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.
22 - CertificateError now derives from GanetiApiError in the RAPI client.
28 *(Released Thu, 12 Apr 2012)*
30 Incompatible/important changes and bugfixes
31 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
33 - The default of the ``/2/instances/[instance_name]/rename`` RAPI
34 resource's ``ip_check`` parameter changed from ``True`` to ``False``
35 to match the underlying LUXI interface.
36 - The ``/2/nodes/[node_name]/evacuate`` RAPI resource was changed to use
37 body parameters, see :doc:`RAPI documentation <rapi>`. The server does
38 not maintain backwards-compatibility as the underlying operation
39 changed in an incompatible way. The RAPI client can talk to old
40 servers, but it needs to be told so as the return value changed.
41 - When creating file-based instances via RAPI, the ``file_driver``
42 parameter no longer defaults to ``loop`` and must be specified.
43 - The deprecated ``bridge`` NIC parameter is no longer supported. Use
45 - Support for the undocumented and deprecated RAPI instance creation
46 request format version 0 has been dropped. Use version 1, supported
47 since Ganeti 2.1.3 and :doc:`documented <rapi>`, instead.
48 - Pyparsing 1.4.6 or above is required, see :doc:`installation
49 documentation <install>`.
50 - The "cluster-verify" hooks are now executed per group by the
51 ``OP_CLUSTER_VERIFY_GROUP`` opcode. This maintains the same behavior
52 if you just run ``gnt-cluster verify``, which generates one opcode per
54 - The environment as passed to the OS scripts is cleared, and thus no
55 environment variables defined in the node daemon's environment will be
56 inherited by the scripts.
57 - The :doc:`iallocator <iallocator>` mode ``multi-evacuate`` has been
59 - :doc:`New iallocator modes <design-multi-reloc>` have been added to
60 support operations involving multiple node groups.
61 - Offline nodes are ignored when failing over an instance.
62 - Support for KVM version 1.0, which changed the version reporting format
64 - TCP/IP ports used by DRBD disks are returned to a pool upon instance
66 - ``Makefile`` is now compatible with Automake 1.11.2
67 - Includes all bugfixes made in the 2.4 series
72 - The ganeti-htools project has been merged into the ganeti-core source
73 tree and will be built as part of Ganeti (see :doc:`install-quick`).
74 - Implemented support for :doc:`shared storage <design-shared-storage>`.
75 - Add support for disks larger than 2 TB in ``lvmstrap`` by supporting
76 GPT-style partition tables (requires `parted
77 <http://www.gnu.org/s/parted/>`_).
78 - Added support for floppy drive and 2nd CD-ROM drive in KVM hypervisor.
79 - Allowed adding tags on instance creation.
80 - Export instance tags to hooks (``INSTANCE_TAGS``, see :doc:`hooks`)
81 - Allow instances to be started in a paused state, enabling the user to
82 see the complete console output on boot using the console.
83 - Added new hypervisor flag to control default reboot behaviour
84 (``reboot_behavior``).
85 - Added support for KVM keymaps (hypervisor parameter ``keymap``).
86 - Improved out-of-band management support:
88 - Added ``gnt-node health`` command reporting the health status of
90 - Added ``gnt-node power`` command to manage power status of nodes.
91 - Added command for emergency power-off (EPO), ``gnt-cluster epo``.
93 - Instance migration can fall back to failover if instance is not
95 - Filters can be used when listing nodes, instances, groups and locks;
96 see *ganeti(7)* manpage.
97 - Added post-execution status as variables to :doc:`hooks <hooks>`
99 - Instance tags are exported/imported together with the instance.
100 - When given an explicit job ID, ``gnt-job info`` will work for archived
102 - Jobs can define dependencies on other jobs (not yet supported via
103 RAPI or command line, but used by internal commands and usable via
106 - Lock monitor (``gnt-debug locks``) shows jobs waiting for
109 - Instance failover is now available as a RAPI resource
110 (``/2/instances/[instance_name]/failover``).
111 - ``gnt-instance info`` defaults to static information if primary node
113 - Opcodes have a new ``comment`` attribute.
114 - Added basic SPICE support to KVM hypervisor.
115 - ``tools/ganeti-listrunner`` allows passing of arguments to executable.
117 Node group improvements
118 ~~~~~~~~~~~~~~~~~~~~~~~
120 - ``gnt-cluster verify`` has been modified to check groups separately,
121 thereby improving performance.
122 - Node group support has been added to ``gnt-cluster verify-disks``,
123 which now operates per node group.
124 - Watcher has been changed to work better with node groups.
126 - One process and state file per node group.
127 - Slow watcher in one group doesn't block other group's watcher.
129 - Added new command, ``gnt-group evacuate``, to move all instances in a
130 node group to other groups.
131 - Added ``gnt-instance change-group`` to move an instance to another
133 - ``gnt-cluster command`` and ``gnt-cluster copyfile`` now support
134 per-group operations.
135 - Node groups can be tagged.
136 - Some operations switch from an exclusive to a shared lock as soon as
138 - Instance's primary and secondary nodes' groups are now available as
139 query fields (``pnode.group``, ``pnode.group.uuid``, ``snodes.group``
140 and ``snodes.group.uuid``).
145 - Numerous updates to documentation and manpages.
147 - :doc:`RAPI <rapi>` documentation now has detailed parameter
149 - Some opcode/job results are now also documented, see :doc:`RAPI
152 - A lockset's internal lock is now also visible in lock monitor.
153 - Log messages from job queue workers now contain information about the
154 opcode they're processing.
155 - ``gnt-instance console`` no longer requires the instance lock.
156 - A short delay when waiting for job changes reduces the number of LUXI
157 requests significantly.
158 - DRBD metadata volumes are overwritten with zeros during disk creation.
159 - Out-of-band commands no longer acquire the cluster lock in exclusive
161 - ``devel/upload`` now uses correct permissions for directories.
