ssconf (possibly useful for
718 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
720 - A partial form of node re-IP is possible via node readd, which now
721 allows changed node primary IP
722 - Command line utilities now show an informational message if the job is
724 - The logs of the master daemon now show the PID/UID/GID of the
731 *(Released Fri, 12 Mar 2010)*
733 During the 2.1.0 long release candidate cycle, a lot of improvements and
734 changes have accumulated with were released later as 2.1.1.
739 The node evacuate command (``gnt-node evacuate``) was significantly
740 rewritten, and as such the IAllocator protocol was changed - a new
741 request type has been added. This unfortunate change during a stable
742 series is designed to improve performance of node evacuations; on
743 clusters with more than about five nodes and which are well-balanced,
744 evacuation should proceed in parallel for all instances of the node
745 being evacuated. As such, any existing IAllocator scripts need to be
746 updated, otherwise the above command will fail due to the unknown
747 request. The provided "dumb" allocator has not been updated; but the
748 ganeti-htools package supports the new protocol since version 0.2.4.
750 Another important change is increased validation of node and instance
751 names. This might create problems in special cases, if invalid host
752 names are being used.
754 Also, a new layer of hypervisor parameters has been added, that sits at
755 OS level between the cluster defaults and the instance ones. This allows
756 customisation of virtualization parameters depending on the installed
757 OS. For example instances with OS 'X' may have a different KVM kernel
758 (or any other parameter) than the cluster defaults. This is intended to
759 help managing a multiple OSes on the same cluster, without manual
760 modification of each instance's parameters.
762 A tool for merging clusters, ``cluster-merge``, has been added in the
768 - Improved the int/float conversions that should make the code more
769 robust in face of errors from the node daemons
770 - Fixed the remove node code in case of internal configuration errors
771 - Fixed the node daemon behaviour in face of inconsistent queue
772 directory (e.g. read-only file-system where we can't open the files
774 - Fixed the behaviour of gnt-node modify for master candidate demotion;
775 now it either aborts cleanly or, if given the new “auto_promote”
776 parameter, will automatically promote other nodes as needed
777 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
778 completely prevent Ganeti from working
779 - Fixed bug for instance export when not all disks were successfully
781 - Fixed behaviour of node add when the new node is slow in starting up
783 - Fixed handling of signals in the LUXI client, which should improve
784 behaviour of command-line scripts
785 - Added checks for invalid node/instance names in the configuration (now
786 flagged during cluster verify)
787 - Fixed watcher behaviour for disk activation errors
788 - Fixed two potentially endless loops in http library, which led to the
789 RAPI daemon hanging and consuming 100% CPU in some cases
790 - Fixed bug in RAPI daemon related to hashed passwords
791 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
793 - Enhanced compatibility with non-Debian OSes, but not using absolute
794 path in some commands and allowing customisation of the ssh
795 configuration directory
796 - Fixed possible future issue with new Python versions by abiding to the
797 proper use of ``__slots__`` attribute on classes
798 - Added checks that should prevent directory traversal attacks
799 - Many documentation fixes based on feedback from users
804 - Added an “early_release” more for instance replace disks and node
805 evacuate, where we release locks earlier and thus allow higher
806 parallelism within the cluster
807 - Added watcher hooks, intended to allow the watcher to restart other
808 daemons (e.g. from the ganeti-nbma project), but they can be used of
809 course for any other purpose
810 - Added a compile-time disable for DRBD barriers, to increase
811 performance if the administrator trusts the power supply or the
812 storage system to not lose writes
813 - Added the option of using syslog for logging instead of, or in
814 addition to, Ganeti's own log files
815 - Removed boot restriction for paravirtual NICs for KVM, recent versions
816 can indeed boot from a paravirtual NIC
817 - Added a generic debug level for many operations; while this is not
818 used widely yet, it allows one to pass the debug value all the way to
820 - Enhanced the hooks environment for instance moves (failovers,
821 migrations) where the primary/secondary nodes changed during the
822 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
823 - Enhanced data validations for many user-supplied values; one important
