10 - Deprecated ``admin_up`` field. Instead, ``admin_state`` is introduced,
11 with 3 possible values -- ``up``, ``down`` and ``offline``.
12 - Replaced ``--disks`` option of ``gnt-instance replace-disks`` with a
13 more flexible ``--disk`` option. Now disk size and mode can be changed
15 - Removed deprecated ``QueryLocks`` LUXI request. Use
16 ``Query(what=QR_LOCK, ...)`` instead.
17 - The LUXI requests :pyeval:`luxi.REQ_QUERY_JOBS`,
18 :pyeval:`luxi.REQ_QUERY_INSTANCES`, :pyeval:`luxi.REQ_QUERY_NODES`,
19 :pyeval:`luxi.REQ_QUERY_GROUPS`, :pyeval:`luxi.REQ_QUERY_EXPORTS` and
20 :pyeval:`luxi.REQ_QUERY_TAGS` are deprecated and will be removed in a
21 future version. :pyeval:`luxi.REQ_QUERY` should be used instead.
27 *(Released Thu, 12 Apr 2012)*
29 Incompatible/important changes and bugfixes
30 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
32 - The default of the ``/2/instances/[instance_name]/rename`` RAPI
33 resource's ``ip_check`` parameter changed from ``True`` to ``False``
34 to match the underlying LUXI interface.
35 - The ``/2/nodes/[node_name]/evacuate`` RAPI resource was changed to use
36 body parameters, see :doc:`RAPI documentation <rapi>`. The server does
37 not maintain backwards-compatibility as the underlying operation
38 changed in an incompatible way. The RAPI client can talk to old
39 servers, but it needs to be told so as the return value changed.
40 - When creating file-based instances via RAPI, the ``file_driver``
41 parameter no longer defaults to ``loop`` and must be specified.
42 - The deprecated ``bridge`` NIC parameter is no longer supported. Use
44 - Support for the undocumented and deprecated RAPI instance creation
45 request format version 0 has been dropped. Use version 1, supported
46 since Ganeti 2.1.3 and :doc:`documented <rapi>`, instead.
47 - Pyparsing 1.4.6 or above is required, see :doc:`installation
48 documentation <install>`.
49 - The "cluster-verify" hooks are now executed per group by the
50 ``OP_CLUSTER_VERIFY_GROUP`` opcode. This maintains the same behavior
51 if you just run ``gnt-cluster verify``, which generates one opcode per
53 - The environment as passed to the OS scripts is cleared, and thus no
54 environment variables defined in the node daemon's environment will be
55 inherited by the scripts.
56 - The :doc:`iallocator <iallocator>` mode ``multi-evacuate`` has been
58 - :doc:`New iallocator modes <design-multi-reloc>` have been added to
59 support operations involving multiple node groups.
60 - Offline nodes are ignored when failing over an instance.
61 - Support for KVM version 1.0, which changed the version reporting format
63 - TCP/IP ports used by DRBD disks are returned to a pool upon instance
65 - ``Makefile`` is now compatible with Automake 1.11.2
66 - Includes all bugfixes made in the 2.4 series
71 - The ganeti-htools project has been merged into the ganeti-core source
72 tree and will be built as part of Ganeti (see :doc:`install-quick`).
73 - Implemented support for :doc:`shared storage <design-shared-storage>`.
74 - Add support for disks larger than 2 TB in ``lvmstrap`` by supporting
75 GPT-style partition tables (requires `parted
76 <http://www.gnu.org/s/parted/>`_).
77 - Added support for floppy drive and 2nd CD-ROM drive in KVM hypervisor.
78 - Allowed adding tags on instance creation.
79 - Export instance tags to hooks (``INSTANCE_TAGS``, see :doc:`hooks`)
80 - Allow instances to be started in a paused state, enabling the user to
81 see the complete console output on boot using the console.
82 - Added new hypervisor flag to control default reboot behaviour
83 (``reboot_behavior``).
84 - Added support for KVM keymaps (hypervisor parameter ``keymap``).
85 - Improved out-of-band management support:
87 - Added ``gnt-node health`` command reporting the health status of
89 - Added ``gnt-node power`` command to manage power status of nodes.
90 - Added command for emergency power-off (EPO), ``gnt-cluster epo``.
92 - Instance migration can fall back to failover if instance is not
94 - Filters can be used when listing nodes, instances, groups and locks;
95 see *ganeti(7)* manpage.
96 - Added post-execution status as variables to :doc:`hooks <hooks>`
98 - Instance tags are exported/imported together with the instance.
99 - When given an explicit job ID, ``gnt-job info`` will work for archived
101 - Jobs can define dependencies on other jobs (not yet supported via
102 RAPI or command line, but used by internal commands and usable via
105 - Lock monitor (``gnt-debug locks``) shows jobs waiting for
108 - Instance failover is now available as a RAPI resource
109 (``/2/instances/[instance_name]/failover``).
110 - ``gnt-instance info`` defaults to static information if primary node
112 - Opcodes have a new ``comment`` attribute.
113 - Added basic SPICE support to KVM hypervisor.
114 - ``tools/ganeti-listrunner`` allows passing of arguments to executable.
116 Node group improvements
117 ~~~~~~~~~~~~~~~~~~~~~~~
119 - ``gnt-cluster verify`` has been modified to check groups separately,
120 thereby improving performance.
121 - Node group support has been added to ``gnt-cluster verify-disks``,
122 which now operates per node group.
123 - Watcher has been changed to work better with node groups.
125 - One process and state file per node group.
126 - Slow watcher in one group doesn't block other group's watcher.
128 - Added new command, ``gnt-group evacuate``, to move all instances in a
129 node group to other groups.
130 - Added ``gnt-instance change-group`` to move an instance to another
132 - ``gnt-cluster command`` and ``gnt-cluster copyfile`` now support
133 per-group operations.
134 - Node groups can be tagged.
135 - Some operations switch from an exclusive to a shared lock as soon as
137 - Instance's primary and secondary nodes' groups are now available as
138 query fields (``pnode.group``, ``pnode.group.uuid``, ``snodes.group``
139 and ``snodes.group.uuid``).
144 - Numerous updates to documentation and manpages.
146 - :doc:`RAPI <rapi>` documentation now has detailed parameter
148 - Some opcode/job results are now also documented, see :doc:`RAPI
151 - A lockset's internal lock is now also visible in lock monitor.
152 - Log messages from job queue workers now contain information about the
153 opcode they're processing.
154 - ``gnt-instance console`` no longer requires the instance lock.
155 - A short delay when waiting for job changes reduces the number of LUXI
156 requests significantly.
157 - DRBD metadata volumes are overwritten with zeros during disk creation.
158 - Out-of-band commands no longer acquire the cluster lock in exclusive
160 - ``devel/upload`` now uses correct permissions for directories.