167 *(Released Fri, 23 Mar 2012)*
169 This was the sixth release candidate of the 2.5 series.
175 *(Released Mon, 9 Jan 2012)*
177 This was the fifth release candidate of the 2.5 series.
183 *(Released Thu, 27 Oct 2011)*
185 This was the fourth release candidate of the 2.5 series.
191 *(Released Wed, 26 Oct 2011)*
193 This was the third release candidate of the 2.5 series.
199 *(Released Tue, 18 Oct 2011)*
201 This was the second release candidate of the 2.5 series.
207 *(Released Tue, 4 Oct 2011)*
209 This was the first release candidate of the 2.5 series.
215 *(Released Wed, 31 Aug 2011)*
217 This was the third beta release of the 2.5 series.
223 *(Released Mon, 22 Aug 2011)*
225 This was the second beta release of the 2.5 series.
231 *(Released Fri, 12 Aug 2011)*
233 This was the first beta release of the 2.5 series.
239 *(Released Thu, 27 Oct 2011)*
241 - Fixed bug when parsing command line parameter values ending in
243 - Fixed assertion error after unclean master shutdown
244 - Disable HTTP client pool for RPC, significantly reducing memory usage
246 - Fixed queue archive creation with wrong permissions
252 *(Released Tue, 23 Aug 2011)*
256 - Fixed documentation for importing with ``--src-dir`` option
257 - Fixed a bug in ``ensure-dirs`` with queue/archive permissions
258 - Fixed a parsing issue with DRBD 8.3.11 in the Linux kernel
264 *(Released Fri, 5 Aug 2011)*
266 Many bug-fixes and a few small features:
268 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
269 issues when you tried to add an instance with two MAC addresses in one
271 - KVM: fixed per-instance stored UID value
272 - KVM: configure bridged NICs at migration start
273 - KVM: Fix a bug where instance will not start with never KVM versions
275 - Added OS search path to ``gnt-cluster info``
276 - Fixed an issue with ``file_storage_dir`` where you were forced to
277 provide an absolute path, but the documentation states it is a
278 relative path, the documentation was right
279 - Added a new parameter to instance stop/start called ``--no-remember``
280 that will make the state change to not be remembered
281 - Implemented ``no_remember`` at RAPI level
282 - Improved the documentation
283 - Node evacuation: don't call IAllocator if node is already empty
284 - Fixed bug in DRBD8 replace disks on current nodes
285 - Fixed bug in recreate-disks for DRBD instances
286 - Moved assertion checking locks in ``gnt-instance replace-disks``
287 causing it to abort with not owning the right locks for some situation
288 - Job queue: Fixed potential race condition when cancelling queued jobs
289 - Fixed off-by-one bug in job serial generation
290 - ``gnt-node volumes``: Fix instance names
291 - Fixed aliases in bash completion
292 - Fixed a bug in reopening log files after being sent a SIGHUP
293 - Added a flag to burnin to allow specifying VCPU count
294 - Bugfixes to non-root Ganeti configuration
300 *(Released Thu, 12 May 2011)*
302 Many bug-fixes and a few new small features:
304 - Fixed a bug related to log opening failures
305 - Fixed a bug in instance listing with orphan instances
306 - Fixed a bug which prevented resetting the cluster-level node parameter
307 ``oob_program`` to the default
308 - Many fixes related to the ``cluster-merge`` tool
309 - Fixed a race condition in the lock monitor, which caused failures
310 during (at least) creation of many instances in parallel
311 - Improved output for gnt-job info
312 - Removed the quiet flag on some ssh calls which prevented debugging
314 - Improved the N+1 failure messages in cluster verify by actually
315 showing the memory values (needed and available)
316 - Increased lock attempt timeouts so that when executing long operations
317 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
318 too early and thus prevent the use of the 'fair' mechanism
319 - Changed instance query data (``gnt-instance info``) to not acquire
320 locks unless needed, thus allowing its use on locked instance if only
321 static information is asked for
322 - Improved behaviour with filesystems that do not support rename on an
324 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
325 kept locks on all nodes during the wipe, which is unneeded
326 - Fixed ``gnt-watcher`` handling of errors during hooks execution
327 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
328 10GiB) which caused the wipe to fail right at the end in some cases
329 - Fixed master IP activation when doing master failover with no-voting
330 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
331 the master node itself
332 - Fixed potential data-loss in under disk full conditions, where Ganeti
333 wouldn't check correctly the return code and would consider
334 partially-written files 'correct'
335 - Fixed bug related to multiple VGs and DRBD disk replacing
336 - Added new disk parameter ``metavg`` that allows placement of the meta
337 device for DRBD in a different volume group
338 - Fixed error handling in the node daemon when the system libc doesn't
339 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
340 - Fixed lock release during replace-disks, which kept cluster-wide locks
341 when doing disk replaces with an iallocator script
342 - Added check for missing bridges in cluster verify
343 - Handle EPIPE errors while writing to the terminal better, so that
344 piping the output to e.g. ``less`` doesn't cause a backtrace
345 - Fixed rare case where a ^C during Luxi calls could have been
346 interpreted as server errors, instead of simply terminating
347 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
349 - Added a few more parameters to the KVM hypervisor, allowing a second
350 CDROM, custom disk type for CDROMs and a floppy image
351 - Removed redundant message in instance rename when the name is given
353 - Added option to ``gnt-instance recreate-disks`` to allow creating the
354 disks on new nodes, allowing recreation when the original instance
355 nodes are completely gone
356 - Added option when converting disk templates to DRBD to skip waiting
357 for the resync, in order to make the instance available sooner
358 - Added two new variables to the OS scripts environment (containing the
360 - Made the root_path and optional parameter for the xen-pvm hypervisor,
361 to allow use of ``pvgrub`` as bootloader
362 - Changed the instance memory modifications to only check out-of-memory
363 conditions on memory increases, and turned the secondary node warnings
364 into errors (they can still be overridden via ``--force``)
365 - Fixed the handling of a corner case when the Python installation gets
366 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
367 to execute a command that doesn't exist
368 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
369 node of the instance returned failures during instance shutdown; this
370 adds the option ``--ignore-consistency`` to gnt-instance move
372 And as usual, various improvements to the error messages, documentation
379 *(Released Wed, 09 Mar 2011)*
381 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
382 the RAPI users file if run twice (even with ``--dry-run``).
384 The release fixes that bug (nothing else changed).
390 *(Released Mon, 07 Mar 2011)*
392 Final 2.4.0 release. Just a few small fixes:
394 - Fixed RAPI node evacuate
395 - Fixed the kvm-ifup script
396 - Fixed internal error handling for special job cases
397 - Updated man page to specify the escaping feature for options
403 *(Released Mon, 28 Feb 2011)*
405 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
406 that this feature wiped the disks of the wrong instance, leading to loss
411 - Fixed title of query field containing instance name
412 - Expanded the glossary in the documentation
413 - Fixed one unittest (internal issue)
419 *(Released Mon, 21 Feb 2011)*
421 A number of bug fixes plus just a couple functionality changes.
423 On the user-visible side, the ``gnt-* list`` command output has changed
424 with respect to "special" field states. The current rc1 style of display
425 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
426 the default output mode special fields states are displayed as follows:
428 - Offline resource: ``*``
429 - Unavailable/not applicable: ``-``
430 - Data missing (RPC failure): ``?``
431 - Unknown field: ``??``
433 Another user-visible change is the addition of ``--force-join`` to
438 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
439 - Fixed regression in RAPI/instance reinstall where all parameters were
440 required (instead of optional)
441 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
442 - Fixed iallocator usage (offline nodes were not considered offline)
443 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
444 - Fixed hypervisor and OS parameter validation with respect to
446 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
448 - Fixed ``tools/listrunner`` with respect to agent-based usage
454 *(Released Fri, 4 Feb 2011)*
456 Many changes and fixes since the beta1 release. While there were some
457 internal changes, the code has been mostly stabilised for the RC
460 Note: the dumb allocator was removed in this release, as it was not kept
461 up-to-date with the IAllocator protocol changes. It is recommended to
462 use the ``hail`` command from the ganeti-htools package.
464 Note: the 2.4 and up versions of Ganeti are not compatible with the
465 0.2.x branch of ganeti-htools. You need to upgrade to
466 ganeti-htools-0.3.0 (or later).