824 item is the restrictions imposed on instance and node names, which
825 might reject some (invalid) host names
826 - Add a configure-time option to disable file-based storage, if it's not
827 needed; this allows greater security separation between the master
828 node and the other nodes from the point of view of the inter-node RPC
830 - Added user notification in interactive tools if job is waiting in the
831 job queue or trying to acquire locks
832 - Added log messages when a job is waiting for locks
833 - Added filtering by node tags in instance operations which admit
834 multiple instances (start, stop, reboot, reinstall)
835 - Added a new tool for cluster mergers, ``cluster-merge``
836 - Parameters from command line which are of the form ``a=b,c=d`` can now
837 use backslash escapes to pass in values which contain commas,
838 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
840 - For KVM, the instance name is the first parameter passed to KVM, so
841 that it's more visible in the process list
847 *(Released Tue, 2 Mar 2010)*
849 Ganeti 2.1 brings many improvements with it. Major changes:
851 - Added infrastructure to ease automated disk repairs
852 - Added new daemon to export configuration data in a cheaper way than
854 - Instance NICs can now be routed instead of being associated with a
856 - Improved job locking logic to reduce impact of jobs acquiring multiple
857 locks waiting for other long-running jobs
859 In-depth implementation details can be found in the Ganeti 2.1 design
865 - Added chroot hypervisor
866 - Added more options to xen-hvm hypervisor (``kernel_path`` and
868 - Added more options to xen-pvm hypervisor (``use_bootloader``,
869 ``bootloader_path`` and ``bootloader_args``)
870 - Added the ``use_localtime`` option for the xen-hvm and kvm
871 hypervisors, and the default value for this has changed to false (in
872 2.0 xen-hvm always enabled it)
873 - Added luxi call to submit multiple jobs in one go
874 - Added cluster initialization option to not modify ``/etc/hosts``
876 - Added network interface parameters
877 - Added dry run mode to some LUs
878 - Added RAPI resources:
880 - ``/2/instances/[instance_name]/info``
881 - ``/2/instances/[instance_name]/replace-disks``
882 - ``/2/nodes/[node_name]/evacuate``
883 - ``/2/nodes/[node_name]/migrate``
884 - ``/2/nodes/[node_name]/role``
885 - ``/2/nodes/[node_name]/storage``
886 - ``/2/nodes/[node_name]/storage/modify``
887 - ``/2/nodes/[node_name]/storage/repair``
889 - Added OpCodes to evacuate or migrate all instances on a node
890 - Added new command to list storage elements on nodes (``gnt-node
891 list-storage``) and modify them (``gnt-node modify-storage``)
892 - Added new ssconf files with master candidate IP address
893 (``ssconf_master_candidates_ips``), node primary IP address
894 (``ssconf_node_primary_ips``) and node secondary IP address
895 (``ssconf_node_secondary_ips``)
896 - Added ``ganeti-confd`` and a client library to query the Ganeti
897 configuration via UDP
898 - Added ability to run hooks after cluster initialization and before
900 - Added automatic mode for disk replace (``gnt-instance replace-disks
902 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
903 after catastrophic data-loss
904 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
906 - Added ``gnt-instance move`` command to move instances
907 - Added ``gnt-cluster watcher`` command to control watcher
908 - Added ``gnt-node powercycle`` command to powercycle nodes
909 - Added new job status field ``lock_status``
910 - Added parseable error codes to cluster verification (``gnt-cluster
911 verify --error-codes``) and made output less verbose (use
912 ``--verbose`` to restore previous behaviour)
913 - Added UUIDs to the main config entities (cluster, nodes, instances)
914 - Added support for OS variants
915 - Added support for hashed passwords in the Ganeti remote API users file
917 - Added option to specify maximum timeout on instance shutdown
918 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
919 - Added new helper script to start and stop Ganeti daemons
920 (``daemon-util``), with the intent to reduce the work necessary to
921 adjust Ganeti for non-Debian distributions and to start/stop daemons
923 - Added more unittests
924 - Fixed critical bug in ganeti-masterd startup
925 - Removed the configure-time ``kvm-migration-port`` parameter, this is
926 now customisable at the cluster level for both the KVM and Xen
927 hypervisors using the new ``migration_port`` parameter
928 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
929 reinstalling an instance
930 - Allowed ``@`` in tag names
931 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
932 - Many documentation updates
933 - Distribute hypervisor files on ``gnt-cluster redist-conf``
934 - ``gnt-instance reinstall`` can now reinstall multiple instances
