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
20 *(Released Mon, 9 Jan 2012)*
22 Incompatible/important changes and bugfixes
23 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
25 - The default of the ``/2/instances/[instance_name]/rename`` RAPI
26 resource's ``ip_check`` parameter changed from ``True`` to ``False``
27 to match the underlying LUXI interface.
28 - The ``/2/nodes/[node_name]/evacuate`` RAPI resource was changed to use
29 body parameters, see :doc:`RAPI documentation <rapi>`. The server does
30 not maintain backwards-compatibility as the underlying operation
31 changed in an incompatible way. The RAPI client can talk to old
32 servers, but it needs to be told so as the return value changed.
33 - When creating file-based instances via RAPI, the ``file_driver``
34 parameter no longer defaults to ``loop`` and must be specified.
35 - The deprecated ``bridge`` NIC parameter is no longer supported. Use
37 - Support for the undocumented and deprecated RAPI instance creation
38 request format version 0 has been dropped. Use version 1, supported
39 since Ganeti 2.1.3 and :doc:`documented <rapi>`, instead.
40 - Pyparsing 1.4.6 or above is required, see :doc:`installation
41 documentation <install>`.
42 - The "cluster-verify" hooks are now executed per group by the
43 ``OP_CLUSTER_VERIFY_GROUP`` opcode. This maintains the same behavior
44 if you just run ``gnt-cluster verify``, which generates one opcode per
46 - The environment as passed to the OS scripts is cleared, and thus no
47 environment variables defined in the node daemon's environment will be
48 inherited by the scripts.
49 - The :doc:`iallocator <iallocator>` mode ``multi-evacuate`` has been
51 - :doc:`New iallocator modes <design-multi-reloc>` have been added to
52 support operations involving multiple node groups.
53 - Offline nodes are ignored when failing over an instance.
54 - Support for KVM version 1.0, which changed the version reporting format
56 - Includes all bugfixes made in the 2.4 series
61 - The ganeti-htools project has been merged into the ganeti-core source
62 tree and will be built as part of Ganeti (see :doc:`install-quick`).
63 - Implemented support for :doc:`shared storage <design-shared-storage>`.
64 - Add support for disks larger than 2 TB in ``lvmstrap`` by supporting
65 GPT-style partition tables (requires `parted
66 <http://www.gnu.org/s/parted/>`_).
67 - Added support for floppy drive and 2nd CD-ROM drive in KVM hypervisor.
68 - Allowed adding tags on instance creation.
69 - Export instance tags to hooks (``INSTANCE_TAGS``, see :doc:`hooks`)
70 - Allow instances to be started in a paused state, enabling the user to
71 see the complete console output on boot using the console.
72 - Added new hypervisor flag to control default reboot behaviour
73 (``reboot_behavior``).
74 - Added support for KVM keymaps (hypervisor parameter ``keymap``).
75 - Improved out-of-band management support:
77 - Added ``gnt-node health`` command reporting the health status of
79 - Added ``gnt-node power`` command to manage power status of nodes.
80 - Added command for emergency power-off (EPO), ``gnt-cluster epo``.
82 - Instance migration can fall back to failover if instance is not
84 - Filters can be used when listing nodes, instances, groups and locks;
85 see *ganeti(7)* manpage.
86 - Added post-execution status as variables to :doc:`hooks <hooks>`
88 - Instance tags are exported/imported together with the instance.
89 - When given an explicit job ID, ``gnt-job info`` will work for archived
91 - Jobs can define dependencies on other jobs (not yet supported via
92 RAPI or command line, but used by internal commands and usable via
95 - Lock monitor (``gnt-debug locks``) shows jobs waiting for
98 - Instance failover is now available as a RAPI resource
99 (``/2/instances/[instance_name]/failover``).
100 - ``gnt-instance info`` defaults to static information if primary node
102 - Opcodes have a new ``comment`` attribute.
103 - Added basic SPICE support to KVM hypervisor.
104 - ``tools/ganeti-listrunner`` allows passing of arguments to executable.
106 Node group improvements
107 ~~~~~~~~~~~~~~~~~~~~~~~
109 - ``gnt-cluster verify`` has been modified to check groups separately,
110 thereby improving performance.
111 - Node group support has been added to ``gnt-cluster verify-disks``,
112 which now operates per node group.
113 - Watcher has been changed to work better with node groups.
115 - One process and state file per node group.
116 - Slow watcher in one group doesn't block other group's watcher.
118 - Added new command, ``gnt-group evacuate``, to move all instances in a
119 node group to other groups.
120 - Added ``gnt-instance change-group`` to move an instance to another
122 - ``gnt-cluster command`` and ``gnt-cluster copyfile`` now support
123 per-group operations.
124 - Node groups can be tagged.
125 - Some operations switch from an exclusive to a shared lock as soon as
127 - Instance's primary and secondary nodes' groups are now available as
128 query fields (``pnode.group``, ``pnode.group.uuid``, ``snodes.group``
129 and ``snodes.group.uuid``).
134 - Numerous updates to documentation and manpages.
136 - :doc:`RAPI <rapi>` documentation now has detailed parameter
138 - Some opcode/job results are now also documented, see :doc:`RAPI
141 - A lockset's internal lock is now also visible in lock monitor.
142 - Log messages from job queue workers now contain information about the
143 opcode they're processing.
144 - ``gnt-instance console`` no longer requires the instance lock.
145 - A short delay when waiting for job changes reduces the number of LUXI
146 requests significantly.
147 - DRBD metadata volumes are overwritten with zeros during disk creation.