468 Regressions fixed from 2.3
469 ~~~~~~~~~~~~~~~~~~~~~~~~~~
471 - Fixed the ``gnt-cluster verify-disks`` command
472 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
474 - Fixed disk adoption breakage
475 - Fixed wrong headers in instance listing for field aliases
480 - Fixed corner case in KVM handling of NICs
481 - Fixed many cases of wrong handling of non-vm_capable nodes
482 - Fixed a bug where a missing instance symlink was not possible to
483 recreate with any ``gnt-*`` command (now ``gnt-instance
484 activate-disks`` does it)
485 - Fixed the volume group name as reported by ``gnt-cluster
487 - Increased timeouts for the import-export code, hopefully leading to
488 fewer aborts due network or instance timeouts
489 - Fixed bug in ``gnt-node list-storage``
490 - Fixed bug where not all daemons were started on cluster
491 initialisation, but only at the first watcher run
492 - Fixed many bugs in the OOB implementation
493 - Fixed watcher behaviour in presence of instances with offline
495 - Fixed instance list output for instances running on the wrong node
496 - a few fixes to the cluster-merge tool, but it still cannot merge
497 multi-node groups (currently it is not recommended to use this tool)
503 - Improved network configuration for the KVM hypervisor
504 - Added e1000 as a supported NIC for Xen-HVM
505 - Improved the lvmstrap tool to also be able to use partitions, as
506 opposed to full disks
507 - Improved speed of disk wiping (the cluster parameter
508 ``prealloc_wipe_disks``, so that it has a low impact on the total time
509 of instance creations
510 - Added documentation for the OS parameters
511 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
512 hypervisor is not responding
513 - Added display of blacklisted and hidden OS information in
515 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
516 NIC and node parameters, which might create problems with currently
517 invalid (but undetected) configuration files, but prevents validation
518 failures when unrelated parameters are modified
519 - Changed cluster initialisation to wait for the master daemon to become
521 - Expanded the RAPI interface:
523 - Added config redistribution resource
524 - Added activation/deactivation of instance disks
525 - Added export of console information
527 - Implemented log file reopening on SIGHUP, which allows using
528 logrotate(8) for the Ganeti log files
529 - Added a basic OOB helper script as an example
535 *(Released Fri, 14 Jan 2011)*
540 - Fixed timezone issues when formatting timestamps
541 - Added support for node groups, available via ``gnt-group`` and other
543 - Added out-of-band framework and management, see :doc:`design
544 document <design-oob>`
545 - Removed support for roman numbers from ``gnt-node list`` and
546 ``gnt-instance list``.
547 - Allowed modification of master network interface via ``gnt-cluster
548 modify --master-netdev``
549 - Accept offline secondaries while shutting down instance disks
550 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
551 - Added support for multiple LVM volume groups
552 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
554 - Added commands to list available fields:
556 - ``gnt-node list-fields``
557 - ``gnt-group list-fields``
558 - ``gnt-instance list-fields``
560 - Updated documentation and man pages
565 - Moved ``rapi_users`` file into separate directory, now named
566 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
568 - Added new tool for running commands on many machines,
569 ``tools/ganeti-listrunner``
570 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
571 improving the ``gnt-instance console`` output
572 - Allowed customisation of disk index separator at ``configure`` time
573 - Export node group allocation policy to :doc:`iallocator <iallocator>`
574 - Added support for non-partitioned md disks in ``lvmstrap``
575 - Added script to gracefully power off KVM instances
576 - Split ``utils`` module into smaller parts
577 - Changed query operations to return more detailed information, e.g.
578 whether an information is unavailable due to an offline node. To use
579 this new functionality, the LUXI call ``Query`` must be used. Field
580 information is now stored by the master daemon and can be retrieved
581 using ``QueryFields``. Instances, nodes and groups can also be queried
582 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
583 exposed via RAPI). The following commands make use of this
584 infrastructure change:
587 - ``gnt-group list-fields``
589 - ``gnt-node list-fields``
590 - ``gnt-instance list``
591 - ``gnt-instance list-fields``
592 - ``gnt-debug locks``
597 - New RAPI resources (see :doc:`rapi`):
601 - ``/2/groups/[group_name]``
602 - ``/2/groups/[group_name]/assign-nodes``
603 - ``/2/groups/[group_name]/modify``
604 - ``/2/groups/[group_name]/rename``
605 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
609 - Implemented ``no_install`` for instance creation
610 - Implemented OS parameters for instance reinstallation, allowing
611 use of special settings on reinstallation (e.g. for preserving data)
616 - Added IPv6 support in import/export
617 - Pause DRBD synchronization while wiping disks on instance creation
618 - Updated unittests and QA scripts
619 - Improved network parameters passed to KVM
620 - Converted man pages from docbook to reStructuredText
626 *(Released Mon, 20 Dec 2010)*
628 Released version 2.3.1~rc1 without any changes.
634 *(Released Wed, 1 Dec 2010)*
636 - impexpd: Disable OpenSSL compression in socat if possible (backport
637 from master, commit e90739d625b, see :doc:`installation guide
638 <install-quick>` for details)
639 - Changed unittest coverage report to exclude test scripts
640 - Added script to check version format
646 *(Released Wed, 1 Dec 2010)*
648 Released version 2.3.0~rc1 without any changes.
654 *(Released Fri, 19 Nov 2010)*
656 A number of bugfixes and documentation updates:
658 - Update ganeti-os-interface documentation
659 - Fixed a bug related to duplicate MACs or similar items which should be
661 - Fix breakage in OS state modify
662 - Reinstall instance: disallow offline secondaries (fixes bug related to
663 OS changing but reinstall failing)
664 - plus all the other fixes between 2.2.1 and 2.2.2
670 *(Released Tue, 2 Nov 2010)*
672 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
673 - Fixed master failover race with watcher
674 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
676 - Accept previously stopped instance for export with instance removal
677 - Simplify and extend the environment variables for instance OS scripts
678 - Added new node flags, ``master_capable`` and ``vm_capable``
679 - Added optional instance disk wiping prior during allocation. This is a
680 cluster-wide option and can be set/modified using
681 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
682 - Added IPv6 support, see :doc:`design document <design-2.3>` and
684 - Added a new watcher option (``--ignore-pause``)
685 - Added option to ignore offline node on instance start/stop
686 (``--ignore-offline``)
687 - Allow overriding OS parameters with ``gnt-instance reinstall``
688 - Added ability to change node's secondary IP address using ``gnt-node
690 - Implemented privilege separation for all daemons except
691 ``ganeti-noded``, see ``configure`` options
692 - Complain if an instance's disk is marked faulty in ``gnt-cluster
694 - Implemented job priorities (see ``ganeti(7)`` manpage)
695 - Ignore failures while shutting down instances during failover from
697 - Exit daemon's bootstrap process only once daemon is ready
698 - Export more information via ``LUInstanceQuery``/remote API
699 - Improved documentation, QA and unittests
700 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
701 restart if the file was created or changed
702 - Added LUXI protocol version sent with each request and response,
703 allowing detection of server/client mismatches
704 - Moved the Python scripts among gnt-* and ganeti-* into modules
705 - Moved all code related to setting up SSH to an external script,
707 - Infrastructure changes for node group support in future versions
713 *(Released Fri, 19 Nov 2010)*
715 A few small bugs fixed, and some improvements to the build system:
717 - Fix documentation regarding conversion to drbd
718 - Fix validation of parameters in cluster modify (``gnt-cluster modify
720 - Fix error handling in node modify with multiple changes
721 - Allow remote imports without checked names
727 *(Released Tue, 19 Oct 2010)*
729 - Disable SSL session ID cache in RPC client
735 *(Released Thu, 14 Oct 2010)*
737 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
738 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
740 - Fix problems with interaction between Curl and Python's HTTP server,
741 resulting in increased speed in many RPC calls
742 - Improve our release script to prevent breakage with older aclocal and
749 *(Released Thu, 7 Oct 2010)*
751 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
752 value retrieved from master
753 - Added support for blacklisted or hidden OS definitions
754 - Added simple lock monitor (accessible via (``gnt-debug locks``)