166 *(Released Fri, 23 Mar 2012)*
168 This was the sixth release candidate of the 2.5 series.
174 *(Released Mon, 9 Jan 2012)*
176 This was the fifth release candidate of the 2.5 series.
182 *(Released Thu, 27 Oct 2011)*
184 This was the fourth release candidate of the 2.5 series.
190 *(Released Wed, 26 Oct 2011)*
192 This was the third release candidate of the 2.5 series.
198 *(Released Tue, 18 Oct 2011)*
200 This was the second release candidate of the 2.5 series.
206 *(Released Tue, 4 Oct 2011)*
208 This was the first release candidate of the 2.5 series.
214 *(Released Wed, 31 Aug 2011)*
216 This was the third beta release of the 2.5 series.
222 *(Released Mon, 22 Aug 2011)*
224 This was the second beta release of the 2.5 series.
230 *(Released Fri, 12 Aug 2011)*
232 This was the first beta release of the 2.5 series.
238 *(Released Thu, 27 Oct 2011)*
240 - Fixed bug when parsing command line parameter values ending in
242 - Fixed assertion error after unclean master shutdown
243 - Disable HTTP client pool for RPC, significantly reducing memory usage
245 - Fixed queue archive creation with wrong permissions
251 *(Released Tue, 23 Aug 2011)*
255 - Fixed documentation for importing with ``--src-dir`` option
256 - Fixed a bug in ``ensure-dirs`` with queue/archive permissions
257 - Fixed a parsing issue with DRBD 8.3.11 in the Linux kernel
263 *(Released Fri, 5 Aug 2011)*
265 Many bug-fixes and a few small features:
267 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
268 issues when you tried to add an instance with two MAC addresses in one
270 - KVM: fixed per-instance stored UID value
271 - KVM: configure bridged NICs at migration start
272 - KVM: Fix a bug where instance will not start with never KVM versions
274 - Added OS search path to ``gnt-cluster info``
275 - Fixed an issue with ``file_storage_dir`` where you were forced to
276 provide an absolute path, but the documentation states it is a
277 relative path, the documentation was right
278 - Added a new parameter to instance stop/start called ``--no-remember``
279 that will make the state change to not be remembered
280 - Implemented ``no_remember`` at RAPI level
281 - Improved the documentation
282 - Node evacuation: don't call IAllocator if node is already empty
283 - Fixed bug in DRBD8 replace disks on current nodes
284 - Fixed bug in recreate-disks for DRBD instances
285 - Moved assertion checking locks in ``gnt-instance replace-disks``
286 causing it to abort with not owning the right locks for some situation
287 - Job queue: Fixed potential race condition when cancelling queued jobs
288 - Fixed off-by-one bug in job serial generation
289 - ``gnt-node volumes``: Fix instance names
290 - Fixed aliases in bash completion
291 - Fixed a bug in reopening log files after being sent a SIGHUP
292 - Added a flag to burnin to allow specifying VCPU count
293 - Bugfixes to non-root Ganeti configuration
299 *(Released Thu, 12 May 2011)*
301 Many bug-fixes and a few new small features:
303 - Fixed a bug related to log opening failures
304 - Fixed a bug in instance listing with orphan instances
305 - Fixed a bug which prevented resetting the cluster-level node parameter
306 ``oob_program`` to the default
307 - Many fixes related to the ``cluster-merge`` tool
308 - Fixed a race condition in the lock monitor, which caused failures
309 during (at least) creation of many instances in parallel
310 - Improved output for gnt-job info
311 - Removed the quiet flag on some ssh calls which prevented debugging
313 - Improved the N+1 failure messages in cluster verify by actually
314 showing the memory values (needed and available)
315 - Increased lock attempt timeouts so that when executing long operations
316 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
317 too early and thus prevent the use of the 'fair' mechanism
318 - Changed instance query data (``gnt-instance info``) to not acquire
319 locks unless needed, thus allowing its use on locked instance if only
320 static information is asked for
321 - Improved behaviour with filesystems that do not support rename on an
323 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
324 kept locks on all nodes during the wipe, which is unneeded
325 - Fixed ``gnt-watcher`` handling of errors during hooks execution
326 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
327 10GiB) which caused the wipe to fail right at the end in some cases
328 - Fixed master IP activation when doing master failover with no-voting
329 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
330 the master node itself
331 - Fixed potential data-loss in under disk full conditions, where Ganeti
332 wouldn't check correctly the return code and would consider
333 partially-written files 'correct'
334 - Fixed bug related to multiple VGs and DRBD disk replacing
335 - Added new disk parameter ``metavg`` that allows placement of the meta
336 device for DRBD in a different volume group
337 - Fixed error handling in the node daemon when the system libc doesn't
338 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
339 - Fixed lock release during replace-disks, which kept cluster-wide locks
340 when doing disk replaces with an iallocator script
341 - Added check for missing bridges in cluster verify
342 - Handle EPIPE errors while writing to the terminal better, so that
343 piping the output to e.g. ``less`` doesn't cause a backtrace
344 - Fixed rare case where a ^C during Luxi calls could have been
345 interpreted as server errors, instead of simply terminating
346 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
348 - Added a few more parameters to the KVM hypervisor, allowing a second
349 CDROM, custom disk type for CDROMs and a floppy image
350 - Removed redundant message in instance rename when the name is given
352 - Added option to ``gnt-instance recreate-disks`` to allow creating the
353 disks on new nodes, allowing recreation when the original instance
354 nodes are completely gone
355 - Added option when converting disk templates to DRBD to skip waiting
356 for the resync, in order to make the instance available sooner
357 - Added two new variables to the OS scripts environment (containing the
359 - Made the root_path and optional parameter for the xen-pvm hypervisor,
360 to allow use of ``pvgrub`` as bootloader
361 - Changed the instance memory modifications to only check out-of-memory
362 conditions on memory increases, and turned the secondary node warnings
363 into errors (they can still be overridden via ``--force``)
364 - Fixed the handling of a corner case when the Python installation gets
365 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
366 to execute a command that doesn't exist
367 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
368 node of the instance returned failures during instance shutdown; this
369 adds the option ``--ignore-consistency`` to gnt-instance move
371 And as usual, various improvements to the error messages, documentation
378 *(Released Wed, 09 Mar 2011)*
380 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
381 the RAPI users file if run twice (even with ``--dry-run``).
383 The release fixes that bug (nothing else changed).
389 *(Released Mon, 07 Mar 2011)*
391 Final 2.4.0 release. Just a few small fixes:
393 - Fixed RAPI node evacuate
394 - Fixed the kvm-ifup script
395 - Fixed internal error handling for special job cases
396 - Updated man page to specify the escaping feature for options
402 *(Released Mon, 28 Feb 2011)*
404 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
405 that this feature wiped the disks of the wrong instance, leading to loss
410 - Fixed title of query field containing instance name
411 - Expanded the glossary in the documentation
412 - Fixed one unittest (internal issue)
418 *(Released Mon, 21 Feb 2011)*
420 A number of bug fixes plus just a couple functionality changes.
422 On the user-visible side, the ``gnt-* list`` command output has changed
423 with respect to "special" field states. The current rc1 style of display
424 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
425 the default output mode special fields states are displayed as follows:
427 - Offline resource: ``*``
428 - Unavailable/not applicable: ``-``
429 - Data missing (RPC failure): ``?``
430 - Unknown field: ``??``
432 Another user-visible change is the addition of ``--force-join`` to
437 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
438 - Fixed regression in RAPI/instance reinstall where all parameters were
439 required (instead of optional)
440 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
441 - Fixed iallocator usage (offline nodes were not considered offline)
442 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
443 - Fixed hypervisor and OS parameter validation with respect to
445 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
447 - Fixed ``tools/listrunner`` with respect to agent-based usage
453 *(Released Fri, 4 Feb 2011)*
455 Many changes and fixes since the beta1 release. While there were some
456 internal changes, the code has been mostly stabilised for the RC
459 Note: the dumb allocator was removed in this release, as it was not kept
460 up-to-date with the IAllocator protocol changes. It is recommended to
461 use the ``hail`` command from the ganeti-htools package.
463 Note: the 2.4 and up versions of Ganeti are not compatible with the
464 0.2.x branch of ganeti-htools. You need to upgrade to
465 ganeti-htools-0.3.0 (or later).