153 *(Released Thu, 27 Oct 2011)*
155 This was the fourth release candidate of the 2.5 series.
161 *(Released Wed, 26 Oct 2011)*
163 This was the third release candidate of the 2.5 series.
169 *(Released Tue, 18 Oct 2011)*
171 This was the second release candidate of the 2.5 series.
177 *(Released Tue, 4 Oct 2011)*
179 This was the first release candidate of the 2.5 series.
185 *(Released Wed, 31 Aug 2011)*
187 This was the third beta release of the 2.5 series.
193 *(Released Mon, 22 Aug 2011)*
195 This was the second beta release of the 2.5 series.
201 *(Released Fri, 12 Aug 2011)*
203 This was the first beta release of the 2.5 series.
209 *(Released Thu, 27 Oct 2011)*
211 - Fixed bug when parsing command line parameter values ending in
213 - Fixed assertion error after unclean master shutdown
214 - Disable HTTP client pool for RPC, significantly reducing memory usage
216 - Fixed queue archive creation with wrong permissions
222 *(Released Tue, 23 Aug 2011)*
226 - Fixed documentation for importing with ``--src-dir`` option
227 - Fixed a bug in ``ensure-dirs`` with queue/archive permissions
228 - Fixed a parsing issue with DRBD 8.3.11 in the Linux kernel
234 *(Released Fri, 5 Aug 2011)*
236 Many bug-fixes and a few small features:
238 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
239 issues when you tried to add an instance with two MAC addresses in one
241 - KVM: fixed per-instance stored UID value
242 - KVM: configure bridged NICs at migration start
243 - KVM: Fix a bug where instance will not start with never KVM versions
245 - Added OS search path to ``gnt-cluster info``
246 - Fixed an issue with ``file_storage_dir`` where you were forced to
247 provide an absolute path, but the documentation states it is a
248 relative path, the documentation was right
249 - Added a new parameter to instance stop/start called ``--no-remember``
250 that will make the state change to not be remembered
251 - Implemented ``no_remember`` at RAPI level
252 - Improved the documentation
253 - Node evacuation: don't call IAllocator if node is already empty
254 - Fixed bug in DRBD8 replace disks on current nodes
255 - Fixed bug in recreate-disks for DRBD instances
256 - Moved assertion checking locks in ``gnt-instance replace-disks``
257 causing it to abort with not owning the right locks for some situation
258 - Job queue: Fixed potential race condition when cancelling queued jobs
259 - Fixed off-by-one bug in job serial generation
260 - ``gnt-node volumes``: Fix instance names
261 - Fixed aliases in bash completion
262 - Fixed a bug in reopening log files after being sent a SIGHUP
263 - Added a flag to burnin to allow specifying VCPU count
264 - Bugfixes to non-root Ganeti configuration
270 *(Released Thu, 12 May 2011)*
272 Many bug-fixes and a few new small features:
274 - Fixed a bug related to log opening failures
275 - Fixed a bug in instance listing with orphan instances
276 - Fixed a bug which prevented resetting the cluster-level node parameter
277 ``oob_program`` to the default
278 - Many fixes related to the ``cluster-merge`` tool
279 - Fixed a race condition in the lock monitor, which caused failures
280 during (at least) creation of many instances in parallel
281 - Improved output for gnt-job info
282 - Removed the quiet flag on some ssh calls which prevented debugging
284 - Improved the N+1 failure messages in cluster verify by actually
285 showing the memory values (needed and available)
286 - Increased lock attempt timeouts so that when executing long operations
287 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
288 too early and thus prevent the use of the 'fair' mechanism
289 - Changed instance query data (``gnt-instance info``) to not acquire
290 locks unless needed, thus allowing its use on locked instance if only
291 static information is asked for
292 - Improved behaviour with filesystems that do not support rename on an
294 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
295 kept locks on all nodes during the wipe, which is unneeded
296 - Fixed ``gnt-watcher`` handling of errors during hooks execution
297 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
298 10GiB) which caused the wipe to fail right at the end in some cases
299 - Fixed master IP activation when doing master failover with no-voting
300 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
301 the master node itself
302 - Fixed potential data-loss in under disk full conditions, where Ganeti
303 wouldn't check correctly the return code and would consider
304 partially-written files 'correct'
305 - Fixed bug related to multiple VGs and DRBD disk replacing
306 - Added new disk parameter ``metavg`` that allows placement of the meta
307 device for DRBD in a different volume group
308 - Fixed error handling in the node daemon when the system libc doesn't
309 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
310 - Fixed lock release during replace-disks, which kept cluster-wide locks
311 when doing disk replaces with an iallocator script
312 - Added check for missing bridges in cluster verify
313 - Handle EPIPE errors while writing to the terminal better, so that
314 piping the output to e.g. ``less`` doesn't cause a backtrace
315 - Fixed rare case where a ^C during Luxi calls could have been
316 interpreted as server errors, instead of simply terminating
317 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
319 - Added a few more parameters to the KVM hypervisor, allowing a second
320 CDROM, custom disk type for CDROMs and a floppy image
321 - Removed redundant message in instance rename when the name is given
323 - Added option to ``gnt-instance recreate-disks`` to allow creating the
324 disks on new nodes, allowing recreation when the original instance
325 nodes are completely gone
326 - Added option when converting disk templates to DRBD to skip waiting
327 for the resync, in order to make the instance available sooner
328 - Added two new variables to the OS scripts environment (containing the
330 - Made the root_path and optional parameter for the xen-pvm hypervisor,
331 to allow use of ``pvgrub`` as bootloader
332 - Changed the instance memory modifications to only check out-of-memory
333 conditions on memory increases, and turned the secondary node warnings
334 into errors (they can still be overridden via ``--force``)
335 - Fixed the handling of a corner case when the Python installation gets
336 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
337 to execute a command that doesn't exist
338 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
339 node of the instance returned failures during instance shutdown; this
340 adds the option ``--ignore-consistency`` to gnt-instance move
342 And as usual, various improvements to the error messages, documentation
349 *(Released Wed, 09 Mar 2011)*
351 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
352 the RAPI users file if run twice (even with ``--dry-run``).
354 The release fixes that bug (nothing else changed).
360 *(Released Mon, 07 Mar 2011)*
362 Final 2.4.0 release. Just a few small fixes:
364 - Fixed RAPI node evacuate
365 - Fixed the kvm-ifup script
366 - Fixed internal error handling for special job cases
367 - Updated man page to specify the escaping feature for options
373 *(Released Mon, 28 Feb 2011)*
375 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
376 that this feature wiped the disks of the wrong instance, leading to loss
381 - Fixed title of query field containing instance name
382 - Expanded the glossary in the documentation
383 - Fixed one unittest (internal issue)
389 *(Released Mon, 21 Feb 2011)*
391 A number of bug fixes plus just a couple functionality changes.
393 On the user-visible side, the ``gnt-* list`` command output has changed
394 with respect to "special" field states. The current rc1 style of display
395 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
396 the default output mode special fields states are displayed as follows:
398 - Offline resource: ``*``
399 - Unavailable/not applicable: ``-``
400 - Data missing (RPC failure): ``?``
401 - Unknown field: ``??``
403 Another user-visible change is the addition of ``--force-join`` to
408 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
409 - Fixed regression in RAPI/instance reinstall where all parameters were
410 required (instead of optional)
411 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
412 - Fixed iallocator usage (offline nodes were not considered offline)
413 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
414 - Fixed hypervisor and OS parameter validation with respect to
416 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
418 - Fixed ``tools/listrunner`` with respect to agent-based usage
424 *(Released Fri, 4 Feb 2011)*
426 Many changes and fixes since the beta1 release. While there were some
427 internal changes, the code has been mostly stabilised for the RC
430 Note: the dumb allocator was removed in this release, as it was not kept
431 up-to-date with the IAllocator protocol changes. It is recommended to
432 use the ``hail`` command from the ganeti-htools package.
434 Note: the 2.4 and up versions of Ganeti are not compatible with the
435 0.2.x branch of ganeti-htools. You need to upgrade to
436 ganeti-htools-0.3.0 (or later).
438 Regressions fixed from 2.3
439 ~~~~~~~~~~~~~~~~~~~~~~~~~~
441 - Fixed the ``gnt-cluster verify-disks`` command
442 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
444 - Fixed disk adoption breakage
445 - Fixed wrong headers in instance listing for field aliases
450 - Fixed corner case in KVM handling of NICs
451 - Fixed many cases of wrong handling of non-vm_capable nodes
452 - Fixed a bug where a missing instance symlink was not possible to
453 recreate with any ``gnt-*`` command (now ``gnt-instance
454 activate-disks`` does it)
455 - Fixed the volume group name as reported by ``gnt-cluster
457 - Increased timeouts for the import-export code, hopefully leading to
458 fewer aborts due network or instance timeouts
459 - Fixed bug in ``gnt-node list-storage``
460 - Fixed bug where not all daemons were started on cluster
461 initialisation, but only at the first watcher run
462 - Fixed many bugs in the OOB implementation
463 - Fixed watcher behaviour in presence of instances with offline
465 - Fixed instance list output for instances running on the wrong node
466 - a few fixes to the cluster-merge tool, but it still cannot merge
467 multi-node groups (currently it is not recommended to use this tool)
473 - Improved network configuration for the KVM hypervisor
474 - Added e1000 as a supported NIC for Xen-HVM
475 - Improved the lvmstrap tool to also be able to use partitions, as
476 opposed to full disks
477 - Improved speed of disk wiping (the cluster parameter
478 ``prealloc_wipe_disks``, so that it has a low impact on the total time
479 of instance creations
480 - Added documentation for the OS parameters
481 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
482 hypervisor is not responding
483 - Added display of blacklisted and hidden OS information in
485 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
486 NIC and node parameters, which might create problems with currently
487 invalid (but undetected) configuration files, but prevents validation
488 failures when unrelated parameters are modified
489 - Changed cluster initialisation to wait for the master daemon to become
491 - Expanded the RAPI interface:
493 - Added config redistribution resource
494 - Added activation/deactivation of instance disks
495 - Added export of console information
497 - Implemented log file reopening on SIGHUP, which allows using
498 logrotate(8) for the Ganeti log files
499 - Added a basic OOB helper script as an example
505 *(Released Fri, 14 Jan 2011)*
510 - Fixed timezone issues when formatting timestamps
511 - Added support for node groups, available via ``gnt-group`` and other
513 - Added out-of-band framework and management, see :doc:`design
514 document <design-oob>`
515 - Removed support for roman numbers from ``gnt-node list`` and
516 ``gnt-instance list``.