755 - Added support for -mem-path in KVM hypervisor abstraction layer
756 - Allow overriding instance parameters in tool for inter-cluster
757 instance moves (``tools/move-instance``)
758 - Improved opcode summaries (e.g. in ``gnt-job list``)
759 - Improve consistency of OS listing by sorting it
760 - Documentation updates
766 *(Released Fri, 8 Oct 2010)*
768 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
774 *(Released Mon, 4 Oct 2010)*
776 - Fixed regression in ``gnt-instance rename``
782 *(Released Wed, 22 Sep 2010)*
784 - Fixed OS_VARIANT variable for OS scripts
785 - Fixed cluster tag operations via RAPI
786 - Made ``setup-ssh`` exit with non-zero code if an error occurred
787 - Disabled RAPI CA checks in watcher
793 *(Released Mon, 23 Aug 2010)*
795 - Support DRBD versions of the format "a.b.c.d"
797 - Re-introduce support for usage from multiple threads in RAPI client
798 - Instance renames and modify via RAPI
799 - Work around race condition between processing and archival in job
801 - Mark opcodes following failed one as failed, too
802 - Job field ``lock_status`` was removed due to difficulties making it
803 work with the changed job queue in Ganeti 2.2; a better way to monitor
804 locks is expected for a later 2.2.x release
805 - Fixed dry-run behaviour with many commands
806 - Support ``ssh-agent`` again when adding nodes
807 - Many additional bugfixes
813 *(Released Fri, 30 Jul 2010)*
815 Important change: the internal RPC mechanism between Ganeti nodes has
816 changed from using a home-grown http library (based on the Python base
817 libraries) to use the PycURL library. This requires that PycURL is
818 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
819 against GnuTLS by default. cURL's support for GnuTLS had known issues
820 before cURL 7.21.0 and we recommend using the latest cURL release or
821 linking against OpenSSL. Most other distributions already link PycURL
822 and cURL against OpenSSL. The command::
824 python -c 'import pycurl; print pycurl.version'
826 can be used to determine the libraries PycURL and cURL are linked
829 Other significant changes:
831 - Rewrote much of the internals of the job queue, in order to achieve
832 better parallelism; this decouples job query operations from the job
833 processing, and it should allow much nicer behaviour of the master
834 daemon under load, and it also has uncovered some long-standing bugs
835 related to the job serialisation (now fixed)
836 - Added a default iallocator setting to the cluster parameters,
837 eliminating the need to always pass nodes or an iallocator for
838 operations that require selection of new node(s)
839 - Added experimental support for the LXC virtualization method
840 - Added support for OS parameters, which allows the installation of
841 instances to pass parameter to OS scripts in order to customise the
843 - Added a hypervisor parameter controlling the migration type (live or
844 non-live), since hypervisors have various levels of reliability; this
845 has renamed the 'live' parameter to 'mode'
846 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
847 logical volumes, meaning that cluster verify will ignore them and not
848 flag their presence as errors
849 - The watcher will now reset the error count for failed instances after
850 8 hours, thus allowing self-healing if the problem that caused the
851 instances to be down/fail to start has cleared in the meantime
852 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
853 check for, and warn, if the drbd module parameter ``usermode_helper``
854 is not consistent with the cluster-wide setting; this is needed to
855 make diagnose easier of failed drbd creations
856 - Started adding base IPv6 support, but this is not yet
857 enabled/available for use
858 - Rename operations (cluster, instance) will now return the new name,
859 which is especially useful if a short name was passed in
860 - Added support for instance migration in RAPI
861 - Added a tool to pre-configure nodes for the SSH setup, before joining
862 them to the cluster; this will allow in the future a simplified model
863 for node joining (but not yet fully enabled in 2.2); this needs the
864 paramiko python library
865 - Fixed handling of name-resolving errors
866 - Fixed consistency of job results on the error path
867 - Fixed master-failover race condition when executed multiple times in
869 - Fixed many bugs related to the job queue (mostly introduced during the
870 2.2 development cycle, so not all are impacting 2.1)
871 - Fixed instance migration with missing disk symlinks
872 - Fixed handling of unknown jobs in ``gnt-job archive``
873 - And many other small fixes/improvements
877 - Enhanced both the unittest and the QA coverage
878 - Switched the opcode validation to a generic model, and extended the
879 validation to all opcode parameters
880 - Changed more parts of the code that write shell scripts to use the
882 - Switched the master daemon to use the asyncore library for the Luxi
889 *(Released Thu, 17 Jun 2010)*
891 - Added tool (``move-instance``) and infrastructure to move instances
892 between separate clusters (see :doc:`separate documentation
893 <move-instance>` and :doc:`design document <design-2.2>`)
894 - Added per-request RPC timeout
895 - RAPI now requires a Content-Type header for requests with a body (e.g.
896 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
897 :rfc:`2616` (HTTP/1.1), section 7.2.1)
898 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
900 - Implemented initial support for running Ganeti daemons as separate
901 users, see configure-time flags ``--with-user-prefix`` and
902 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
904 - Instances can be removed after export (``gnt-backup export
906 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
907 key (instead of 1024 bit)
908 - Added new cluster configuration file for cluster domain secret
909 - Import/export now use SSL instead of SSH
910 - Added support for showing estimated time when exporting an instance,
911 see the ``ganeti-os-interface(7)`` manpage and look for
918 *(Released Tue, 16 Nov 2010)*
920 Some more bugfixes. Unless critical bugs occur, this will be the last
923 - Fix case of MAC special-values
924 - Fix mac checker regex
925 - backend: Fix typo causing "out of range" error
926 - Add missing --units in gnt-instance list man page
932 *(Released Tue, 24 Aug 2010)*
935 - Don't ignore secondary node silently on non-mirrored disk templates
937 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
938 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
939 - Properly document the usb_mouse parameter
940 - Fix path in ganeti-rapi(8) (issue 116)
941 - Adjust error message when the ganeti user's .ssh directory is
943 - Add same-node-check when changing the disk template to drbd
949 *(Released Fri, 16 Jul 2010)*
952 - Add an option to only select some reboot types during qa/burnin.
953 (on some hypervisors consequent reboots are not supported)
954 - Fix infrequent race condition in master failover. Sometimes the old
955 master ip address would be still detected as up for a short time
956 after it was removed, causing failover to fail.
957 - Decrease mlockall warnings when the ctypes module is missing. On
958 Python 2.4 we support running even if no ctypes module is installed,
959 but we were too verbose about this issue.
960 - Fix building on old distributions, on which man doesn't have a
962 - Fix RAPI not to ignore the MAC address on instance creation
963 - Implement the old instance creation format in the RAPI client.
969 *(Released Thu, 01 Jul 2010)*
971 A small bugfix release:
972 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
973 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
974 a non-existing option
975 - Fix instance create: the --force-variant option was ignored
976 - Improve pylint 0.21 compatibility and warnings with Python 2.6
977 - Fix modify node storage with non-FQDN arguments
978 - Fix RAPI client to authenticate under Python 2.6 when used
979 for more than 5 requests needing authentication
980 - Fix gnt-instance modify -t (storage) giving a wrong error message
981 when converting a non-shutdown drbd instance to plain
987 *(Released Fri, 18 Jun 2010)*
989 A small bugfix release:
991 - Fix live migration of KVM instances started with older Ganeti
992 versions which had fewer hypervisor parameters
993 - Fix gnt-instance grow-disk on down instances
994 - Fix an error-reporting bug during instance migration
995 - Better checking of the ``--net`` and ``--disk`` values, to avoid
996 silently ignoring broken ones
997 - Fix an RPC error reporting bug affecting, for example, RAPI client
999 - Fix bug triggered by different API version os-es on different nodes
1000 - Fix a bug in instance startup with custom hvparams: OS level
1001 parameters would fail to be applied.
1002 - Fix the RAPI client under Python 2.6 (but more work is needed to
1003 make it work completely well with OpenSSL)
1004 - Fix handling of errors when resolving names from DNS
1010 *(Released Thu, 3 Jun 2010)*
1012 A medium sized development cycle. Some new features, and some
1013 fixes/small improvements/cleanups.