935 - Updated many command line parameters
936 - Introduced new OS API version 15
937 - No longer support a default hypervisor
938 - Treat virtual LVs as inexistent
939 - Improved job locking logic to reduce lock contention
940 - Match instance and node names case insensitively
941 - Reimplemented bash completion script to be more complete
948 *(Released Thu, 4 Feb 2010)*
950 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
951 - Fix a string formatting bug
952 - Improve safety of the code in some error paths
953 - Improve data validation in the master of values returned from nodes
959 *(Released Thu, 17 Dec 2009)*
961 - Fix security issue due to missing validation of iallocator names; this
962 allows local and remote execution of arbitrary executables
963 - Fix failure of gnt-node list during instance removal
964 - Ship the RAPI documentation in the archive
970 *(Released Wed, 30 Sep 2009)*
972 - Fixed many wrong messages
973 - Fixed a few bugs related to the locking library
974 - Fixed MAC checking at instance creation time
975 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
976 - Fixed a few issues related to signal handling in both daemons and
978 - Fixed the example startup script provided
979 - Fixed insserv dependencies in the example startup script (patch from
981 - Fixed handling of drained nodes in the iallocator framework
982 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
984 - Fixed error related to invalid job IDs in job polling
985 - Fixed job/opcode persistence on unclean master shutdown
986 - Fixed handling of partial job processing after unclean master
988 - Fixed error reporting from LUs, previously all errors were converted
989 into execution errors
990 - Fixed error reporting from burnin
991 - Decreased significantly the memory usage of the job queue
992 - Optimised slightly multi-job submission
993 - Optimised slightly opcode loading
994 - Backported the multi-job submit framework from the development
995 branch; multi-instance start and stop should be faster
996 - Added script to clean archived jobs after 21 days; this will reduce
997 the size of the queue directory
998 - Added some extra checks in disk size tracking
999 - Added an example ethers hook script
1000 - Added a cluster parameter that prevents Ganeti from modifying of
1002 - Added more node information to RAPI responses
1003 - Added a ``gnt-job watch`` command that allows following the ouput of a
1005 - Added a bind-address option to ganeti-rapi
1006 - Added more checks to the configuration verify
1007 - Enhanced the burnin script such that some operations can be retried
1009 - Converted instance reinstall to multi-instance model
1015 *(Released Fri, 7 Aug 2009)*
1017 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
1018 to allow using the pre-2.0.2 behaviour in activation, if any existing
1019 instances have mismatched disk sizes in the configuration
1020 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
1021 any configuration mismatches for disk sizes
1022 - Added ``gnt-master cluste-failover --no-voting`` to allow master
1023 failover to work on two-node clusters
1024 - Fixed the ``--net`` option of ``gnt-backup import``, which was
1026 - Fixed detection of OS script errors in ``gnt-backup export``
1027 - Fixed exit code of ``gnt-backup export``
1033 *(Released Fri, 17 Jul 2009)*
1035 - Added experimental support for stripped logical volumes; this should
1036 enhance performance but comes with a higher complexity in the block
1037 device handling; stripping is only enabled when passing
1038 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
1039 affected even in the non-stripped mode
1040 - Improved resiliency against transient failures at the end of DRBD
1041 resyncs, and in general of DRBD resync checks
1042 - Fixed a couple of issues with exports and snapshot errors
1043 - Fixed a couple of issues in instance listing
1044 - Added display of the disk size in ``gnt-instance info``
1045 - Fixed checking for valid OSes in instance creation
1046 - Fixed handling of the "vcpus" parameter in instance listing and in
1047 general of invalid parameters
1048 - Fixed http server library, and thus RAPI, to handle invalid
1049 username/password combinations correctly; this means that now they
1050 report unauthorized for queries too, not only for modifications,
1051 allowing earlier detect of configuration problems
1052 - Added a new "role" node list field, equivalent to the master/master
1053 candidate/drained/offline flags combinations
1054 - Fixed cluster modify and changes of candidate pool size
1055 - Fixed cluster verify error messages for wrong files on regular nodes
1056 - Fixed a couple of issues with node demotion from master candidate role
1057 - Fixed node readd issues
1058 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
1059 - Added a new ``--no-voting`` option for masterfailover to fix failover