517 - Allowed modification of master network interface via ``gnt-cluster
518 modify --master-netdev``
519 - Accept offline secondaries while shutting down instance disks
520 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
521 - Added support for multiple LVM volume groups
522 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
524 - Added commands to list available fields:
526 - ``gnt-node list-fields``
527 - ``gnt-group list-fields``
528 - ``gnt-instance list-fields``
530 - Updated documentation and man pages
535 - Moved ``rapi_users`` file into separate directory, now named
536 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
538 - Added new tool for running commands on many machines,
539 ``tools/ganeti-listrunner``
540 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
541 improving the ``gnt-instance console`` output
542 - Allowed customisation of disk index separator at ``configure`` time
543 - Export node group allocation policy to :doc:`iallocator <iallocator>`
544 - Added support for non-partitioned md disks in ``lvmstrap``
545 - Added script to gracefully power off KVM instances
546 - Split ``utils`` module into smaller parts
547 - Changed query operations to return more detailed information, e.g.
548 whether an information is unavailable due to an offline node. To use
549 this new functionality, the LUXI call ``Query`` must be used. Field
550 information is now stored by the master daemon and can be retrieved
551 using ``QueryFields``. Instances, nodes and groups can also be queried
552 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
553 exposed via RAPI). The following commands make use of this
554 infrastructure change:
557 - ``gnt-group list-fields``
559 - ``gnt-node list-fields``
560 - ``gnt-instance list``
561 - ``gnt-instance list-fields``
562 - ``gnt-debug locks``
567 - New RAPI resources (see :doc:`rapi`):
571 - ``/2/groups/[group_name]``
572 - ``/2/groups/[group_name]/assign-nodes``
573 - ``/2/groups/[group_name]/modify``
574 - ``/2/groups/[group_name]/rename``
575 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
579 - Implemented ``no_install`` for instance creation
580 - Implemented OS parameters for instance reinstallation, allowing
581 use of special settings on reinstallation (e.g. for preserving data)
586 - Added IPv6 support in import/export
587 - Pause DRBD synchronization while wiping disks on instance creation
588 - Updated unittests and QA scripts
589 - Improved network parameters passed to KVM
590 - Converted man pages from docbook to reStructuredText
596 *(Released Mon, 20 Dec 2010)*
598 Released version 2.3.1~rc1 without any changes.
604 *(Released Wed, 1 Dec 2010)*
606 - impexpd: Disable OpenSSL compression in socat if possible (backport
607 from master, commit e90739d625b, see :doc:`installation guide
608 <install-quick>` for details)
609 - Changed unittest coverage report to exclude test scripts
610 - Added script to check version format
616 *(Released Wed, 1 Dec 2010)*
618 Released version 2.3.0~rc1 without any changes.
624 *(Released Fri, 19 Nov 2010)*
626 A number of bugfixes and documentation updates:
628 - Update ganeti-os-interface documentation
629 - Fixed a bug related to duplicate MACs or similar items which should be
631 - Fix breakage in OS state modify
632 - Reinstall instance: disallow offline secondaries (fixes bug related to
633 OS changing but reinstall failing)
634 - plus all the other fixes between 2.2.1 and 2.2.2
640 *(Released Tue, 2 Nov 2010)*
642 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
643 - Fixed master failover race with watcher
644 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
646 - Accept previously stopped instance for export with instance removal
647 - Simplify and extend the environment variables for instance OS scripts
648 - Added new node flags, ``master_capable`` and ``vm_capable``
649 - Added optional instance disk wiping prior during allocation. This is a
650 cluster-wide option and can be set/modified using
651 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
652 - Added IPv6 support, see :doc:`design document <design-2.3>` and
654 - Added a new watcher option (``--ignore-pause``)
655 - Added option to ignore offline node on instance start/stop
656 (``--ignore-offline``)
657 - Allow overriding OS parameters with ``gnt-instance reinstall``
658 - Added ability to change node's secondary IP address using ``gnt-node
660 - Implemented privilege separation for all daemons except
661 ``ganeti-noded``, see ``configure`` options
662 - Complain if an instance's disk is marked faulty in ``gnt-cluster
664 - Implemented job priorities (see ``ganeti(7)`` manpage)
665 - Ignore failures while shutting down instances during failover from
667 - Exit daemon's bootstrap process only once daemon is ready
668 - Export more information via ``LUInstanceQuery``/remote API
669 - Improved documentation, QA and unittests
670 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
671 restart if the file was created or changed
672 - Added LUXI protocol version sent with each request and response,
673 allowing detection of server/client mismatches
674 - Moved the Python scripts among gnt-* and ganeti-* into modules
675 - Moved all code related to setting up SSH to an external script,
677 - Infrastructure changes for node group support in future versions
683 *(Released Fri, 19 Nov 2010)*
685 A few small bugs fixed, and some improvements to the build system:
687 - Fix documentation regarding conversion to drbd
688 - Fix validation of parameters in cluster modify (``gnt-cluster modify
690 - Fix error handling in node modify with multiple changes
691 - Allow remote imports without checked names
697 *(Released Tue, 19 Oct 2010)*
699 - Disable SSL session ID cache in RPC client
705 *(Released Thu, 14 Oct 2010)*
707 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
708 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
710 - Fix problems with interaction between Curl and Python's HTTP server,
711 resulting in increased speed in many RPC calls
712 - Improve our release script to prevent breakage with older aclocal and
719 *(Released Thu, 7 Oct 2010)*
721 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
722 value retrieved from master
723 - Added support for blacklisted or hidden OS definitions
724 - Added simple lock monitor (accessible via (``gnt-debug locks``)
725 - Added support for -mem-path in KVM hypervisor abstraction layer
726 - Allow overriding instance parameters in tool for inter-cluster
727 instance moves (``tools/move-instance``)
728 - Improved opcode summaries (e.g. in ``gnt-job list``)
729 - Improve consistency of OS listing by sorting it
730 - Documentation updates
736 *(Released Fri, 8 Oct 2010)*
738 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
744 *(Released Mon, 4 Oct 2010)*
746 - Fixed regression in ``gnt-instance rename``
752 *(Released Wed, 22 Sep 2010)*
754 - Fixed OS_VARIANT variable for OS scripts
755 - Fixed cluster tag operations via RAPI
756 - Made ``setup-ssh`` exit with non-zero code if an error occurred
757 - Disabled RAPI CA checks in watcher
763 *(Released Mon, 23 Aug 2010)*
765 - Support DRBD versions of the format "a.b.c.d"
767 - Re-introduce support for usage from multiple threads in RAPI client
768 - Instance renames and modify via RAPI
769 - Work around race condition between processing and archival in job
771 - Mark opcodes following failed one as failed, too
772 - Job field ``lock_status`` was removed due to difficulties making it
773 work with the changed job queue in Ganeti 2.2; a better way to monitor
774 locks is expected for a later 2.2.x release
775 - Fixed dry-run behaviour with many commands
776 - Support ``ssh-agent`` again when adding nodes
777 - Many additional bugfixes
783 *(Released Fri, 30 Jul 2010)*