1015 Significant features
1016 ~~~~~~~~~~~~~~~~~~~~
1018 The node deamon now tries to mlock itself into memory, unless the
1019 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
1020 its logs, and falls back to console logging. This allows emergency
1021 features such as ``gnt-node powercycle`` to work even in the event of a
1022 broken node disk (tested offlining the disk hosting the node's
1023 filesystem and dropping its memory caches; don't try this at home)
1025 KVM: add vhost-net acceleration support. It can be tested with a new
1026 enough version of the kernel and of qemu-kvm.
1028 KVM: Add instance chrooting feature. If you use privilege dropping for
1029 your VMs you can also now force them to chroot to an empty directory,
1030 before starting the emulated guest.
1032 KVM: Add maximum migration bandwith and maximum downtime tweaking
1033 support (requires a new-enough version of qemu-kvm).
1035 Cluster verify will now warn if the master node doesn't have the master
1036 ip configured on it.
1038 Add a new (incompatible) instance creation request format to RAPI which
1039 supports all parameters (previously only a subset was supported, and it
1040 wasn't possible to extend the old format to accomodate all the new
1041 features. The old format is still supported, and a client can check for
1042 this feature, before using it, by checking for its presence in the
1043 ``features`` RAPI resource.
1045 Now with ancient latin support. Try it passing the ``--roman`` option to
1046 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
1047 (requires the python-roman module to be installed, in order to work).
1052 As usual many internal code refactorings, documentation updates, and
1055 - Lots of improvements and cleanups to the experimental Remote API
1056 (RAPI) client library.
1057 - A new unit test suite for the core daemon libraries.
1058 - A fix to creating missing directories makes sure the umask is not
1059 applied anymore. This enforces the same directory permissions
1061 - Better handling terminating daemons with ctrl+c (used when running
1062 them in debugging mode).
1063 - Fix a race condition in live migrating a KVM instance, when stat()
1064 on the old proc status file returned EINVAL, which is an unexpected
1066 - Fixed manpage checking with newer man and utf-8 charachters. But now
1067 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
1073 *(Released Fri, 7 May 2010)*
1075 Fix a bug which prevented untagged KVM instances from starting.
1081 *(Released Fri, 7 May 2010)*
1083 Another release with a long development cycle, during which many
1084 different features were added.
1086 Significant features
1087 ~~~~~~~~~~~~~~~~~~~~
1089 The KVM hypervisor now can run the individual instances as non-root, to
1090 reduce the impact of a VM being hijacked due to bugs in the
1091 hypervisor. It is possible to run all instances as a single (non-root)
1092 user, to manually specify a user for each instance, or to dynamically
1093 allocate a user out of a cluster-wide pool to each instance, with the
1094 guarantee that no two instances will run under the same user ID on any
1097 An experimental RAPI client library, that can be used standalone
1098 (without the other Ganeti libraries), is provided in the source tree as
1099 ``lib/rapi/client.py``. Note this client might change its interface in
1100 the future, as we iterate on its capabilities.
1102 A new command, ``gnt-cluster renew-crypto`` has been added to easily
1103 replace the cluster's certificates and crypto keys. This might help in
1104 case they have been compromised, or have simply expired.
1106 A new disk option for instance creation has been added that allows one
1107 to "adopt" currently existing logical volumes, with data
1108 preservation. This should allow easier migration to Ganeti from
1109 unmanaged (or managed via other software) instances.
1111 Another disk improvement is the possibility to convert between redundant
1112 (DRBD) and plain (LVM) disk configuration for an instance. This should
1113 allow better scalability (starting with one node and growing the
1114 cluster, or shrinking a two-node cluster to one node).
1116 A new feature that could help with automated node failovers has been
1117 implemented: if a node sees itself as offline (by querying the master
1118 candidates), it will try to shutdown (hard) all instances and any active
1119 DRBD devices. This reduces the risk of duplicate instances if an
1120 external script automatically failovers the instances on such nodes. To
1121 enable this, the cluster parameter ``maintain_node_health`` should be
1122 enabled; in the future this option (per the name) will enable other
1123 automatic maintenance features.
1125 Instance export/import now will reuse the original instance
1126 specifications for all parameters; that means exporting an instance,
1127 deleting it and the importing it back should give an almost identical
1128 instance. Note that the default import behaviour has changed from
1129 before, where it created only one NIC; now it recreates the original
1132 Cluster verify has added a few new checks: SSL certificates validity,
1133 /etc/hosts consistency across the cluster, etc.
1138 As usual, many internal changes were done, documentation fixes,
1141 - Fixed cluster initialization with disabled cluster storage (regression
1142 introduced in 2.1.1)
1143 - File-based storage supports growing the disks
1144 - Fixed behaviour of node role changes
1145 - Fixed cluster verify for some corner cases, plus a general rewrite of
1146 cluster verify to allow future extension with more checks
1147 - Fixed log spamming by watcher and node daemon (regression introduced
1149 - Fixed possible validation issues when changing the list of enabled
1151 - Fixed cleanup of /etc/hosts during node removal
1152 - Fixed RAPI response for invalid methods
1153 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
1154 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
1155 from ide disks, etc.)
1156 - Allow OS changes without re-installation (to record a changed OS
1157 outside of Ganeti, or to allow OS renames)
1158 - Allow instance creation without OS installation (useful for example if
1159 the OS will be installed manually, or restored from a backup not in
1161 - Implemented option to make cluster ``copyfile`` use the replication
1163 - Added list of enabled hypervisors to ssconf (possibly useful for
1165 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
1167 - A partial form of node re-IP is possible via node readd, which now
1168 allows changed node primary IP
1169 - Command line utilities now show an informational message if the job is
1171 - The logs of the master daemon now show the PID/UID/GID of the
1178 *(Released Fri, 12 Mar 2010)*
1180 During the 2.1.0 long release candidate cycle, a lot of improvements and
1181 changes have accumulated with were released later as 2.1.1.
1186 The node evacuate command (``gnt-node evacuate``) was significantly
1187 rewritten, and as such the IAllocator protocol was changed - a new
1188 request type has been added. This unfortunate change during a stable
1189 series is designed to improve performance of node evacuations; on
1190 clusters with more than about five nodes and which are well-balanced,
1191 evacuation should proceed in parallel for all instances of the node
1192 being evacuated. As such, any existing IAllocator scripts need to be
1193 updated, otherwise the above command will fail due to the unknown
1194 request. The provided "dumb" allocator has not been updated; but the
1195 ganeti-htools package supports the new protocol since version 0.2.4.
1197 Another important change is increased validation of node and instance
1198 names. This might create problems in special cases, if invalid host
1199 names are being used.
1201 Also, a new layer of hypervisor parameters has been added, that sits at
1202 OS level between the cluster defaults and the instance ones. This allows
1203 customisation of virtualization parameters depending on the installed
1204 OS. For example instances with OS 'X' may have a different KVM kernel
1205 (or any other parameter) than the cluster defaults. This is intended to
1206 help managing a multiple OSes on the same cluster, without manual
1207 modification of each instance's parameters.
1209 A tool for merging clusters, ``cluster-merge``, has been added in the
1210 tools sub-directory.