1060 on two-nodes clusters when the former master node is unreachable
1061 - Added instance reinstall over RAPI
1067 *(Released Tue, 16 Jun 2009)*
1069 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
1070 allow re-adding the start in single-user option (regression from 1.2)
1071 - the watcher writes the instance status to a file, to allow monitoring
1072 to report the instance status (from the master) based on cached
1073 results of the watcher's queries; while this can get stale if the
1074 watcher is being locked due to other work on the cluster, this is
1075 still an improvement
1076 - the watcher now also restarts the node daemon and the rapi daemon if
1078 - fixed the watcher to handle full and drained queue cases
1079 - hooks export more instance data in the environment, which helps if
1080 hook scripts need to take action based on the instance's properties
1081 (no longer need to query back into ganeti)
1082 - instance failovers when the instance is stopped do not check for free
1083 RAM, so that failing over a stopped instance is possible in low memory
1085 - rapi uses queries for tags instead of jobs (for less job traffic), and
1086 for cluster tags it won't talk to masterd at all but read them from
1088 - a couple of error handling fixes in RAPI
1089 - drbd handling: improved the error handling of inconsistent disks after
1090 resync to reduce the frequency of "there are some degraded disks for
1091 this instance" messages
1092 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
1093 error handling path called a wrong function name)
1099 *(Released Wed, 27 May 2009)*
1101 - no changes from rc5
1104 Version 2.0 release candidate 5
1105 -------------------------------
1107 *(Released Wed, 20 May 2009)*
1109 - fix a couple of bugs (validation, argument checks)
1110 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
1111 - some small improvements to RAPI and IAllocator
1112 - make watcher automatically start the master daemon if down
1115 Version 2.0 release candidate 4
1116 -------------------------------
1118 *(Released Mon, 27 Apr 2009)*
1120 - change the OS list to not require locks; this helps with big clusters
1121 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
1122 volume group is broken
1123 - ``gnt-instance info``, without any arguments, doesn't run for all
1124 instances anymore; either pass ``--all`` or pass the desired
1125 instances; this helps against mistakes on big clusters where listing
1126 the information for all instances takes a long time
1127 - miscellaneous doc and man pages fixes
1130 Version 2.0 release candidate 3
1131 -------------------------------
1133 *(Released Wed, 8 Apr 2009)*
1135 - Change the internal locking model of some ``gnt-node`` commands, in
1136 order to reduce contention (and blocking of master daemon) when
1137 batching many creation/reinstall jobs
1138 - Fixes to Xen soft reboot
1139 - No longer build documentation at build time, instead distribute it in
1140 the archive, in order to reduce the need for the whole docbook/rst
1144 Version 2.0 release candidate 2
1145 -------------------------------
1147 *(Released Fri, 27 Mar 2009)*
1149 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
1150 - Fix watcher startup sequence, improves the behaviour of busy clusters
1151 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
1152 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
1153 bind address and other places
1154 - Some documentation fixes and updates
1157 Version 2.0 release candidate 1
1158 -------------------------------
1160 *(Released Mon, 2 Mar 2009)*
1162 - More documentation updates, now all docs should be more-or-less
1164 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
1166 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
1168 - Fix an issue related to $libdir/run/ganeti and cluster creation
1174 *(Released Thu, 19 Feb 2009)*
1176 - Xen PVM and KVM have switched the default value for the instance root
1177 disk to the first partition on the first drive, instead of the whole
1178 drive; this means that the OS installation scripts must be changed
1180 - Man pages have been updated
1181 - RAPI has been switched by default to HTTPS, and the exported functions
1182 should all work correctly
1183 - RAPI v1 has been removed
1184 - Many improvements to the KVM hypervisor
1185 - Block device errors are now better reported
1186 - Many other bugfixes and small improvements
1192 *(Released Mon, 26 Jan 2009)*
1194 - Version 2 is a general rewrite of the code and therefore the
1195 differences are too many to list, see the design document for 2.0 in
1196 the ``doc/`` subdirectory for more details
1197 - In this beta version there is not yet a migration path from 1.2 (there
1198 will be one in the final 2.0 release)
1199 - A few significant changes are:
1201 - all commands are executed by a daemon (``ganeti-masterd``) and the