785 Important change: the internal RPC mechanism between Ganeti nodes has
786 changed from using a home-grown http library (based on the Python base
787 libraries) to use the PycURL library. This requires that PycURL is
788 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
789 against GnuTLS by default. cURL's support for GnuTLS had known issues
790 before cURL 7.21.0 and we recommend using the latest cURL release or
791 linking against OpenSSL. Most other distributions already link PycURL
792 and cURL against OpenSSL. The command::
794 python -c 'import pycurl; print pycurl.version'
796 can be used to determine the libraries PycURL and cURL are linked
799 Other significant changes:
801 - Rewrote much of the internals of the job queue, in order to achieve
802 better parallelism; this decouples job query operations from the job
803 processing, and it should allow much nicer behaviour of the master
804 daemon under load, and it also has uncovered some long-standing bugs
805 related to the job serialisation (now fixed)
806 - Added a default iallocator setting to the cluster parameters,
807 eliminating the need to always pass nodes or an iallocator for
808 operations that require selection of new node(s)
809 - Added experimental support for the LXC virtualization method
810 - Added support for OS parameters, which allows the installation of
811 instances to pass parameter to OS scripts in order to customise the
813 - Added a hypervisor parameter controlling the migration type (live or
814 non-live), since hypervisors have various levels of reliability; this
815 has renamed the 'live' parameter to 'mode'
816 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
817 logical volumes, meaning that cluster verify will ignore them and not
818 flag their presence as errors
819 - The watcher will now reset the error count for failed instances after
820 8 hours, thus allowing self-healing if the problem that caused the
821 instances to be down/fail to start has cleared in the meantime
822 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
823 check for, and warn, if the drbd module parameter ``usermode_helper``
824 is not consistent with the cluster-wide setting; this is needed to
825 make diagnose easier of failed drbd creations
826 - Started adding base IPv6 support, but this is not yet
827 enabled/available for use
828 - Rename operations (cluster, instance) will now return the new name,
829 which is especially useful if a short name was passed in
830 - Added support for instance migration in RAPI
831 - Added a tool to pre-configure nodes for the SSH setup, before joining
832 them to the cluster; this will allow in the future a simplified model
833 for node joining (but not yet fully enabled in 2.2); this needs the
834 paramiko python library
835 - Fixed handling of name-resolving errors
836 - Fixed consistency of job results on the error path
837 - Fixed master-failover race condition when executed multiple times in
839 - Fixed many bugs related to the job queue (mostly introduced during the
840 2.2 development cycle, so not all are impacting 2.1)
841 - Fixed instance migration with missing disk symlinks
842 - Fixed handling of unknown jobs in ``gnt-job archive``
843 - And many other small fixes/improvements
847 - Enhanced both the unittest and the QA coverage
848 - Switched the opcode validation to a generic model, and extended the
849 validation to all opcode parameters
850 - Changed more parts of the code that write shell scripts to use the
852 - Switched the master daemon to use the asyncore library for the Luxi
859 *(Released Thu, 17 Jun 2010)*
861 - Added tool (``move-instance``) and infrastructure to move instances
862 between separate clusters (see :doc:`separate documentation
863 <move-instance>` and :doc:`design document <design-2.2>`)
864 - Added per-request RPC timeout
865 - RAPI now requires a Content-Type header for requests with a body (e.g.
866 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
867 :rfc:`2616` (HTTP/1.1), section 7.2.1)
868 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
870 - Implemented initial support for running Ganeti daemons as separate
871 users, see configure-time flags ``--with-user-prefix`` and
872 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
874 - Instances can be removed after export (``gnt-backup export
876 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
877 key (instead of 1024 bit)
878 - Added new cluster configuration file for cluster domain secret
879 - Import/export now use SSL instead of SSH
880 - Added support for showing estimated time when exporting an instance,
881 see the ``ganeti-os-interface(7)`` manpage and look for
888 *(Released Tue, 16 Nov 2010)*
890 Some more bugfixes. Unless critical bugs occur, this will be the last
893 - Fix case of MAC special-values
894 - Fix mac checker regex
895 - backend: Fix typo causing "out of range" error
896 - Add missing --units in gnt-instance list man page
902 *(Released Tue, 24 Aug 2010)*
905 - Don't ignore secondary node silently on non-mirrored disk templates
907 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
908 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
909 - Properly document the usb_mouse parameter
910 - Fix path in ganeti-rapi(8) (issue 116)
911 - Adjust error message when the ganeti user's .ssh directory is
913 - Add same-node-check when changing the disk template to drbd
919 *(Released Fri, 16 Jul 2010)*
922 - Add an option to only select some reboot types during qa/burnin.
923 (on some hypervisors consequent reboots are not supported)
924 - Fix infrequent race condition in master failover. Sometimes the old
925 master ip address would be still detected as up for a short time
926 after it was removed, causing failover to fail.
927 - Decrease mlockall warnings when the ctypes module is missing. On
928 Python 2.4 we support running even if no ctypes module is installed,
929 but we were too verbose about this issue.
930 - Fix building on old distributions, on which man doesn't have a
932 - Fix RAPI not to ignore the MAC address on instance creation
933 - Implement the old instance creation format in the RAPI client.
939 *(Released Thu, 01 Jul 2010)*
941 A small bugfix release:
942 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
943 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
944 a non-existing option
945 - Fix instance create: the --force-variant option was ignored
946 - Improve pylint 0.21 compatibility and warnings with Python 2.6
947 - Fix modify node storage with non-FQDN arguments
948 - Fix RAPI client to authenticate under Python 2.6 when used
949 for more than 5 requests needing authentication
950 - Fix gnt-instance modify -t (storage) giving a wrong error message
951 when converting a non-shutdown drbd instance to plain
957 *(Released Fri, 18 Jun 2010)*
959 A small bugfix release:
961 - Fix live migration of KVM instances started with older Ganeti
962 versions which had fewer hypervisor parameters
963 - Fix gnt-instance grow-disk on down instances
964 - Fix an error-reporting bug during instance migration
965 - Better checking of the ``--net`` and ``--disk`` values, to avoid
966 silently ignoring broken ones
967 - Fix an RPC error reporting bug affecting, for example, RAPI client
969 - Fix bug triggered by different API version os-es on different nodes
970 - Fix a bug in instance startup with custom hvparams: OS level
971 parameters would fail to be applied.
972 - Fix the RAPI client under Python 2.6 (but more work is needed to
973 make it work completely well with OpenSSL)
974 - Fix handling of errors when resolving names from DNS
980 *(Released Thu, 3 Jun 2010)*
982 A medium sized development cycle. Some new features, and some
983 fixes/small improvements/cleanups.