1215 - Improved the int/float conversions that should make the code more
1216 robust in face of errors from the node daemons
1217 - Fixed the remove node code in case of internal configuration errors
1218 - Fixed the node daemon behaviour in face of inconsistent queue
1219 directory (e.g. read-only file-system where we can't open the files
1221 - Fixed the behaviour of gnt-node modify for master candidate demotion;
1222 now it either aborts cleanly or, if given the new "auto_promote"
1223 parameter, will automatically promote other nodes as needed
1224 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
1225 completely prevent Ganeti from working
1226 - Fixed bug for instance export when not all disks were successfully
1228 - Fixed behaviour of node add when the new node is slow in starting up
1230 - Fixed handling of signals in the LUXI client, which should improve
1231 behaviour of command-line scripts
1232 - Added checks for invalid node/instance names in the configuration (now
1233 flagged during cluster verify)
1234 - Fixed watcher behaviour for disk activation errors
1235 - Fixed two potentially endless loops in http library, which led to the
1236 RAPI daemon hanging and consuming 100% CPU in some cases
1237 - Fixed bug in RAPI daemon related to hashed passwords
1238 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
1240 - Enhanced compatibility with non-Debian OSes, but not using absolute
1241 path in some commands and allowing customisation of the ssh
1242 configuration directory
1243 - Fixed possible future issue with new Python versions by abiding to the
1244 proper use of ``__slots__`` attribute on classes
1245 - Added checks that should prevent directory traversal attacks
1246 - Many documentation fixes based on feedback from users
1251 - Added an "early_release" more for instance replace disks and node
1252 evacuate, where we release locks earlier and thus allow higher
1253 parallelism within the cluster
1254 - Added watcher hooks, intended to allow the watcher to restart other
1255 daemons (e.g. from the ganeti-nbma project), but they can be used of
1256 course for any other purpose
1257 - Added a compile-time disable for DRBD barriers, to increase
1258 performance if the administrator trusts the power supply or the
1259 storage system to not lose writes
1260 - Added the option of using syslog for logging instead of, or in
1261 addition to, Ganeti's own log files
1262 - Removed boot restriction for paravirtual NICs for KVM, recent versions
1263 can indeed boot from a paravirtual NIC
1264 - Added a generic debug level for many operations; while this is not
1265 used widely yet, it allows one to pass the debug value all the way to
1267 - Enhanced the hooks environment for instance moves (failovers,
1268 migrations) where the primary/secondary nodes changed during the
1269 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
1270 - Enhanced data validations for many user-supplied values; one important
1271 item is the restrictions imposed on instance and node names, which
1272 might reject some (invalid) host names
1273 - Add a configure-time option to disable file-based storage, if it's not
1274 needed; this allows greater security separation between the master
1275 node and the other nodes from the point of view of the inter-node RPC
1277 - Added user notification in interactive tools if job is waiting in the
1278 job queue or trying to acquire locks
1279 - Added log messages when a job is waiting for locks
1280 - Added filtering by node tags in instance operations which admit
1281 multiple instances (start, stop, reboot, reinstall)
1282 - Added a new tool for cluster mergers, ``cluster-merge``
1283 - Parameters from command line which are of the form ``a=b,c=d`` can now
1284 use backslash escapes to pass in values which contain commas,
1285 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
1287 - For KVM, the instance name is the first parameter passed to KVM, so
1288 that it's more visible in the process list
1294 *(Released Tue, 2 Mar 2010)*
1296 Ganeti 2.1 brings many improvements with it. Major changes:
1298 - Added infrastructure to ease automated disk repairs
1299 - Added new daemon to export configuration data in a cheaper way than
1300 using the remote API
1301 - Instance NICs can now be routed instead of being associated with a
1303 - Improved job locking logic to reduce impact of jobs acquiring multiple
1304 locks waiting for other long-running jobs
1306 In-depth implementation details can be found in the Ganeti 2.1 design
1312 - Added chroot hypervisor
1313 - Added more options to xen-hvm hypervisor (``kernel_path`` and
1315 - Added more options to xen-pvm hypervisor (``use_bootloader``,
1316 ``bootloader_path`` and ``bootloader_args``)
1317 - Added the ``use_localtime`` option for the xen-hvm and kvm
1318 hypervisors, and the default value for this has changed to false (in
1319 2.0 xen-hvm always enabled it)
1320 - Added luxi call to submit multiple jobs in one go
1321 - Added cluster initialization option to not modify ``/etc/hosts``
1323 - Added network interface parameters
1324 - Added dry run mode to some LUs
1325 - Added RAPI resources:
1327 - ``/2/instances/[instance_name]/info``
1328 - ``/2/instances/[instance_name]/replace-disks``
1329 - ``/2/nodes/[node_name]/evacuate``
1330 - ``/2/nodes/[node_name]/migrate``
1331 - ``/2/nodes/[node_name]/role``
1332 - ``/2/nodes/[node_name]/storage``
1333 - ``/2/nodes/[node_name]/storage/modify``
1334 - ``/2/nodes/[node_name]/storage/repair``
1336 - Added OpCodes to evacuate or migrate all instances on a node
1337 - Added new command to list storage elements on nodes (``gnt-node
1338 list-storage``) and modify them (``gnt-node modify-storage``)
1339 - Added new ssconf files with master candidate IP address
1340 (``ssconf_master_candidates_ips``), node primary IP address
1341 (``ssconf_node_primary_ips``) and node secondary IP address
1342 (``ssconf_node_secondary_ips``)
1343 - Added ``ganeti-confd`` and a client library to query the Ganeti
1344 configuration via UDP
1345 - Added ability to run hooks after cluster initialization and before
1347 - Added automatic mode for disk replace (``gnt-instance replace-disks
1349 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
1350 after catastrophic data-loss
1351 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
1353 - Added ``gnt-instance move`` command to move instances
1354 - Added ``gnt-cluster watcher`` command to control watcher
1355 - Added ``gnt-node powercycle`` command to powercycle nodes
1356 - Added new job status field ``lock_status``
1357 - Added parseable error codes to cluster verification (``gnt-cluster
1358 verify --error-codes``) and made output less verbose (use
1359 ``--verbose`` to restore previous behaviour)
1360 - Added UUIDs to the main config entities (cluster, nodes, instances)
1361 - Added support for OS variants
1362 - Added support for hashed passwords in the Ganeti remote API users file
1364 - Added option to specify maximum timeout on instance shutdown
1365 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
1366 - Added new helper script to start and stop Ganeti daemons
1367 (``daemon-util``), with the intent to reduce the work necessary to
1368 adjust Ganeti for non-Debian distributions and to start/stop daemons
1370 - Added more unittests
1371 - Fixed critical bug in ganeti-masterd startup
1372 - Removed the configure-time ``kvm-migration-port`` parameter, this is
1373 now customisable at the cluster level for both the KVM and Xen
1374 hypervisors using the new ``migration_port`` parameter
1375 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
1376 reinstalling an instance
1377 - Allowed ``@`` in tag names
1378 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
1379 - Many documentation updates
1380 - Distribute hypervisor files on ``gnt-cluster redist-conf``
1381 - ``gnt-instance reinstall`` can now reinstall multiple instances
1382 - Updated many command line parameters
1383 - Introduced new OS API version 15
1384 - No longer support a default hypervisor
1385 - Treat virtual LVs as inexistent
1386 - Improved job locking logic to reduce lock contention
1387 - Match instance and node names case insensitively
1388 - Reimplemented bash completion script to be more complete
1395 *(Released Thu, 4 Feb 2010)*
1397 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
1398 - Fix a string formatting bug
1399 - Improve safety of the code in some error paths
1400 - Improve data validation in the master of values returned from nodes
1406 *(Released Thu, 17 Dec 2009)*
1408 - Fix security issue due to missing validation of iallocator names; this
1409 allows local and remote execution of arbitrary executables
1410 - Fix failure of gnt-node list during instance removal
1411 - Ship the RAPI documentation in the archive
1417 *(Released Wed, 30 Sep 2009)*
1419 - Fixed many wrong messages
1420 - Fixed a few bugs related to the locking library
1421 - Fixed MAC checking at instance creation time
1422 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
1423 - Fixed a few issues related to signal handling in both daemons and
1425 - Fixed the example startup script provided
1426 - Fixed insserv dependencies in the example startup script (patch from
1428 - Fixed handling of drained nodes in the iallocator framework