467 Regressions fixed from 2.3
468 ~~~~~~~~~~~~~~~~~~~~~~~~~~
470 - Fixed the ``gnt-cluster verify-disks`` command
471 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
473 - Fixed disk adoption breakage
474 - Fixed wrong headers in instance listing for field aliases
479 - Fixed corner case in KVM handling of NICs
480 - Fixed many cases of wrong handling of non-vm_capable nodes
481 - Fixed a bug where a missing instance symlink was not possible to
482 recreate with any ``gnt-*`` command (now ``gnt-instance
483 activate-disks`` does it)
484 - Fixed the volume group name as reported by ``gnt-cluster
486 - Increased timeouts for the import-export code, hopefully leading to
487 fewer aborts due network or instance timeouts
488 - Fixed bug in ``gnt-node list-storage``
489 - Fixed bug where not all daemons were started on cluster
490 initialisation, but only at the first watcher run
491 - Fixed many bugs in the OOB implementation
492 - Fixed watcher behaviour in presence of instances with offline
494 - Fixed instance list output for instances running on the wrong node
495 - a few fixes to the cluster-merge tool, but it still cannot merge
496 multi-node groups (currently it is not recommended to use this tool)
502 - Improved network configuration for the KVM hypervisor
503 - Added e1000 as a supported NIC for Xen-HVM
504 - Improved the lvmstrap tool to also be able to use partitions, as
505 opposed to full disks
506 - Improved speed of disk wiping (the cluster parameter
507 ``prealloc_wipe_disks``, so that it has a low impact on the total time
508 of instance creations
509 - Added documentation for the OS parameters
510 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
511 hypervisor is not responding
512 - Added display of blacklisted and hidden OS information in
514 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
515 NIC and node parameters, which might create problems with currently
516 invalid (but undetected) configuration files, but prevents validation
517 failures when unrelated parameters are modified
518 - Changed cluster initialisation to wait for the master daemon to become
520 - Expanded the RAPI interface:
522 - Added config redistribution resource
523 - Added activation/deactivation of instance disks
524 - Added export of console information
526 - Implemented log file reopening on SIGHUP, which allows using
527 logrotate(8) for the Ganeti log files
528 - Added a basic OOB helper script as an example
534 *(Released Fri, 14 Jan 2011)*
539 - Fixed timezone issues when formatting timestamps
540 - Added support for node groups, available via ``gnt-group`` and other
542 - Added out-of-band framework and management, see :doc:`design
543 document <design-oob>`
544 - Removed support for roman numbers from ``gnt-node list`` and
545 ``gnt-instance list``.
546 - Allowed modification of master network interface via ``gnt-cluster
547 modify --master-netdev``
548 - Accept offline secondaries while shutting down instance disks
549 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
550 - Added support for multiple LVM volume groups
551 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
553 - Added commands to list available fields:
555 - ``gnt-node list-fields``
556 - ``gnt-group list-fields``
557 - ``gnt-instance list-fields``
559 - Updated documentation and man pages
564 - Moved ``rapi_users`` file into separate directory, now named
565 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
567 - Added new tool for running commands on many machines,
568 ``tools/ganeti-listrunner``
569 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
570 improving the ``gnt-instance console`` output
571 - Allowed customisation of disk index separator at ``configure`` time
572 - Export node group allocation policy to :doc:`iallocator <iallocator>`
573 - Added support for non-partitioned md disks in ``lvmstrap``
574 - Added script to gracefully power off KVM instances
575 - Split ``utils`` module into smaller parts
576 - Changed query operations to return more detailed information, e.g.
577 whether an information is unavailable due to an offline node. To use
578 this new functionality, the LUXI call ``Query`` must be used. Field
579 information is now stored by the master daemon and can be retrieved
580 using ``QueryFields``. Instances, nodes and groups can also be queried
581 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
582 exposed via RAPI). The following commands make use of this
583 infrastructure change:
586 - ``gnt-group list-fields``
588 - ``gnt-node list-fields``
589 - ``gnt-instance list``
590 - ``gnt-instance list-fields``
591 - ``gnt-debug locks``
596 - New RAPI resources (see :doc:`rapi`):
600 - ``/2/groups/[group_name]``
601 - ``/2/groups/[group_name]/assign-nodes``
602 - ``/2/groups/[group_name]/modify``
603 - ``/2/groups/[group_name]/rename``
604 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
608 - Implemented ``no_install`` for instance creation
609 - Implemented OS parameters for instance reinstallation, allowing
610 use of special settings on reinstallation (e.g. for preserving data)
615 - Added IPv6 support in import/export
616 - Pause DRBD synchronization while wiping disks on instance creation
617 - Updated unittests and QA scripts
618 - Improved network parameters passed to KVM
619 - Converted man pages from docbook to reStructuredText
625 *(Released Mon, 20 Dec 2010)*
627 Released version 2.3.1~rc1 without any changes.
633 *(Released Wed, 1 Dec 2010)*
635 - impexpd: Disable OpenSSL compression in socat if possible (backport
636 from master, commit e90739d625b, see :doc:`installation guide
637 <install-quick>` for details)
638 - Changed unittest coverage report to exclude test scripts
639 - Added script to check version format
645 *(Released Wed, 1 Dec 2010)*
647 Released version 2.3.0~rc1 without any changes.
653 *(Released Fri, 19 Nov 2010)*
655 A number of bugfixes and documentation updates:
657 - Update ganeti-os-interface documentation
658 - Fixed a bug related to duplicate MACs or similar items which should be
660 - Fix breakage in OS state modify
661 - Reinstall instance: disallow offline secondaries (fixes bug related to
662 OS changing but reinstall failing)
663 - plus all the other fixes between 2.2.1 and 2.2.2
669 *(Released Tue, 2 Nov 2010)*
671 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
672 - Fixed master failover race with watcher
673 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
675 - Accept previously stopped instance for export with instance removal
676 - Simplify and extend the environment variables for instance OS scripts
677 - Added new node flags, ``master_capable`` and ``vm_capable``
678 - Added optional instance disk wiping prior during allocation. This is a
679 cluster-wide option and can be set/modified using
680 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
681 - Added IPv6 support, see :doc:`design document <design-2.3>` and
683 - Added a new watcher option (``--ignore-pause``)
684 - Added option to ignore offline node on instance start/stop
685 (``--ignore-offline``)
686 - Allow overriding OS parameters with ``gnt-instance reinstall``
687 - Added ability to change node's secondary IP address using ``gnt-node
689 - Implemented privilege separation for all daemons except
690 ``ganeti-noded``, see ``configure`` options
691 - Complain if an instance's disk is marked faulty in ``gnt-cluster
693 - Implemented job priorities (see ``ganeti(7)`` manpage)
694 - Ignore failures while shutting down instances during failover from
696 - Exit daemon's bootstrap process only once daemon is ready
697 - Export more information via ``LUInstanceQuery``/remote API
698 - Improved documentation, QA and unittests
699 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
700 restart if the file was created or changed
701 - Added LUXI protocol version sent with each request and response,
702 allowing detection of server/client mismatches
703 - Moved the Python scripts among gnt-* and ganeti-* into modules
704 - Moved all code related to setting up SSH to an external script,
706 - Infrastructure changes for node group support in future versions
712 *(Released Fri, 19 Nov 2010)*
714 A few small bugs fixed, and some improvements to the build system:
716 - Fix documentation regarding conversion to drbd
717 - Fix validation of parameters in cluster modify (``gnt-cluster modify
719 - Fix error handling in node modify with multiple changes
720 - Allow remote imports without checked names
726 *(Released Tue, 19 Oct 2010)*
728 - Disable SSL session ID cache in RPC client
734 *(Released Thu, 14 Oct 2010)*
736 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
737 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
739 - Fix problems with interaction between Curl and Python's HTTP server,
740 resulting in increased speed in many RPC calls
741 - Improve our release script to prevent breakage with older aclocal and
748 *(Released Thu, 7 Oct 2010)*
750 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