988 The node deamon now tries to mlock itself into memory, unless the
989 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
990 its logs, and falls back to console logging. This allows emergency
991 features such as ``gnt-node powercycle`` to work even in the event of a
992 broken node disk (tested offlining the disk hosting the node's
993 filesystem and dropping its memory caches; don't try this at home)
995 KVM: add vhost-net acceleration support. It can be tested with a new
996 enough version of the kernel and of qemu-kvm.
998 KVM: Add instance chrooting feature. If you use privilege dropping for
999 your VMs you can also now force them to chroot to an empty directory,
1000 before starting the emulated guest.
1002 KVM: Add maximum migration bandwith and maximum downtime tweaking
1003 support (requires a new-enough version of qemu-kvm).
1005 Cluster verify will now warn if the master node doesn't have the master
1006 ip configured on it.
1008 Add a new (incompatible) instance creation request format to RAPI which
1009 supports all parameters (previously only a subset was supported, and it
1010 wasn't possible to extend the old format to accomodate all the new
1011 features. The old format is still supported, and a client can check for
1012 this feature, before using it, by checking for its presence in the
1013 ``features`` RAPI resource.
1015 Now with ancient latin support. Try it passing the ``--roman`` option to
1016 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
1017 (requires the python-roman module to be installed, in order to work).
1022 As usual many internal code refactorings, documentation updates, and
1025 - Lots of improvements and cleanups to the experimental Remote API
1026 (RAPI) client library.
1027 - A new unit test suite for the core daemon libraries.
1028 - A fix to creating missing directories makes sure the umask is not
1029 applied anymore. This enforces the same directory permissions
1031 - Better handling terminating daemons with ctrl+c (used when running
1032 them in debugging mode).
1033 - Fix a race condition in live migrating a KVM instance, when stat()
1034 on the old proc status file returned EINVAL, which is an unexpected
1036 - Fixed manpage checking with newer man and utf-8 charachters. But now
1037 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
1043 *(Released Fri, 7 May 2010)*
1045 Fix a bug which prevented untagged KVM instances from starting.
1051 *(Released Fri, 7 May 2010)*
1053 Another release with a long development cycle, during which many
1054 different features were added.
1056 Significant features
1057 ~~~~~~~~~~~~~~~~~~~~
1059 The KVM hypervisor now can run the individual instances as non-root, to
1060 reduce the impact of a VM being hijacked due to bugs in the
1061 hypervisor. It is possible to run all instances as a single (non-root)
1062 user, to manually specify a user for each instance, or to dynamically
1063 allocate a user out of a cluster-wide pool to each instance, with the
1064 guarantee that no two instances will run under the same user ID on any
1067 An experimental RAPI client library, that can be used standalone
1068 (without the other Ganeti libraries), is provided in the source tree as
1069 ``lib/rapi/client.py``. Note this client might change its interface in
1070 the future, as we iterate on its capabilities.
1072 A new command, ``gnt-cluster renew-crypto`` has been added to easily
1073 replace the cluster's certificates and crypto keys. This might help in
1074 case they have been compromised, or have simply expired.
1076 A new disk option for instance creation has been added that allows one
1077 to "adopt" currently existing logical volumes, with data
1078 preservation. This should allow easier migration to Ganeti from
1079 unmanaged (or managed via other software) instances.
1081 Another disk improvement is the possibility to convert between redundant
1082 (DRBD) and plain (LVM) disk configuration for an instance. This should
1083 allow better scalability (starting with one node and growing the
1084 cluster, or shrinking a two-node cluster to one node).
1086 A new feature that could help with automated node failovers has been
1087 implemented: if a node sees itself as offline (by querying the master
1088 candidates), it will try to shutdown (hard) all instances and any active
1089 DRBD devices. This reduces the risk of duplicate instances if an
1090 external script automatically failovers the instances on such nodes. To
1091 enable this, the cluster parameter ``maintain_node_health`` should be
1092 enabled; in the future this option (per the name) will enable other
1093 automatic maintenance features.
1095 Instance export/import now will reuse the original instance
1096 specifications for all parameters; that means exporting an instance,
1097 deleting it and the importing it back should give an almost identical
1098 instance. Note that the default import behaviour has changed from
1099 before, where it created only one NIC; now it recreates the original
1102 Cluster verify has added a few new checks: SSL certificates validity,
1103 /etc/hosts consistency across the cluster, etc.
1108 As usual, many internal changes were done, documentation fixes,
1111 - Fixed cluster initialization with disabled cluster storage (regression
1112 introduced in 2.1.1)
1113 - File-based storage supports growing the disks
1114 - Fixed behaviour of node role changes
1115 - Fixed cluster verify for some corner cases, plus a general rewrite of
1116 cluster verify to allow future extension with more checks
1117 - Fixed log spamming by watcher and node daemon (regression introduced
1119 - Fixed possible validation issues when changing the list of enabled
1121 - Fixed cleanup of /etc/hosts during node removal
1122 - Fixed RAPI response for invalid methods
1123 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
1124 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
1125 from ide disks, etc.)
1126 - Allow OS changes without re-installation (to record a changed OS
1127 outside of Ganeti, or to allow OS renames)
1128 - Allow instance creation without OS installation (useful for example if
1129 the OS will be installed manually, or restored from a backup not in
1131 - Implemented option to make cluster ``copyfile`` use the replication
1133 - Added list of enabled hypervisors to ssconf (possibly useful for
1135 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
1137 - A partial form of node re-IP is possible via node readd, which now
1138 allows changed node primary IP
1139 - Command line utilities now show an informational message if the job is
1141 - The logs of the master daemon now show the PID/UID/GID of the
1148 *(Released Fri, 12 Mar 2010)*
1150 During the 2.1.0 long release candidate cycle, a lot of improvements and
1151 changes have accumulated with were released later as 2.1.1.
1156 The node evacuate command (``gnt-node evacuate``) was significantly
1157 rewritten, and as such the IAllocator protocol was changed - a new
1158 request type has been added. This unfortunate change during a stable
1159 series is designed to improve performance of node evacuations; on
1160 clusters with more than about five nodes and which are well-balanced,
1161 evacuation should proceed in parallel for all instances of the node
1162 being evacuated. As such, any existing IAllocator scripts need to be
1163 updated, otherwise the above command will fail due to the unknown
1164 request. The provided "dumb" allocator has not been updated; but the
1165 ganeti-htools package supports the new protocol since version 0.2.4.
1167 Another important change is increased validation of node and instance
1168 names. This might create problems in special cases, if invalid host
1169 names are being used.
1171 Also, a new layer of hypervisor parameters has been added, that sits at
1172 OS level between the cluster defaults and the instance ones. This allows
1173 customisation of virtualization parameters depending on the installed
1174 OS. For example instances with OS 'X' may have a different KVM kernel
1175 (or any other parameter) than the cluster defaults. This is intended to
1176 help managing a multiple OSes on the same cluster, without manual
1177 modification of each instance's parameters.
1179 A tool for merging clusters, ``cluster-merge``, has been added in the
1180 tools sub-directory.