1429 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
1431 - Fixed error related to invalid job IDs in job polling
1432 - Fixed job/opcode persistence on unclean master shutdown
1433 - Fixed handling of partial job processing after unclean master
1435 - Fixed error reporting from LUs, previously all errors were converted
1436 into execution errors
1437 - Fixed error reporting from burnin
1438 - Decreased significantly the memory usage of the job queue
1439 - Optimised slightly multi-job submission
1440 - Optimised slightly opcode loading
1441 - Backported the multi-job submit framework from the development
1442 branch; multi-instance start and stop should be faster
1443 - Added script to clean archived jobs after 21 days; this will reduce
1444 the size of the queue directory
1445 - Added some extra checks in disk size tracking
1446 - Added an example ethers hook script
1447 - Added a cluster parameter that prevents Ganeti from modifying of
1449 - Added more node information to RAPI responses
1450 - Added a ``gnt-job watch`` command that allows following the ouput of a
1452 - Added a bind-address option to ganeti-rapi
1453 - Added more checks to the configuration verify
1454 - Enhanced the burnin script such that some operations can be retried
1456 - Converted instance reinstall to multi-instance model
1462 *(Released Fri, 7 Aug 2009)*
1464 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
1465 to allow using the pre-2.0.2 behaviour in activation, if any existing
1466 instances have mismatched disk sizes in the configuration
1467 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
1468 any configuration mismatches for disk sizes
1469 - Added ``gnt-master cluste-failover --no-voting`` to allow master
1470 failover to work on two-node clusters
1471 - Fixed the ``--net`` option of ``gnt-backup import``, which was
1473 - Fixed detection of OS script errors in ``gnt-backup export``
1474 - Fixed exit code of ``gnt-backup export``
1480 *(Released Fri, 17 Jul 2009)*
1482 - Added experimental support for stripped logical volumes; this should
1483 enhance performance but comes with a higher complexity in the block
1484 device handling; stripping is only enabled when passing
1485 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
1486 affected even in the non-stripped mode
1487 - Improved resiliency against transient failures at the end of DRBD
1488 resyncs, and in general of DRBD resync checks
1489 - Fixed a couple of issues with exports and snapshot errors
1490 - Fixed a couple of issues in instance listing
1491 - Added display of the disk size in ``gnt-instance info``
1492 - Fixed checking for valid OSes in instance creation
1493 - Fixed handling of the "vcpus" parameter in instance listing and in
1494 general of invalid parameters
1495 - Fixed http server library, and thus RAPI, to handle invalid
1496 username/password combinations correctly; this means that now they
1497 report unauthorized for queries too, not only for modifications,
1498 allowing earlier detect of configuration problems
1499 - Added a new "role" node list field, equivalent to the master/master
1500 candidate/drained/offline flags combinations
1501 - Fixed cluster modify and changes of candidate pool size
1502 - Fixed cluster verify error messages for wrong files on regular nodes
1503 - Fixed a couple of issues with node demotion from master candidate role
1504 - Fixed node readd issues
1505 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
1506 - Added a new ``--no-voting`` option for masterfailover to fix failover
1507 on two-nodes clusters when the former master node is unreachable
1508 - Added instance reinstall over RAPI
1514 *(Released Tue, 16 Jun 2009)*
1516 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
1517 allow re-adding the start in single-user option (regression from 1.2)
1518 - the watcher writes the instance status to a file, to allow monitoring
1519 to report the instance status (from the master) based on cached
1520 results of the watcher's queries; while this can get stale if the
1521 watcher is being locked due to other work on the cluster, this is
1522 still an improvement
1523 - the watcher now also restarts the node daemon and the rapi daemon if
1525 - fixed the watcher to handle full and drained queue cases
1526 - hooks export more instance data in the environment, which helps if
1527 hook scripts need to take action based on the instance's properties
1528 (no longer need to query back into ganeti)
1529 - instance failovers when the instance is stopped do not check for free
1530 RAM, so that failing over a stopped instance is possible in low memory
1532 - rapi uses queries for tags instead of jobs (for less job traffic), and
1533 for cluster tags it won't talk to masterd at all but read them from
1535 - a couple of error handling fixes in RAPI
1536 - drbd handling: improved the error handling of inconsistent disks after
1537 resync to reduce the frequency of "there are some degraded disks for
1538 this instance" messages
1539 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
1540 error handling path called a wrong function name)
1546 *(Released Wed, 27 May 2009)*
1548 - no changes from rc5
1554 *(Released Wed, 20 May 2009)*
1556 - fix a couple of bugs (validation, argument checks)
1557 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
1558 - some small improvements to RAPI and IAllocator
1559 - make watcher automatically start the master daemon if down
1565 *(Released Mon, 27 Apr 2009)*
1567 - change the OS list to not require locks; this helps with big clusters
1568 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
1569 volume group is broken
1570 - ``gnt-instance info``, without any arguments, doesn't run for all
1571 instances anymore; either pass ``--all`` or pass the desired
1572 instances; this helps against mistakes on big clusters where listing
1573 the information for all instances takes a long time
1574 - miscellaneous doc and man pages fixes
1580 *(Released Wed, 8 Apr 2009)*
1582 - Change the internal locking model of some ``gnt-node`` commands, in
1583 order to reduce contention (and blocking of master daemon) when
1584 batching many creation/reinstall jobs
1585 - Fixes to Xen soft reboot
1586 - No longer build documentation at build time, instead distribute it in
1587 the archive, in order to reduce the need for the whole docbook/rst
1594 *(Released Fri, 27 Mar 2009)*
1596 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
1597 - Fix watcher startup sequence, improves the behaviour of busy clusters
1598 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
1599 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
1600 bind address and other places
1601 - Some documentation fixes and updates
1607 *(Released Mon, 2 Mar 2009)*
1609 - More documentation updates, now all docs should be more-or-less
1611 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
1613 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
1615 - Fix an issue related to $libdir/run/ganeti and cluster creation
1621 *(Released Thu, 19 Feb 2009)*
1623 - Xen PVM and KVM have switched the default value for the instance root
1624 disk to the first partition on the first drive, instead of the whole
1625 drive; this means that the OS installation scripts must be changed
1627 - Man pages have been updated
1628 - RAPI has been switched by default to HTTPS, and the exported functions
1629 should all work correctly
1630 - RAPI v1 has been removed
1631 - Many improvements to the KVM hypervisor
1632 - Block device errors are now better reported
1633 - Many other bugfixes and small improvements
1639 *(Released Mon, 26 Jan 2009)*
1641 - Version 2 is a general rewrite of the code and therefore the
1642 differences are too many to list, see the design document for 2.0 in
1643 the ``doc/`` subdirectory for more details
1644 - In this beta version there is not yet a migration path from 1.2 (there
1645 will be one in the final 2.0 release)
1646 - A few significant changes are:
1648 - all commands are executed by a daemon (``ganeti-masterd``) and the
1649 various ``gnt-*`` commands are just front-ends to it
1650 - all the commands are entered into, and executed from a job queue,
1651 see the ``gnt-job(8)`` manpage
1652 - the RAPI daemon supports read-write operations, secured by basic
1653 HTTP authentication on top of HTTPS
1654 - DRBD version 0.7 support has been removed, DRBD 8 is the only
1655 supported version (when migrating from Ganeti 1.2 to 2.0, you need
1656 to migrate to DRBD 8 first while still running Ganeti 1.2)
1657 - DRBD devices are using statically allocated minor numbers, which
1658 will be assigned to existing instances during the migration process
1659 - there is support for both Xen PVM and Xen HVM instances running on
1661 - KVM virtualization is supported too
1662 - file-based storage has been implemented, which means that it is
1663 possible to run the cluster without LVM and DRBD storage, for
1664 example using a shared filesystem exported from shared storage (and
1665 still have live migration)
1671 *(Released Tue, 13 Jan 2009)*
1673 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
1674 - Reuse the old instance mac address by default on instance import, if
1675 the instance name is the same.