751 value retrieved from master
752 - Added support for blacklisted or hidden OS definitions
753 - Added simple lock monitor (accessible via (``gnt-debug locks``)
754 - Added support for -mem-path in KVM hypervisor abstraction layer
755 - Allow overriding instance parameters in tool for inter-cluster
756 instance moves (``tools/move-instance``)
757 - Improved opcode summaries (e.g. in ``gnt-job list``)
758 - Improve consistency of OS listing by sorting it
759 - Documentation updates
765 *(Released Fri, 8 Oct 2010)*
767 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
773 *(Released Mon, 4 Oct 2010)*
775 - Fixed regression in ``gnt-instance rename``
781 *(Released Wed, 22 Sep 2010)*
783 - Fixed OS_VARIANT variable for OS scripts
784 - Fixed cluster tag operations via RAPI
785 - Made ``setup-ssh`` exit with non-zero code if an error occurred
786 - Disabled RAPI CA checks in watcher
792 *(Released Mon, 23 Aug 2010)*
794 - Support DRBD versions of the format "a.b.c.d"
796 - Re-introduce support for usage from multiple threads in RAPI client
797 - Instance renames and modify via RAPI
798 - Work around race condition between processing and archival in job
800 - Mark opcodes following failed one as failed, too
801 - Job field ``lock_status`` was removed due to difficulties making it
802 work with the changed job queue in Ganeti 2.2; a better way to monitor
803 locks is expected for a later 2.2.x release
804 - Fixed dry-run behaviour with many commands
805 - Support ``ssh-agent`` again when adding nodes
806 - Many additional bugfixes
812 *(Released Fri, 30 Jul 2010)*
814 Important change: the internal RPC mechanism between Ganeti nodes has
815 changed from using a home-grown http library (based on the Python base
816 libraries) to use the PycURL library. This requires that PycURL is
817 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
818 against GnuTLS by default. cURL's support for GnuTLS had known issues
819 before cURL 7.21.0 and we recommend using the latest cURL release or
820 linking against OpenSSL. Most other distributions already link PycURL
821 and cURL against OpenSSL. The command::
823 python -c 'import pycurl; print pycurl.version'
825 can be used to determine the libraries PycURL and cURL are linked
828 Other significant changes:
830 - Rewrote much of the internals of the job queue, in order to achieve
831 better parallelism; this decouples job query operations from the job
832 processing, and it should allow much nicer behaviour of the master
833 daemon under load, and it also has uncovered some long-standing bugs
834 related to the job serialisation (now fixed)
835 - Added a default iallocator setting to the cluster parameters,
836 eliminating the need to always pass nodes or an iallocator for
837 operations that require selection of new node(s)
838 - Added experimental support for the LXC virtualization method
839 - Added support for OS parameters, which allows the installation of
840 instances to pass parameter to OS scripts in order to customise the
842 - Added a hypervisor parameter controlling the migration type (live or
843 non-live), since hypervisors have various levels of reliability; this
844 has renamed the 'live' parameter to 'mode'
845 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
846 logical volumes, meaning that cluster verify will ignore them and not
847 flag their presence as errors
848 - The watcher will now reset the error count for failed instances after
849 8 hours, thus allowing self-healing if the problem that caused the
850 instances to be down/fail to start has cleared in the meantime
851 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
852 check for, and warn, if the drbd module parameter ``usermode_helper``
853 is not consistent with the cluster-wide setting; this is needed to
854 make diagnose easier of failed drbd creations
855 - Started adding base IPv6 support, but this is not yet
856 enabled/available for use
857 - Rename operations (cluster, instance) will now return the new name,
858 which is especially useful if a short name was passed in
859 - Added support for instance migration in RAPI
860 - Added a tool to pre-configure nodes for the SSH setup, before joining
861 them to the cluster; this will allow in the future a simplified model
862 for node joining (but not yet fully enabled in 2.2); this needs the
863 paramiko python library
864 - Fixed handling of name-resolving errors
865 - Fixed consistency of job results on the error path
866 - Fixed master-failover race condition when executed multiple times in
868 - Fixed many bugs related to the job queue (mostly introduced during the
869 2.2 development cycle, so not all are impacting 2.1)
870 - Fixed instance migration with missing disk symlinks
871 - Fixed handling of unknown jobs in ``gnt-job archive``
872 - And many other small fixes/improvements
876 - Enhanced both the unittest and the QA coverage
877 - Switched the opcode validation to a generic model, and extended the
878 validation to all opcode parameters
879 - Changed more parts of the code that write shell scripts to use the
881 - Switched the master daemon to use the asyncore library for the Luxi
888 *(Released Thu, 17 Jun 2010)*
890 - Added tool (``move-instance``) and infrastructure to move instances
891 between separate clusters (see :doc:`separate documentation
892 <move-instance>` and :doc:`design document <design-2.2>`)
893 - Added per-request RPC timeout
894 - RAPI now requires a Content-Type header for requests with a body (e.g.
895 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
896 :rfc:`2616` (HTTP/1.1), section 7.2.1)
897 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
899 - Implemented initial support for running Ganeti daemons as separate
900 users, see configure-time flags ``--with-user-prefix`` and
901 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
903 - Instances can be removed after export (``gnt-backup export
905 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
906 key (instead of 1024 bit)
907 - Added new cluster configuration file for cluster domain secret
908 - Import/export now use SSL instead of SSH
909 - Added support for showing estimated time when exporting an instance,
910 see the ``ganeti-os-interface(7)`` manpage and look for
917 *(Released Tue, 16 Nov 2010)*
919 Some more bugfixes. Unless critical bugs occur, this will be the last
922 - Fix case of MAC special-values
923 - Fix mac checker regex
924 - backend: Fix typo causing "out of range" error
925 - Add missing --units in gnt-instance list man page
931 *(Released Tue, 24 Aug 2010)*
934 - Don't ignore secondary node silently on non-mirrored disk templates
936 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
937 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
938 - Properly document the usb_mouse parameter
939 - Fix path in ganeti-rapi(8) (issue 116)
940 - Adjust error message when the ganeti user's .ssh directory is
942 - Add same-node-check when changing the disk template to drbd
948 *(Released Fri, 16 Jul 2010)*
951 - Add an option to only select some reboot types during qa/burnin.
952 (on some hypervisors consequent reboots are not supported)
953 - Fix infrequent race condition in master failover. Sometimes the old
954 master ip address would be still detected as up for a short time
955 after it was removed, causing failover to fail.
956 - Decrease mlockall warnings when the ctypes module is missing. On
957 Python 2.4 we support running even if no ctypes module is installed,
958 but we were too verbose about this issue.
959 - Fix building on old distributions, on which man doesn't have a
961 - Fix RAPI not to ignore the MAC address on instance creation
962 - Implement the old instance creation format in the RAPI client.
968 *(Released Thu, 01 Jul 2010)*
970 A small bugfix release:
971 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
972 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
973 a non-existing option
974 - Fix instance create: the --force-variant option was ignored
975 - Improve pylint 0.21 compatibility and warnings with Python 2.6
976 - Fix modify node storage with non-FQDN arguments
977 - Fix RAPI client to authenticate under Python 2.6 when used
978 for more than 5 requests needing authentication
979 - Fix gnt-instance modify -t (storage) giving a wrong error message
980 when converting a non-shutdown drbd instance to plain
986 *(Released Fri, 18 Jun 2010)*
988 A small bugfix release:
990 - Fix live migration of KVM instances started with older Ganeti
991 versions which had fewer hypervisor parameters
992 - Fix gnt-instance grow-disk on down instances
993 - Fix an error-reporting bug during instance migration
994 - Better checking of the ``--net`` and ``--disk`` values, to avoid
995 silently ignoring broken ones
996 - Fix an RPC error reporting bug affecting, for example, RAPI client
998 - Fix bug triggered by different API version os-es on different nodes
999 - Fix a bug in instance startup with custom hvparams: OS level
1000 parameters would fail to be applied.