1185 - Improved the int/float conversions that should make the code more
1186 robust in face of errors from the node daemons
1187 - Fixed the remove node code in case of internal configuration errors
1188 - Fixed the node daemon behaviour in face of inconsistent queue
1189 directory (e.g. read-only file-system where we can't open the files
1191 - Fixed the behaviour of gnt-node modify for master candidate demotion;
1192 now it either aborts cleanly or, if given the new "auto_promote"
1193 parameter, will automatically promote other nodes as needed
1194 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
1195 completely prevent Ganeti from working
1196 - Fixed bug for instance export when not all disks were successfully
1198 - Fixed behaviour of node add when the new node is slow in starting up
1200 - Fixed handling of signals in the LUXI client, which should improve
1201 behaviour of command-line scripts
1202 - Added checks for invalid node/instance names in the configuration (now
1203 flagged during cluster verify)
1204 - Fixed watcher behaviour for disk activation errors
1205 - Fixed two potentially endless loops in http library, which led to the
1206 RAPI daemon hanging and consuming 100% CPU in some cases
1207 - Fixed bug in RAPI daemon related to hashed passwords
1208 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
1210 - Enhanced compatibility with non-Debian OSes, but not using absolute
1211 path in some commands and allowing customisation of the ssh
1212 configuration directory
1213 - Fixed possible future issue with new Python versions by abiding to the
1214 proper use of ``__slots__`` attribute on classes
1215 - Added checks that should prevent directory traversal attacks
1216 - Many documentation fixes based on feedback from users
1221 - Added an "early_release" more for instance replace disks and node
1222 evacuate, where we release locks earlier and thus allow higher
1223 parallelism within the cluster
1224 - Added watcher hooks, intended to allow the watcher to restart other
1225 daemons (e.g. from the ganeti-nbma project), but they can be used of
1226 course for any other purpose
1227 - Added a compile-time disable for DRBD barriers, to increase
1228 performance if the administrator trusts the power supply or the
1229 storage system to not lose writes
1230 - Added the option of using syslog for logging instead of, or in
1231 addition to, Ganeti's own log files
1232 - Removed boot restriction for paravirtual NICs for KVM, recent versions
1233 can indeed boot from a paravirtual NIC
1234 - Added a generic debug level for many operations; while this is not
1235 used widely yet, it allows one to pass the debug value all the way to
1237 - Enhanced the hooks environment for instance moves (failovers,
1238 migrations) where the primary/secondary nodes changed during the
1239 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
1240 - Enhanced data validations for many user-supplied values; one important
1241 item is the restrictions imposed on instance and node names, which
1242 might reject some (invalid) host names
1243 - Add a configure-time option to disable file-based storage, if it's not
1244 needed; this allows greater security separation between the master
1245 node and the other nodes from the point of view of the inter-node RPC
1247 - Added user notification in interactive tools if job is waiting in the
1248 job queue or trying to acquire locks
1249 - Added log messages when a job is waiting for locks
1250 - Added filtering by node tags in instance operations which admit
1251 multiple instances (start, stop, reboot, reinstall)
1252 - Added a new tool for cluster mergers, ``cluster-merge``
1253 - Parameters from command line which are of the form ``a=b,c=d`` can now
1254 use backslash escapes to pass in values which contain commas,
1255 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
1257 - For KVM, the instance name is the first parameter passed to KVM, so
1258 that it's more visible in the process list
1264 *(Released Tue, 2 Mar 2010)*
1266 Ganeti 2.1 brings many improvements with it. Major changes:
1268 - Added infrastructure to ease automated disk repairs
1269 - Added new daemon to export configuration data in a cheaper way than
1270 using the remote API
1271 - Instance NICs can now be routed instead of being associated with a
1273 - Improved job locking logic to reduce impact of jobs acquiring multiple
1274 locks waiting for other long-running jobs
1276 In-depth implementation details can be found in the Ganeti 2.1 design
1282 - Added chroot hypervisor
1283 - Added more options to xen-hvm hypervisor (``kernel_path`` and
1285 - Added more options to xen-pvm hypervisor (``use_bootloader``,
1286 ``bootloader_path`` and ``bootloader_args``)
1287 - Added the ``use_localtime`` option for the xen-hvm and kvm
1288 hypervisors, and the default value for this has changed to false (in
1289 2.0 xen-hvm always enabled it)
1290 - Added luxi call to submit multiple jobs in one go
1291 - Added cluster initialization option to not modify ``/etc/hosts``
1293 - Added network interface parameters
1294 - Added dry run mode to some LUs
1295 - Added RAPI resources:
1297 - ``/2/instances/[instance_name]/info``
1298 - ``/2/instances/[instance_name]/replace-disks``
1299 - ``/2/nodes/[node_name]/evacuate``
1300 - ``/2/nodes/[node_name]/migrate``
1301 - ``/2/nodes/[node_name]/role``
1302 - ``/2/nodes/[node_name]/storage``
1303 - ``/2/nodes/[node_name]/storage/modify``
1304 - ``/2/nodes/[node_name]/storage/repair``
1306 - Added OpCodes to evacuate or migrate all instances on a node
1307 - Added new command to list storage elements on nodes (``gnt-node
1308 list-storage``) and modify them (``gnt-node modify-storage``)
1309 - Added new ssconf files with master candidate IP address
1310 (``ssconf_master_candidates_ips``), node primary IP address
1311 (``ssconf_node_primary_ips``) and node secondary IP address
1312 (``ssconf_node_secondary_ips``)
1313 - Added ``ganeti-confd`` and a client library to query the Ganeti
1314 configuration via UDP
1315 - Added ability to run hooks after cluster initialization and before
1317 - Added automatic mode for disk replace (``gnt-instance replace-disks
1319 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
1320 after catastrophic data-loss
1321 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
1323 - Added ``gnt-instance move`` command to move instances
1324 - Added ``gnt-cluster watcher`` command to control watcher
1325 - Added ``gnt-node powercycle`` command to powercycle nodes
1326 - Added new job status field ``lock_status``
1327 - Added parseable error codes to cluster verification (``gnt-cluster
1328 verify --error-codes``) and made output less verbose (use
1329 ``--verbose`` to restore previous behaviour)
1330 - Added UUIDs to the main config entities (cluster, nodes, instances)
1331 - Added support for OS variants
1332 - Added support for hashed passwords in the Ganeti remote API users file
1334 - Added option to specify maximum timeout on instance shutdown
1335 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
1336 - Added new helper script to start and stop Ganeti daemons
1337 (``daemon-util``), with the intent to reduce the work necessary to
1338 adjust Ganeti for non-Debian distributions and to start/stop daemons
1340 - Added more unittests
1341 - Fixed critical bug in ganeti-masterd startup
1342 - Removed the configure-time ``kvm-migration-port`` parameter, this is
1343 now customisable at the cluster level for both the KVM and Xen
1344 hypervisors using the new ``migration_port`` parameter
1345 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
1346 reinstalling an instance
1347 - Allowed ``@`` in tag names
1348 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
1349 - Many documentation updates
1350 - Distribute hypervisor files on ``gnt-cluster redist-conf``
1351 - ``gnt-instance reinstall`` can now reinstall multiple instances
1352 - Updated many command line parameters
1353 - Introduced new OS API version 15
1354 - No longer support a default hypervisor
1355 - Treat virtual LVs as inexistent
1356 - Improved job locking logic to reduce lock contention
1357 - Match instance and node names case insensitively
1358 - Reimplemented bash completion script to be more complete
1365 *(Released Thu, 4 Feb 2010)*
1367 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
1368 - Fix a string formatting bug
1369 - Improve safety of the code in some error paths
1370 - Improve data validation in the master of values returned from nodes
1376 *(Released Thu, 17 Dec 2009)*
1378 - Fix security issue due to missing validation of iallocator names; this
1379 allows local and remote execution of arbitrary executables
1380 - Fix failure of gnt-node list during instance removal
1381 - Ship the RAPI documentation in the archive
1387 *(Released Wed, 30 Sep 2009)*
1389 - Fixed many wrong messages
1390 - Fixed a few bugs related to the locking library
1391 - Fixed MAC checking at instance creation time
1392 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
1393 - Fixed a few issues related to signal handling in both daemons and
1395 - Fixed the example startup script provided
1396 - Fixed insserv dependencies in the example startup script (patch from
1398 - Fixed handling of drained nodes in the iallocator framework