1202 various ``gnt-*`` commands are just front-ends to it
1203 - all the commands are entered into, and executed from a job queue,
1204 see the ``gnt-job(8)`` manpage
1205 - the RAPI daemon supports read-write operations, secured by basic
1206 HTTP authentication on top of HTTPS
1207 - DRBD version 0.7 support has been removed, DRBD 8 is the only
1208 supported version (when migrating from Ganeti 1.2 to 2.0, you need
1209 to migrate to DRBD 8 first while still running Ganeti 1.2)
1210 - DRBD devices are using statically allocated minor numbers, which
1211 will be assigned to existing instances during the migration process
1212 - there is support for both Xen PVM and Xen HVM instances running on
1214 - KVM virtualization is supported too
1215 - file-based storage has been implemented, which means that it is
1216 possible to run the cluster without LVM and DRBD storage, for
1217 example using a shared filesystem exported from shared storage (and
1218 still have live migration)
1224 *(Released Tue, 13 Jan 2009)*
1226 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
1227 - Reuse the old instance mac address by default on instance import, if
1228 the instance name is the same.
1229 - Handle situations in which the node info rpc returns incomplete
1231 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
1232 - Improved version of batcher:
1234 - state file support
1235 - instance mac address support
1236 - support for HVM clusters/instances
1238 - Add an option to show the number of cpu sockets and nodes in
1240 - Support OSes that handle more than one version of the OS api (but do
1241 not change the current API in any other way)
1242 - Fix ``gnt-node migrate``
1243 - ``gnt-debug`` man page
1244 - Fixes various more typos and small issues
1245 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
1251 *(Released Wed, 24 Sep 2008)*
1253 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
1254 type of disk exported to fully virtualized instances.
1255 - provide access to the serial console of HVM instances
1256 - instance auto_balance flag, set by default. If turned off it will
1257 avoid warnings on cluster verify if there is not enough memory to fail
1258 over an instance. in the future it will prevent automatically failing
1259 it over when we will support that.
1260 - batcher tool for instance creation, see ``tools/README.batcher``
1261 - ``gnt-instance reinstall --select-os`` to interactively select a new
1262 operating system when reinstalling an instance.
1263 - when changing the memory amount on instance modify a check has been
1264 added that the instance will be able to start. also warnings are
1265 emitted if the instance will not be able to fail over, if auto_balance
1267 - documentation fixes
1268 - sync fields between ``gnt-instance list/modify/add/import``
1269 - fix a race condition in drbd when the sync speed was set after giving
1270 the device a remote peer.
1276 *(Released Tue, 22 Jul 2008)*
1278 - note: the allowed size and number of tags per object were reduced
1279 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
1280 - fixed twisted 8.x compatibility
1281 - fixed ``gnt-instance replace-disks`` with iallocator
1282 - add TCP keepalives on twisted connections to detect restarted nodes
1283 - disk increase support, see ``gnt-instance grow-disk``
1284 - implement bulk node/instance query for RAPI
1285 - add tags in node/instance listing (optional)
1286 - experimental migration (and live migration) support, read the man page
1287 for ``gnt-instance migrate``
1288 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
1289 has some small improvements in handling its state file
1295 *(Released Fri, 13 Jun 2008)*
1297 - Experimental readonly, REST-based remote API implementation;
1298 automatically started on master node, TCP port 5080, if enabled by
1299 ``--enable-rapi`` parameter to configure script.
1300 - Instance allocator support. Add and import instance accept a
1301 ``--iallocator`` parameter, and call that instance allocator to decide
1302 which node to use for the instance. The iallocator document describes
1303 what's expected from an allocator script.
1304 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
1305 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
1306 node is lost there is still enough memory to fail over the instances
1308 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
1309 ``gnt-cluster verify``, to check for site-specific compliance. All the
1310 hooks will run, and their output, if any, will be displayed. Any
1311 failing hook will make the verification return an error value.
1312 - ``gnt-cluster verify`` now checks that its peers are reachable on the
1313 primary and secondary interfaces
1314 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
1315 that is still declared as part of the cluster and has failed.