1001 - Fix the RAPI client under Python 2.6 (but more work is needed to
1002 make it work completely well with OpenSSL)
1003 - Fix handling of errors when resolving names from DNS
1009 *(Released Thu, 3 Jun 2010)*
1011 A medium sized development cycle. Some new features, and some
1012 fixes/small improvements/cleanups.
1014 Significant features
1015 ~~~~~~~~~~~~~~~~~~~~
1017 The node deamon now tries to mlock itself into memory, unless the
1018 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
1019 its logs, and falls back to console logging. This allows emergency
1020 features such as ``gnt-node powercycle`` to work even in the event of a
1021 broken node disk (tested offlining the disk hosting the node's
1022 filesystem and dropping its memory caches; don't try this at home)
1024 KVM: add vhost-net acceleration support. It can be tested with a new
1025 enough version of the kernel and of qemu-kvm.
1027 KVM: Add instance chrooting feature. If you use privilege dropping for
1028 your VMs you can also now force them to chroot to an empty directory,
1029 before starting the emulated guest.
1031 KVM: Add maximum migration bandwith and maximum downtime tweaking
1032 support (requires a new-enough version of qemu-kvm).
1034 Cluster verify will now warn if the master node doesn't have the master
1035 ip configured on it.
1037 Add a new (incompatible) instance creation request format to RAPI which
1038 supports all parameters (previously only a subset was supported, and it
1039 wasn't possible to extend the old format to accomodate all the new
1040 features. The old format is still supported, and a client can check for
1041 this feature, before using it, by checking for its presence in the
1042 ``features`` RAPI resource.
1044 Now with ancient latin support. Try it passing the ``--roman`` option to
1045 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
1046 (requires the python-roman module to be installed, in order to work).
1051 As usual many internal code refactorings, documentation updates, and
1054 - Lots of improvements and cleanups to the experimental Remote API
1055 (RAPI) client library.
1056 - A new unit test suite for the core daemon libraries.
1057 - A fix to creating missing directories makes sure the umask is not
1058 applied anymore. This enforces the same directory permissions
1060 - Better handling terminating daemons with ctrl+c (used when running
1061 them in debugging mode).
1062 - Fix a race condition in live migrating a KVM instance, when stat()
1063 on the old proc status file returned EINVAL, which is an unexpected
1065 - Fixed manpage checking with newer man and utf-8 charachters. But now
1066 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
1072 *(Released Fri, 7 May 2010)*
1074 Fix a bug which prevented untagged KVM instances from starting.
1080 *(Released Fri, 7 May 2010)*
1082 Another release with a long development cycle, during which many
1083 different features were added.
1085 Significant features
1086 ~~~~~~~~~~~~~~~~~~~~
1088 The KVM hypervisor now can run the individual instances as non-root, to
1089 reduce the impact of a VM being hijacked due to bugs in the
1090 hypervisor. It is possible to run all instances as a single (non-root)
1091 user, to manually specify a user for each instance, or to dynamically
1092 allocate a user out of a cluster-wide pool to each instance, with the
1093 guarantee that no two instances will run under the same user ID on any
1096 An experimental RAPI client library, that can be used standalone
1097 (without the other Ganeti libraries), is provided in the source tree as
1098 ``lib/rapi/client.py``. Note this client might change its interface in
1099 the future, as we iterate on its capabilities.
1101 A new command, ``gnt-cluster renew-crypto`` has been added to easily
1102 replace the cluster's certificates and crypto keys. This might help in
1103 case they have been compromised, or have simply expired.
1105 A new disk option for instance creation has been added that allows one
1106 to "adopt" currently existing logical volumes, with data
1107 preservation. This should allow easier migration to Ganeti from
1108 unmanaged (or managed via other software) instances.
1110 Another disk improvement is the possibility to convert between redundant
1111 (DRBD) and plain (LVM) disk configuration for an instance. This should
1112 allow better scalability (starting with one node and growing the
1113 cluster, or shrinking a two-node cluster to one node).
1115 A new feature that could help with automated node failovers has been
1116 implemented: if a node sees itself as offline (by querying the master
1117 candidates), it will try to shutdown (hard) all instances and any active
1118 DRBD devices. This reduces the risk of duplicate instances if an
1119 external script automatically failovers the instances on such nodes. To
1120 enable this, the cluster parameter ``maintain_node_health`` should be
1121 enabled; in the future this option (per the name) will enable other
1122 automatic maintenance features.
1124 Instance export/import now will reuse the original instance
1125 specifications for all parameters; that means exporting an instance,
1126 deleting it and the importing it back should give an almost identical
1127 instance. Note that the default import behaviour has changed from
1128 before, where it created only one NIC; now it recreates the original
1131 Cluster verify has added a few new checks: SSL certificates validity,
1132 /etc/hosts consistency across the cluster, etc.
1137 As usual, many internal changes were done, documentation fixes,
1140 - Fixed cluster initialization with disabled cluster storage (regression
1141 introduced in 2.1.1)
1142 - File-based storage supports growing the disks
1143 - Fixed behaviour of node role changes
1144 - Fixed cluster verify for some corner cases, plus a general rewrite of
1145 cluster verify to allow future extension with more checks
1146 - Fixed log spamming by watcher and node daemon (regression introduced
1148 - Fixed possible validation issues when changing the list of enabled
1150 - Fixed cleanup of /etc/hosts during node removal
1151 - Fixed RAPI response for invalid methods
1152 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
1153 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
1154 from ide disks, etc.)
1155 - Allow OS changes without re-installation (to record a changed OS
1156 outside of Ganeti, or to allow OS renames)
1157 - Allow instance creation without OS installation (useful for example if
1158 the OS will be installed manually, or restored from a backup not in
1160 - Implemented option to make cluster ``copyfile`` use the replication
1162 - Added list of enabled hypervisors to ssconf (possibly useful for
1164 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
1166 - A partial form of node re-IP is possible via node readd, which now
1167 allows changed node primary IP
1168 - Command line utilities now show an informational message if the job is
1170 - The logs of the master daemon now show the PID/UID/GID of the
1177 *(Released Fri, 12 Mar 2010)*
1179 During the 2.1.0 long release candidate cycle, a lot of improvements and
1180 changes have accumulated with were released later as 2.1.1.
1185 The node evacuate command (``gnt-node evacuate``) was significantly
1186 rewritten, and as such the IAllocator protocol was changed - a new
1187 request type has been added. This unfortunate change during a stable
1188 series is designed to improve performance of node evacuations; on
1189 clusters with more than about five nodes and which are well-balanced,
1190 evacuation should proceed in parallel for all instances of the node
1191 being evacuated. As such, any existing IAllocator scripts need to be
1192 updated, otherwise the above command will fail due to the unknown
1193 request. The provided "dumb" allocator has not been updated; but the
1194 ganeti-htools package supports the new protocol since version 0.2.4.
1196 Another important change is increased validation of node and instance
1197 names. This might create problems in special cases, if invalid host
1198 names are being used.
1200 Also, a new layer of hypervisor parameters has been added, that sits at
1201 OS level between the cluster defaults and the instance ones. This allows
1202 customisation of virtualization parameters depending on the installed
1203 OS. For example instances with OS 'X' may have a different KVM kernel
1204 (or any other parameter) than the cluster defaults. This is intended to
1205 help managing a multiple OSes on the same cluster, without manual
1206 modification of each instance's parameters.
1208 A tool for merging clusters, ``cluster-merge``, has been added in the
1209 tools sub-directory.