1399 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
1401 - Fixed error related to invalid job IDs in job polling
1402 - Fixed job/opcode persistence on unclean master shutdown
1403 - Fixed handling of partial job processing after unclean master
1405 - Fixed error reporting from LUs, previously all errors were converted
1406 into execution errors
1407 - Fixed error reporting from burnin
1408 - Decreased significantly the memory usage of the job queue
1409 - Optimised slightly multi-job submission
1410 - Optimised slightly opcode loading
1411 - Backported the multi-job submit framework from the development
1412 branch; multi-instance start and stop should be faster
1413 - Added script to clean archived jobs after 21 days; this will reduce
1414 the size of the queue directory
1415 - Added some extra checks in disk size tracking
1416 - Added an example ethers hook script
1417 - Added a cluster parameter that prevents Ganeti from modifying of
1419 - Added more node information to RAPI responses
1420 - Added a ``gnt-job watch`` command that allows following the ouput of a
1422 - Added a bind-address option to ganeti-rapi
1423 - Added more checks to the configuration verify
1424 - Enhanced the burnin script such that some operations can be retried
1426 - Converted instance reinstall to multi-instance model
1432 *(Released Fri, 7 Aug 2009)*
1434 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
1435 to allow using the pre-2.0.2 behaviour in activation, if any existing
1436 instances have mismatched disk sizes in the configuration
1437 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
1438 any configuration mismatches for disk sizes
1439 - Added ``gnt-master cluste-failover --no-voting`` to allow master
1440 failover to work on two-node clusters
1441 - Fixed the ``--net`` option of ``gnt-backup import``, which was
1443 - Fixed detection of OS script errors in ``gnt-backup export``
1444 - Fixed exit code of ``gnt-backup export``
1450 *(Released Fri, 17 Jul 2009)*
1452 - Added experimental support for stripped logical volumes; this should
1453 enhance performance but comes with a higher complexity in the block
1454 device handling; stripping is only enabled when passing
1455 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
1456 affected even in the non-stripped mode
1457 - Improved resiliency against transient failures at the end of DRBD
1458 resyncs, and in general of DRBD resync checks
1459 - Fixed a couple of issues with exports and snapshot errors
1460 - Fixed a couple of issues in instance listing
1461 - Added display of the disk size in ``gnt-instance info``
1462 - Fixed checking for valid OSes in instance creation
1463 - Fixed handling of the "vcpus" parameter in instance listing and in
1464 general of invalid parameters
1465 - Fixed http server library, and thus RAPI, to handle invalid
1466 username/password combinations correctly; this means that now they
1467 report unauthorized for queries too, not only for modifications,
1468 allowing earlier detect of configuration problems
1469 - Added a new "role" node list field, equivalent to the master/master
1470 candidate/drained/offline flags combinations
1471 - Fixed cluster modify and changes of candidate pool size
1472 - Fixed cluster verify error messages for wrong files on regular nodes
1473 - Fixed a couple of issues with node demotion from master candidate role
1474 - Fixed node readd issues
1475 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
1476 - Added a new ``--no-voting`` option for masterfailover to fix failover
1477 on two-nodes clusters when the former master node is unreachable
1478 - Added instance reinstall over RAPI
1484 *(Released Tue, 16 Jun 2009)*
1486 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
1487 allow re-adding the start in single-user option (regression from 1.2)
1488 - the watcher writes the instance status to a file, to allow monitoring
1489 to report the instance status (from the master) based on cached
1490 results of the watcher's queries; while this can get stale if the
1491 watcher is being locked due to other work on the cluster, this is
1492 still an improvement
1493 - the watcher now also restarts the node daemon and the rapi daemon if
1495 - fixed the watcher to handle full and drained queue cases
1496 - hooks export more instance data in the environment, which helps if
1497 hook scripts need to take action based on the instance's properties
1498 (no longer need to query back into ganeti)
1499 - instance failovers when the instance is stopped do not check for free
1500 RAM, so that failing over a stopped instance is possible in low memory
1502 - rapi uses queries for tags instead of jobs (for less job traffic), and
1503 for cluster tags it won't talk to masterd at all but read them from
1505 - a couple of error handling fixes in RAPI
1506 - drbd handling: improved the error handling of inconsistent disks after
1507 resync to reduce the frequency of "there are some degraded disks for
1508 this instance" messages
1509 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
1510 error handling path called a wrong function name)
1516 *(Released Wed, 27 May 2009)*
1518 - no changes from rc5
1524 *(Released Wed, 20 May 2009)*
1526 - fix a couple of bugs (validation, argument checks)
1527 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
1528 - some small improvements to RAPI and IAllocator
1529 - make watcher automatically start the master daemon if down
1535 *(Released Mon, 27 Apr 2009)*
1537 - change the OS list to not require locks; this helps with big clusters
1538 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
1539 volume group is broken
1540 - ``gnt-instance info``, without any arguments, doesn't run for all
1541 instances anymore; either pass ``--all`` or pass the desired
1542 instances; this helps against mistakes on big clusters where listing
1543 the information for all instances takes a long time
1544 - miscellaneous doc and man pages fixes
1550 *(Released Wed, 8 Apr 2009)*
1552 - Change the internal locking model of some ``gnt-node`` commands, in
1553 order to reduce contention (and blocking of master daemon) when
1554 batching many creation/reinstall jobs
1555 - Fixes to Xen soft reboot
1556 - No longer build documentation at build time, instead distribute it in
1557 the archive, in order to reduce the need for the whole docbook/rst
1564 *(Released Fri, 27 Mar 2009)*
1566 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
1567 - Fix watcher startup sequence, improves the behaviour of busy clusters
1568 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
1569 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
1570 bind address and other places
1571 - Some documentation fixes and updates
1577 *(Released Mon, 2 Mar 2009)*
1579 - More documentation updates, now all docs should be more-or-less
1581 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
1583 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
1585 - Fix an issue related to $libdir/run/ganeti and cluster creation
1591 *(Released Thu, 19 Feb 2009)*
1593 - Xen PVM and KVM have switched the default value for the instance root
1594 disk to the first partition on the first drive, instead of the whole
1595 drive; this means that the OS installation scripts must be changed
1597 - Man pages have been updated
1598 - RAPI has been switched by default to HTTPS, and the exported functions
1599 should all work correctly
1600 - RAPI v1 has been removed
1601 - Many improvements to the KVM hypervisor
1602 - Block device errors are now better reported
1603 - Many other bugfixes and small improvements
1609 *(Released Mon, 26 Jan 2009)*
1611 - Version 2 is a general rewrite of the code and therefore the
1612 differences are too many to list, see the design document for 2.0 in
1613 the ``doc/`` subdirectory for more details
1614 - In this beta version there is not yet a migration path from 1.2 (there
1615 will be one in the final 2.0 release)
1616 - A few significant changes are:
1618 - all commands are executed by a daemon (``ganeti-masterd``) and the
1619 various ``gnt-*`` commands are just front-ends to it
1620 - all the commands are entered into, and executed from a job queue,
1621 see the ``gnt-job(8)`` manpage
1622 - the RAPI daemon supports read-write operations, secured by basic
1623 HTTP authentication on top of HTTPS
1624 - DRBD version 0.7 support has been removed, DRBD 8 is the only
1625 supported version (when migrating from Ganeti 1.2 to 2.0, you need
1626 to migrate to DRBD 8 first while still running Ganeti 1.2)
1627 - DRBD devices are using statically allocated minor numbers, which
1628 will be assigned to existing instances during the migration process
1629 - there is support for both Xen PVM and Xen HVM instances running on
1631 - KVM virtualization is supported too
1632 - file-based storage has been implemented, which means that it is
1633 possible to run the cluster without LVM and DRBD storage, for
1634 example using a shared filesystem exported from shared storage (and
1635 still have live migration)
1641 *(Released Tue, 13 Jan 2009)*
1643 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
1644 - Reuse the old instance mac address by default on instance import, if
1645 the instance name is the same.