1676 - Handle situations in which the node info rpc returns incomplete
1678 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
1679 - Improved version of batcher:
1681 - state file support
1682 - instance mac address support
1683 - support for HVM clusters/instances
1685 - Add an option to show the number of cpu sockets and nodes in
1687 - Support OSes that handle more than one version of the OS api (but do
1688 not change the current API in any other way)
1689 - Fix ``gnt-node migrate``
1690 - ``gnt-debug`` man page
1691 - Fixes various more typos and small issues
1692 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
1698 *(Released Wed, 24 Sep 2008)*
1700 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
1701 type of disk exported to fully virtualized instances.
1702 - provide access to the serial console of HVM instances
1703 - instance auto_balance flag, set by default. If turned off it will
1704 avoid warnings on cluster verify if there is not enough memory to fail
1705 over an instance. in the future it will prevent automatically failing
1706 it over when we will support that.
1707 - batcher tool for instance creation, see ``tools/README.batcher``
1708 - ``gnt-instance reinstall --select-os`` to interactively select a new
1709 operating system when reinstalling an instance.
1710 - when changing the memory amount on instance modify a check has been
1711 added that the instance will be able to start. also warnings are
1712 emitted if the instance will not be able to fail over, if auto_balance
1714 - documentation fixes
1715 - sync fields between ``gnt-instance list/modify/add/import``
1716 - fix a race condition in drbd when the sync speed was set after giving
1717 the device a remote peer.
1723 *(Released Tue, 22 Jul 2008)*
1725 - note: the allowed size and number of tags per object were reduced
1726 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
1727 - fixed twisted 8.x compatibility
1728 - fixed ``gnt-instance replace-disks`` with iallocator
1729 - add TCP keepalives on twisted connections to detect restarted nodes
1730 - disk increase support, see ``gnt-instance grow-disk``
1731 - implement bulk node/instance query for RAPI
1732 - add tags in node/instance listing (optional)
1733 - experimental migration (and live migration) support, read the man page
1734 for ``gnt-instance migrate``
1735 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
1736 has some small improvements in handling its state file
1742 *(Released Fri, 13 Jun 2008)*
1744 - Experimental readonly, REST-based remote API implementation;
1745 automatically started on master node, TCP port 5080, if enabled by
1746 ``--enable-rapi`` parameter to configure script.
1747 - Instance allocator support. Add and import instance accept a
1748 ``--iallocator`` parameter, and call that instance allocator to decide
1749 which node to use for the instance. The iallocator document describes
1750 what's expected from an allocator script.
1751 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
1752 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
1753 node is lost there is still enough memory to fail over the instances
1755 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
1756 ``gnt-cluster verify``, to check for site-specific compliance. All the
1757 hooks will run, and their output, if any, will be displayed. Any
1758 failing hook will make the verification return an error value.
1759 - ``gnt-cluster verify`` now checks that its peers are reachable on the
1760 primary and secondary interfaces
1761 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
1762 that is still declared as part of the cluster and has failed.
1763 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
1764 specifying output fields, that just adds the chosen fields to the
1766 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
1767 export from the filesystem.
1768 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
1769 have been added. Using them you can enable/disable acpi and pae
1770 support, and specify a path for a cd image to be exported to the
1771 instance. These parameters as the name suggest only work on HVM
1773 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
1774 PAE support will be set to the previously hardcoded values, but the
1775 (previously hardcoded) path to the CDROM ISO image will be unset and
1776 if required, needs to be set manually with ``gnt-instance modify``
1778 - The address to which an instance's VNC console is bound is now
1779 selectable per-instance, rather than being cluster wide. Of course
1780 this only applies to instances controlled via VNC, so currently just
1781 applies to HVM clusters.
1787 *(Released Mon, 18 Feb 2008)*
1789 - more tweaks to the disk activation code (especially helpful for DRBD)
1790 - change the default ``gnt-instance list`` output format, now there is
1791 one combined status field (see the manpage for the exact values this
1793 - some more fixes for the mac export to hooks change
1794 - make Ganeti not break with DRBD 8.2.x (which changed the version
1795 format in ``/proc/drbd``) (issue 24)
1796 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
1797 template, allowing migration from DRBD0.7+MD to DRBD8
1803 *(Released Wed, 30 Jan 2008)*
1805 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
1807 - add command aliases infrastructure and a few aliases
1808 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
1809 man pages and the ``--help`` option of ``gnt-node
1810 list``/``gnt-instance list``
1811 - fix ``gnt-backup list`` with down nodes (issue 21)
1812 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
1813 - fix the dist archive and add a check for including svn/git files in
1815 - some developer-related changes: improve the burnin and the QA suite,
1816 add an upload script for testing during development
1822 *(Released Wed, 16 Jan 2008)*
1824 - experimental HVM support, read the install document, section
1825 "Initializing the cluster"
1826 - allow for the PVM hypervisor per-instance kernel and initrd paths
1827 - add a new command ``gnt-cluster verify-disks`` which uses a new
1828 algorithm to improve the reconnection of the DRBD pairs if the device
1829 on the secondary node has gone away
1830 - make logical volume code auto-activate LVs at disk activation time
1831 - slightly improve the speed of activating disks
1832 - allow specification of the MAC address at instance creation time, and
1833 changing it later via ``gnt-instance modify``
1834 - fix handling of external commands that generate lots of output on
1836 - update documentation with regard to minimum version of DRBD8 supported
1842 *(Released Tue, 4 Dec 2007)*
1844 - Log the ``xm create`` output to the node daemon log on failure (to
1845 help diagnosing the error)
1846 - In debug mode, log all external commands output if failed to the logs
1847 - Change parsing of lvm commands to ignore stderr
1853 *(Released Wed, 28 Nov 2007)*
1855 - Another round of updates to the DRBD 8 code to deal with more failures
1856 in the replace secondary node operation
1857 - Some more logging of failures in disk operations (lvm, drbd)
1858 - A few documentation updates
1865 *(Released Tue, 13 Nov 2007)*
1867 - Change configuration file format from Python's Pickle to JSON.
1868 Upgrading is possible using the cfgupgrade utility.
1869 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
1870 faster replace disks and is more stable (DRBD 8 has many improvements
1871 compared to DRBD 0.7)
1872 - Added command line tags support (see man pages for ``gnt-instance``,
1873 ``gnt-node``, ``gnt-cluster``)
1874 - Added instance rename support
1875 - Added multi-instance startup/shutdown
1876 - Added cluster rename support
1877 - Added ``gnt-node evacuate`` to simplify some node operations
1878 - Added instance reboot operation that can speedup reboot as compared to
1880 - Soften the requirement that hostnames are in FQDN format
1881 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
1883 - Removed dependency on debian's patched fping that uses the
1884 non-standard ``-S`` option
1885 - Now the OS definitions are searched for in multiple, configurable
1886 paths (easier for distros to package)
1887 - Some changes to the hooks infrastructure (especially the new
1888 post-configuration update hook)
1889 - Other small bugfixes
1891 .. vim: set textwidth=72 syntax=rst :