1214 - Improved the int/float conversions that should make the code more
1215 robust in face of errors from the node daemons
1216 - Fixed the remove node code in case of internal configuration errors
1217 - Fixed the node daemon behaviour in face of inconsistent queue
1218 directory (e.g. read-only file-system where we can't open the files
1220 - Fixed the behaviour of gnt-node modify for master candidate demotion;
1221 now it either aborts cleanly or, if given the new "auto_promote"
1222 parameter, will automatically promote other nodes as needed
1223 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
1224 completely prevent Ganeti from working
1225 - Fixed bug for instance export when not all disks were successfully
1227 - Fixed behaviour of node add when the new node is slow in starting up
1229 - Fixed handling of signals in the LUXI client, which should improve
1230 behaviour of command-line scripts
1231 - Added checks for invalid node/instance names in the configuration (now
1232 flagged during cluster verify)
1233 - Fixed watcher behaviour for disk activation errors
1234 - Fixed two potentially endless loops in http library, which led to the
1235 RAPI daemon hanging and consuming 100% CPU in some cases
1236 - Fixed bug in RAPI daemon related to hashed passwords
1237 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
1239 - Enhanced compatibility with non-Debian OSes, but not using absolute
1240 path in some commands and allowing customisation of the ssh
1241 configuration directory
1242 - Fixed possible future issue with new Python versions by abiding to the
1243 proper use of ``__slots__`` attribute on classes
1244 - Added checks that should prevent directory traversal attacks
1245 - Many documentation fixes based on feedback from users
1250 - Added an "early_release" more for instance replace disks and node
1251 evacuate, where we release locks earlier and thus allow higher
1252 parallelism within the cluster
1253 - Added watcher hooks, intended to allow the watcher to restart other
1254 daemons (e.g. from the ganeti-nbma project), but they can be used of
1255 course for any other purpose
1256 - Added a compile-time disable for DRBD barriers, to increase
1257 performance if the administrator trusts the power supply or the
1258 storage system to not lose writes
1259 - Added the option of using syslog for logging instead of, or in
1260 addition to, Ganeti's own log files
1261 - Removed boot restriction for paravirtual NICs for KVM, recent versions
1262 can indeed boot from a paravirtual NIC
1263 - Added a generic debug level for many operations; while this is not
1264 used widely yet, it allows one to pass the debug value all the way to
1266 - Enhanced the hooks environment for instance moves (failovers,
1267 migrations) where the primary/secondary nodes changed during the
1268 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
1269 - Enhanced data validations for many user-supplied values; one important
1270 item is the restrictions imposed on instance and node names, which
1271 might reject some (invalid) host names
1272 - Add a configure-time option to disable file-based storage, if it's not
1273 needed; this allows greater security separation between the master
1274 node and the other nodes from the point of view of the inter-node RPC
1276 - Added user notification in interactive tools if job is waiting in the
1277 job queue or trying to acquire locks
1278 - Added log messages when a job is waiting for locks
1279 - Added filtering by node tags in instance operations which admit
1280 multiple instances (start, stop, reboot, reinstall)
1281 - Added a new tool for cluster mergers, ``cluster-merge``
1282 - Parameters from command line which are of the form ``a=b,c=d`` can now
1283 use backslash escapes to pass in values which contain commas,
1284 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
1286 - For KVM, the instance name is the first parameter passed to KVM, so
1287 that it's more visible in the process list
1293 *(Released Tue, 2 Mar 2010)*
1295 Ganeti 2.1 brings many improvements with it. Major changes:
1297 - Added infrastructure to ease automated disk repairs
1298 - Added new daemon to export configuration data in a cheaper way than
1299 using the remote API
1300 - Instance NICs can now be routed instead of being associated with a
1302 - Improved job locking logic to reduce impact of jobs acquiring multiple
1303 locks waiting for other long-running jobs
1305 In-depth implementation details can be found in the Ganeti 2.1 design
1311 - Added chroot hypervisor
1312 - Added more options to xen-hvm hypervisor (``kernel_path`` and
1314 - Added more options to xen-pvm hypervisor (``use_bootloader``,
1315 ``bootloader_path`` and ``bootloader_args``)
1316 - Added the ``use_localtime`` option for the xen-hvm and kvm
1317 hypervisors, and the default value for this has changed to false (in
1318 2.0 xen-hvm always enabled it)
1319 - Added luxi call to submit multiple jobs in one go
1320 - Added cluster initialization option to not modify ``/etc/hosts``
1322 - Added network interface parameters
1323 - Added dry run mode to some LUs
1324 - Added RAPI resources:
1326 - ``/2/instances/[instance_name]/info``
1327 - ``/2/instances/[instance_name]/replace-disks``
1328 - ``/2/nodes/[node_name]/evacuate``
1329 - ``/2/nodes/[node_name]/migrate``
1330 - ``/2/nodes/[node_name]/role``
1331 - ``/2/nodes/[node_name]/storage``
1332 - ``/2/nodes/[node_name]/storage/modify``
1333 - ``/2/nodes/[node_name]/storage/repair``
1335 - Added OpCodes to evacuate or migrate all instances on a node
1336 - Added new command to list storage elements on nodes (``gnt-node
1337 list-storage``) and modify them (``gnt-node modify-storage``)
1338 - Added new ssconf files with master candidate IP address
1339 (``ssconf_master_candidates_ips``), node primary IP address
1340 (``ssconf_node_primary_ips``) and node secondary IP address
1341 (``ssconf_node_secondary_ips``)
1342 - Added ``ganeti-confd`` and a client library to query the Ganeti
1343 configuration via UDP
1344 - Added ability to run hooks after cluster initialization and before
1346 - Added automatic mode for disk replace (``gnt-instance replace-disks
1348 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
1349 after catastrophic data-loss
1350 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
1352 - Added ``gnt-instance move`` command to move instances
1353 - Added ``gnt-cluster watcher`` command to control watcher
1354 - Added ``gnt-node powercycle`` command to powercycle nodes
1355 - Added new job status field ``lock_status``
1356 - Added parseable error codes to cluster verification (``gnt-cluster
1357 verify --error-codes``) and made output less verbose (use
1358 ``--verbose`` to restore previous behaviour)
1359 - Added UUIDs to the main config entities (cluster, nodes, instances)
1360 - Added support for OS variants
1361 - Added support for hashed passwords in the Ganeti remote API users file
1363 - Added option to specify maximum timeout on instance shutdown
1364 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
1365 - Added new helper script to start and stop Ganeti daemons
1366 (``daemon-util``), with the intent to reduce the work necessary to
1367 adjust Ganeti for non-Debian distributions and to start/stop daemons
1369 - Added more unittests
1370 - Fixed critical bug in ganeti-masterd startup
1371 - Removed the configure-time ``kvm-migration-port`` parameter, this is
1372 now customisable at the cluster level for both the KVM and Xen
1373 hypervisors using the new ``migration_port`` parameter
1374 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
1375 reinstalling an instance
1376 - Allowed ``@`` in tag names
1377 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
1378 - Many documentation updates
1379 - Distribute hypervisor files on ``gnt-cluster redist-conf``
1380 - ``gnt-instance reinstall`` can now reinstall multiple instances
1381 - Updated many command line parameters
1382 - Introduced new OS API version 15
1383 - No longer support a default hypervisor
1384 - Treat virtual LVs as inexistent
1385 - Improved job locking logic to reduce lock contention
1386 - Match instance and node names case insensitively
1387 - Reimplemented bash completion script to be more complete
1394 *(Released Thu, 4 Feb 2010)*
1396 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
1397 - Fix a string formatting bug
1398 - Improve safety of the code in some error paths
1399 - Improve data validation in the master of values returned from nodes
1405 *(Released Thu, 17 Dec 2009)*
1407 - Fix security issue due to missing validation of iallocator names; this
1408 allows local and remote execution of arbitrary executables
1409 - Fix failure of gnt-node list during instance removal
1410 - Ship the RAPI documentation in the archive
1416 *(Released Wed, 30 Sep 2009)*
1418 - Fixed many wrong messages
1419 - Fixed a few bugs related to the locking library
1420 - Fixed MAC checking at instance creation time
1421 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
1422 - Fixed a few issues related to signal handling in both daemons and
1424 - Fixed the example startup script provided
1425 - Fixed insserv dependencies in the example startup script (patch from
1427 - Fixed handling of drained nodes in the iallocator framework