1316 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
1317 specifying output fields, that just adds the chosen fields to the
1319 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
1320 export from the filesystem.
1321 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
1322 have been added. Using them you can enable/disable acpi and pae
1323 support, and specify a path for a cd image to be exported to the
1324 instance. These parameters as the name suggest only work on HVM
1326 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
1327 PAE support will be set to the previously hardcoded values, but the
1328 (previously hardcoded) path to the CDROM ISO image will be unset and
1329 if required, needs to be set manually with ``gnt-instance modify``
1331 - The address to which an instance's VNC console is bound is now
1332 selectable per-instance, rather than being cluster wide. Of course
1333 this only applies to instances controlled via VNC, so currently just
1334 applies to HVM clusters.
1340 *(Released Mon, 18 Feb 2008)*
1342 - more tweaks to the disk activation code (especially helpful for DRBD)
1343 - change the default ``gnt-instance list`` output format, now there is
1344 one combined status field (see the manpage for the exact values this
1346 - some more fixes for the mac export to hooks change
1347 - make Ganeti not break with DRBD 8.2.x (which changed the version
1348 format in ``/proc/drbd``) (issue 24)
1349 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
1350 template, allowing migration from DRBD0.7+MD to DRBD8
1356 *(Released Wed, 30 Jan 2008)*
1358 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
1360 - add command aliases infrastructure and a few aliases
1361 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
1362 man pages and the ``--help`` option of ``gnt-node
1363 list``/``gnt-instance list``
1364 - fix ``gnt-backup list`` with down nodes (issue 21)
1365 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
1366 - fix the dist archive and add a check for including svn/git files in
1368 - some developer-related changes: improve the burnin and the QA suite,
1369 add an upload script for testing during development
1375 *(Released Wed, 16 Jan 2008)*
1377 - experimental HVM support, read the install document, section
1378 "Initializing the cluster"
1379 - allow for the PVM hypervisor per-instance kernel and initrd paths
1380 - add a new command ``gnt-cluster verify-disks`` which uses a new
1381 algorithm to improve the reconnection of the DRBD pairs if the device
1382 on the secondary node has gone away
1383 - make logical volume code auto-activate LVs at disk activation time
1384 - slightly improve the speed of activating disks
1385 - allow specification of the MAC address at instance creation time, and
1386 changing it later via ``gnt-instance modify``
1387 - fix handling of external commands that generate lots of output on
1389 - update documentation with regard to minimum version of DRBD8 supported
1395 *(Released Tue, 4 Dec 2007)*
1397 - Log the ``xm create`` output to the node daemon log on failure (to
1398 help diagnosing the error)
1399 - In debug mode, log all external commands output if failed to the logs
1400 - Change parsing of lvm commands to ignore stderr
1406 *(Released Wed, 28 Nov 2007)*
1408 - Another round of updates to the DRBD 8 code to deal with more failures
1409 in the replace secondary node operation
1410 - Some more logging of failures in disk operations (lvm, drbd)
1411 - A few documentation updates
1418 *(Released Tue, 13 Nov 2007)*
1420 - Change configuration file format from Python's Pickle to JSON.
1421 Upgrading is possible using the cfgupgrade utility.
1422 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
1423 faster replace disks and is more stable (DRBD 8 has many improvements
1424 compared to DRBD 0.7)
1425 - Added command line tags support (see man pages for ``gnt-instance``,
1426 ``gnt-node``, ``gnt-cluster``)
1427 - Added instance rename support
1428 - Added multi-instance startup/shutdown
1429 - Added cluster rename support
1430 - Added ``gnt-node evacuate`` to simplify some node operations
1431 - Added instance reboot operation that can speedup reboot as compared to
1433 - Soften the requirement that hostnames are in FQDN format
1434 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
1436 - Removed dependency on debian's patched fping that uses the
1437 non-standard ``-S`` option
1438 - Now the OS definitions are searched for in multiple, configurable
1439 paths (easier for distros to package)
1440 - Some changes to the hooks infrastructure (especially the new
1441 post-configuration update hook)
1442 - Other small bugfixes
1444 .. vim: set textwidth=72 syntax=rst :