1428 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
1430 - Fixed error related to invalid job IDs in job polling
1431 - Fixed job/opcode persistence on unclean master shutdown
1432 - Fixed handling of partial job processing after unclean master
1434 - Fixed error reporting from LUs, previously all errors were converted
1435 into execution errors
1436 - Fixed error reporting from burnin
1437 - Decreased significantly the memory usage of the job queue
1438 - Optimised slightly multi-job submission
1439 - Optimised slightly opcode loading
1440 - Backported the multi-job submit framework from the development
1441 branch; multi-instance start and stop should be faster
1442 - Added script to clean archived jobs after 21 days; this will reduce
1443 the size of the queue directory
1444 - Added some extra checks in disk size tracking
1445 - Added an example ethers hook script
1446 - Added a cluster parameter that prevents Ganeti from modifying of
1448 - Added more node information to RAPI responses
1449 - Added a ``gnt-job watch`` command that allows following the ouput of a
1451 - Added a bind-address option to ganeti-rapi
1452 - Added more checks to the configuration verify
1453 - Enhanced the burnin script such that some operations can be retried
1455 - Converted instance reinstall to multi-instance model
1461 *(Released Fri, 7 Aug 2009)*
1463 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
1464 to allow using the pre-2.0.2 behaviour in activation, if any existing
1465 instances have mismatched disk sizes in the configuration
1466 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
1467 any configuration mismatches for disk sizes
1468 - Added ``gnt-master cluste-failover --no-voting`` to allow master
1469 failover to work on two-node clusters
1470 - Fixed the ``--net`` option of ``gnt-backup import``, which was
1472 - Fixed detection of OS script errors in ``gnt-backup export``
1473 - Fixed exit code of ``gnt-backup export``
1479 *(Released Fri, 17 Jul 2009)*
1481 - Added experimental support for stripped logical volumes; this should
1482 enhance performance but comes with a higher complexity in the block
1483 device handling; stripping is only enabled when passing
1484 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
1485 affected even in the non-stripped mode
1486 - Improved resiliency against transient failures at the end of DRBD
1487 resyncs, and in general of DRBD resync checks
1488 - Fixed a couple of issues with exports and snapshot errors
1489 - Fixed a couple of issues in instance listing
1490 - Added display of the disk size in ``gnt-instance info``
1491 - Fixed checking for valid OSes in instance creation
1492 - Fixed handling of the "vcpus" parameter in instance listing and in
1493 general of invalid parameters
1494 - Fixed http server library, and thus RAPI, to handle invalid
1495 username/password combinations correctly; this means that now they
1496 report unauthorized for queries too, not only for modifications,
1497 allowing earlier detect of configuration problems
1498 - Added a new "role" node list field, equivalent to the master/master
1499 candidate/drained/offline flags combinations
1500 - Fixed cluster modify and changes of candidate pool size
1501 - Fixed cluster verify error messages for wrong files on regular nodes
1502 - Fixed a couple of issues with node demotion from master candidate role
1503 - Fixed node readd issues
1504 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
1505 - Added a new ``--no-voting`` option for masterfailover to fix failover
1506 on two-nodes clusters when the former master node is unreachable
1507 - Added instance reinstall over RAPI
1513 *(Released Tue, 16 Jun 2009)*
1515 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
1516 allow re-adding the start in single-user option (regression from 1.2)
1517 - the watcher writes the instance status to a file, to allow monitoring
1518 to report the instance status (from the master) based on cached
1519 results of the watcher's queries; while this can get stale if the
1520 watcher is being locked due to other work on the cluster, this is
1521 still an improvement
1522 - the watcher now also restarts the node daemon and the rapi daemon if
1524 - fixed the watcher to handle full and drained queue cases
1525 - hooks export more instance data in the environment, which helps if
1526 hook scripts need to take action based on the instance's properties
1527 (no longer need to query back into ganeti)
1528 - instance failovers when the instance is stopped do not check for free
1529 RAM, so that failing over a stopped instance is possible in low memory
1531 - rapi uses queries for tags instead of jobs (for less job traffic), and
1532 for cluster tags it won't talk to masterd at all but read them from
1534 - a couple of error handling fixes in RAPI
1535 - drbd handling: improved the error handling of inconsistent disks after
1536 resync to reduce the frequency of "there are some degraded disks for
1537 this instance" messages
1538 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
1539 error handling path called a wrong function name)
1545 *(Released Wed, 27 May 2009)*
1547 - no changes from rc5
1553 *(Released Wed, 20 May 2009)*
1555 - fix a couple of bugs (validation, argument checks)
1556 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
1557 - some small improvements to RAPI and IAllocator
1558 - make watcher automatically start the master daemon if down
1564 *(Released Mon, 27 Apr 2009)*
1566 - change the OS list to not require locks; this helps with big clusters
1567 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
1568 volume group is broken
1569 - ``gnt-instance info``, without any arguments, doesn't run for all
1570 instances anymore; either pass ``--all`` or pass the desired
1571 instances; this helps against mistakes on big clusters where listing
1572 the information for all instances takes a long time
1573 - miscellaneous doc and man pages fixes
1579 *(Released Wed, 8 Apr 2009)*
1581 - Change the internal locking model of some ``gnt-node`` commands, in
1582 order to reduce contention (and blocking of master daemon) when
1583 batching many creation/reinstall jobs
1584 - Fixes to Xen soft reboot
1585 - No longer build documentation at build time, instead distribute it in
1586 the archive, in order to reduce the need for the whole docbook/rst
1593 *(Released Fri, 27 Mar 2009)*
1595 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
1596 - Fix watcher startup sequence, improves the behaviour of busy clusters
1597 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
1598 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
1599 bind address and other places
1600 - Some documentation fixes and updates
1606 *(Released Mon, 2 Mar 2009)*
1608 - More documentation updates, now all docs should be more-or-less
1610 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
1612 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
1614 - Fix an issue related to $libdir/run/ganeti and cluster creation
1620 *(Released Thu, 19 Feb 2009)*
1622 - Xen PVM and KVM have switched the default value for the instance root
1623 disk to the first partition on the first drive, instead of the whole
1624 drive; this means that the OS installation scripts must be changed
1626 - Man pages have been updated
1627 - RAPI has been switched by default to HTTPS, and the exported functions
1628 should all work correctly
1629 - RAPI v1 has been removed
1630 - Many improvements to the KVM hypervisor
1631 - Block device errors are now better reported
1632 - Many other bugfixes and small improvements
1638 *(Released Mon, 26 Jan 2009)*
1640 - Version 2 is a general rewrite of the code and therefore the
1641 differences are too many to list, see the design document for 2.0 in
1642 the ``doc/`` subdirectory for more details
1643 - In this beta version there is not yet a migration path from 1.2 (there
1644 will be one in the final 2.0 release)
1645 - A few significant changes are:
1647 - all commands are executed by a daemon (``ganeti-masterd``) and the
1648 various ``gnt-*`` commands are just front-ends to it
1649 - all the commands are entered into, and executed from a job queue,
1650 see the ``gnt-job(8)`` manpage
1651 - the RAPI daemon supports read-write operations, secured by basic
1652 HTTP authentication on top of HTTPS
1653 - DRBD version 0.7 support has been removed, DRBD 8 is the only
1654 supported version (when migrating from Ganeti 1.2 to 2.0, you need
1655 to migrate to DRBD 8 first while still running Ganeti 1.2)
1656 - DRBD devices are using statically allocated minor numbers, which
1657 will be assigned to existing instances during the migration process
1658 - there is support for both Xen PVM and Xen HVM instances running on
1660 - KVM virtualization is supported too
1661 - file-based storage has been implemented, which means that it is
1662 possible to run the cluster without LVM and DRBD storage, for
1663 example using a shared filesystem exported from shared storage (and
1664 still have live migration)
1670 *(Released Tue, 13 Jan 2009)*
1672 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
1673 - Reuse the old instance mac address by default on instance import, if
1674 the instance name is the same.
1675 - Handle situations in which the node info rpc returns incomplete
1677 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
1678 - Improved version of batcher:
1680 - state file support
1681 - instance mac address support
1682 - support for HVM clusters/instances
1684 - Add an option to show the number of cpu sockets and nodes in
1686 - Support OSes that handle more than one version of the OS api (but do
1687 not change the current API in any other way)
1688 - Fix ``gnt-node migrate``
1689 - ``gnt-debug`` man page
1690 - Fixes various more typos and small issues
1691 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
1697 *(Released Wed, 24 Sep 2008)*
1699 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
1700 type of disk exported to fully virtualized instances.
1701 - provide access to the serial console of HVM instances
1702 - instance auto_balance flag, set by default. If turned off it will
1703 avoid warnings on cluster verify if there is not enough memory to fail
1704 over an instance. in the future it will prevent automatically failing
1705 it over when we will support that.
1706 - batcher tool for instance creation, see ``tools/README.batcher``
1707 - ``gnt-instance reinstall --select-os`` to interactively select a new
1708 operating system when reinstalling an instance.
1709 - when changing the memory amount on instance modify a check has been
1710 added that the instance will be able to start. also warnings are
1711 emitted if the instance will not be able to fail over, if auto_balance
1713 - documentation fixes
1714 - sync fields between ``gnt-instance list/modify/add/import``
1715 - fix a race condition in drbd when the sync speed was set after giving
1716 the device a remote peer.
1722 *(Released Tue, 22 Jul 2008)*
1724 - note: the allowed size and number of tags per object were reduced
1725 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
1726 - fixed twisted 8.x compatibility
1727 - fixed ``gnt-instance replace-disks`` with iallocator
1728 - add TCP keepalives on twisted connections to detect restarted nodes
1729 - disk increase support, see ``gnt-instance grow-disk``
1730 - implement bulk node/instance query for RAPI
1731 - add tags in node/instance listing (optional)
1732 - experimental migration (and live migration) support, read the man page
1733 for ``gnt-instance migrate``
1734 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
1735 has some small improvements in handling its state file
1741 *(Released Fri, 13 Jun 2008)*
1743 - Experimental readonly, REST-based remote API implementation;
1744 automatically started on master node, TCP port 5080, if enabled by
1745 ``--enable-rapi`` parameter to configure script.
1746 - Instance allocator support. Add and import instance accept a
1747 ``--iallocator`` parameter, and call that instance allocator to decide
1748 which node to use for the instance. The iallocator document describes
1749 what's expected from an allocator script.
1750 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
1751 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
1752 node is lost there is still enough memory to fail over the instances
1754 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
1755 ``gnt-cluster verify``, to check for site-specific compliance. All the
1756 hooks will run, and their output, if any, will be displayed. Any
1757 failing hook will make the verification return an error value.
1758 - ``gnt-cluster verify`` now checks that its peers are reachable on the
1759 primary and secondary interfaces
1760 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
1761 that is still declared as part of the cluster and has failed.
1762 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
1763 specifying output fields, that just adds the chosen fields to the
1765 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
1766 export from the filesystem.
1767 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
1768 have been added. Using them you can enable/disable acpi and pae
1769 support, and specify a path for a cd image to be exported to the
1770 instance. These parameters as the name suggest only work on HVM
1772 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
1773 PAE support will be set to the previously hardcoded values, but the
1774 (previously hardcoded) path to the CDROM ISO image will be unset and
1775 if required, needs to be set manually with ``gnt-instance modify``
1777 - The address to which an instance's VNC console is bound is now
1778 selectable per-instance, rather than being cluster wide. Of course
1779 this only applies to instances controlled via VNC, so currently just
1780 applies to HVM clusters.
1786 *(Released Mon, 18 Feb 2008)*
1788 - more tweaks to the disk activation code (especially helpful for DRBD)
1789 - change the default ``gnt-instance list`` output format, now there is
1790 one combined status field (see the manpage for the exact values this
1792 - some more fixes for the mac export to hooks change
1793 - make Ganeti not break with DRBD 8.2.x (which changed the version
1794 format in ``/proc/drbd``) (issue 24)
1795 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
1796 template, allowing migration from DRBD0.7+MD to DRBD8
1802 *(Released Wed, 30 Jan 2008)*
1804 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
1806 - add command aliases infrastructure and a few aliases
1807 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
1808 man pages and the ``--help`` option of ``gnt-node
1809 list``/``gnt-instance list``
1810 - fix ``gnt-backup list`` with down nodes (issue 21)
1811 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
1812 - fix the dist archive and add a check for including svn/git files in
1814 - some developer-related changes: improve the burnin and the QA suite,
1815 add an upload script for testing during development
1821 *(Released Wed, 16 Jan 2008)*
1823 - experimental HVM support, read the install document, section
1824 "Initializing the cluster"
1825 - allow for the PVM hypervisor per-instance kernel and initrd paths
1826 - add a new command ``gnt-cluster verify-disks`` which uses a new
1827 algorithm to improve the reconnection of the DRBD pairs if the device
1828 on the secondary node has gone away
1829 - make logical volume code auto-activate LVs at disk activation time
1830 - slightly improve the speed of activating disks
1831 - allow specification of the MAC address at instance creation time, and
1832 changing it later via ``gnt-instance modify``
1833 - fix handling of external commands that generate lots of output on
1835 - update documentation with regard to minimum version of DRBD8 supported
1841 *(Released Tue, 4 Dec 2007)*
1843 - Log the ``xm create`` output to the node daemon log on failure (to
1844 help diagnosing the error)
1845 - In debug mode, log all external commands output if failed to the logs
1846 - Change parsing of lvm commands to ignore stderr
1852 *(Released Wed, 28 Nov 2007)*
1854 - Another round of updates to the DRBD 8 code to deal with more failures
1855 in the replace secondary node operation
1856 - Some more logging of failures in disk operations (lvm, drbd)
1857 - A few documentation updates
1864 *(Released Tue, 13 Nov 2007)*
1866 - Change configuration file format from Python's Pickle to JSON.
1867 Upgrading is possible using the cfgupgrade utility.
1868 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
1869 faster replace disks and is more stable (DRBD 8 has many improvements
1870 compared to DRBD 0.7)
1871 - Added command line tags support (see man pages for ``gnt-instance``,
1872 ``gnt-node``, ``gnt-cluster``)
1873 - Added instance rename support
1874 - Added multi-instance startup/shutdown
1875 - Added cluster rename support
1876 - Added ``gnt-node evacuate`` to simplify some node operations
1877 - Added instance reboot operation that can speedup reboot as compared to
1879 - Soften the requirement that hostnames are in FQDN format
1880 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
1882 - Removed dependency on debian's patched fping that uses the
1883 non-standard ``-S`` option
1884 - Now the OS definitions are searched for in multiple, configurable
1885 paths (easier for distros to package)
1886 - Some changes to the hooks infrastructure (especially the new
1887 post-configuration update hook)
1888 - Other small bugfixes
1890 .. vim: set textwidth=72 syntax=rst :