1646 - Handle situations in which the node info rpc returns incomplete
1648 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
1649 - Improved version of batcher:
1651 - state file support
1652 - instance mac address support
1653 - support for HVM clusters/instances
1655 - Add an option to show the number of cpu sockets and nodes in
1657 - Support OSes that handle more than one version of the OS api (but do
1658 not change the current API in any other way)
1659 - Fix ``gnt-node migrate``
1660 - ``gnt-debug`` man page
1661 - Fixes various more typos and small issues
1662 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
1668 *(Released Wed, 24 Sep 2008)*
1670 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
1671 type of disk exported to fully virtualized instances.
1672 - provide access to the serial console of HVM instances
1673 - instance auto_balance flag, set by default. If turned off it will
1674 avoid warnings on cluster verify if there is not enough memory to fail
1675 over an instance. in the future it will prevent automatically failing
1676 it over when we will support that.
1677 - batcher tool for instance creation, see ``tools/README.batcher``
1678 - ``gnt-instance reinstall --select-os`` to interactively select a new
1679 operating system when reinstalling an instance.
1680 - when changing the memory amount on instance modify a check has been
1681 added that the instance will be able to start. also warnings are
1682 emitted if the instance will not be able to fail over, if auto_balance
1684 - documentation fixes
1685 - sync fields between ``gnt-instance list/modify/add/import``
1686 - fix a race condition in drbd when the sync speed was set after giving
1687 the device a remote peer.
1693 *(Released Tue, 22 Jul 2008)*
1695 - note: the allowed size and number of tags per object were reduced
1696 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
1697 - fixed twisted 8.x compatibility
1698 - fixed ``gnt-instance replace-disks`` with iallocator
1699 - add TCP keepalives on twisted connections to detect restarted nodes
1700 - disk increase support, see ``gnt-instance grow-disk``
1701 - implement bulk node/instance query for RAPI
1702 - add tags in node/instance listing (optional)
1703 - experimental migration (and live migration) support, read the man page
1704 for ``gnt-instance migrate``
1705 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
1706 has some small improvements in handling its state file
1712 *(Released Fri, 13 Jun 2008)*
1714 - Experimental readonly, REST-based remote API implementation;
1715 automatically started on master node, TCP port 5080, if enabled by
1716 ``--enable-rapi`` parameter to configure script.
1717 - Instance allocator support. Add and import instance accept a
1718 ``--iallocator`` parameter, and call that instance allocator to decide
1719 which node to use for the instance. The iallocator document describes
1720 what's expected from an allocator script.
1721 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
1722 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
1723 node is lost there is still enough memory to fail over the instances
1725 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
1726 ``gnt-cluster verify``, to check for site-specific compliance. All the
1727 hooks will run, and their output, if any, will be displayed. Any
1728 failing hook will make the verification return an error value.
1729 - ``gnt-cluster verify`` now checks that its peers are reachable on the
1730 primary and secondary interfaces
1731 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
1732 that is still declared as part of the cluster and has failed.
1733 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
1734 specifying output fields, that just adds the chosen fields to the
1736 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
1737 export from the filesystem.
1738 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
1739 have been added. Using them you can enable/disable acpi and pae
1740 support, and specify a path for a cd image to be exported to the
1741 instance. These parameters as the name suggest only work on HVM
1743 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
1744 PAE support will be set to the previously hardcoded values, but the
1745 (previously hardcoded) path to the CDROM ISO image will be unset and
1746 if required, needs to be set manually with ``gnt-instance modify``
1748 - The address to which an instance's VNC console is bound is now
1749 selectable per-instance, rather than being cluster wide. Of course
1750 this only applies to instances controlled via VNC, so currently just
1751 applies to HVM clusters.
1757 *(Released Mon, 18 Feb 2008)*
1759 - more tweaks to the disk activation code (especially helpful for DRBD)
1760 - change the default ``gnt-instance list`` output format, now there is
1761 one combined status field (see the manpage for the exact values this
1763 - some more fixes for the mac export to hooks change
1764 - make Ganeti not break with DRBD 8.2.x (which changed the version
1765 format in ``/proc/drbd``) (issue 24)
1766 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
1767 template, allowing migration from DRBD0.7+MD to DRBD8
1773 *(Released Wed, 30 Jan 2008)*
1775 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
1777 - add command aliases infrastructure and a few aliases
1778 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
1779 man pages and the ``--help`` option of ``gnt-node
1780 list``/``gnt-instance list``
1781 - fix ``gnt-backup list`` with down nodes (issue 21)
1782 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
1783 - fix the dist archive and add a check for including svn/git files in
1785 - some developer-related changes: improve the burnin and the QA suite,
1786 add an upload script for testing during development
1792 *(Released Wed, 16 Jan 2008)*
1794 - experimental HVM support, read the install document, section
1795 "Initializing the cluster"
1796 - allow for the PVM hypervisor per-instance kernel and initrd paths
1797 - add a new command ``gnt-cluster verify-disks`` which uses a new
1798 algorithm to improve the reconnection of the DRBD pairs if the device
1799 on the secondary node has gone away
1800 - make logical volume code auto-activate LVs at disk activation time
1801 - slightly improve the speed of activating disks
1802 - allow specification of the MAC address at instance creation time, and
1803 changing it later via ``gnt-instance modify``
1804 - fix handling of external commands that generate lots of output on
1806 - update documentation with regard to minimum version of DRBD8 supported
1812 *(Released Tue, 4 Dec 2007)*
1814 - Log the ``xm create`` output to the node daemon log on failure (to
1815 help diagnosing the error)
1816 - In debug mode, log all external commands output if failed to the logs
1817 - Change parsing of lvm commands to ignore stderr
1823 *(Released Wed, 28 Nov 2007)*
1825 - Another round of updates to the DRBD 8 code to deal with more failures
1826 in the replace secondary node operation
1827 - Some more logging of failures in disk operations (lvm, drbd)
1828 - A few documentation updates
1835 *(Released Tue, 13 Nov 2007)*
1837 - Change configuration file format from Python's Pickle to JSON.
1838 Upgrading is possible using the cfgupgrade utility.
1839 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
1840 faster replace disks and is more stable (DRBD 8 has many improvements
1841 compared to DRBD 0.7)
1842 - Added command line tags support (see man pages for ``gnt-instance``,
1843 ``gnt-node``, ``gnt-cluster``)
1844 - Added instance rename support
1845 - Added multi-instance startup/shutdown
1846 - Added cluster rename support
1847 - Added ``gnt-node evacuate`` to simplify some node operations
1848 - Added instance reboot operation that can speedup reboot as compared to
1850 - Soften the requirement that hostnames are in FQDN format
1851 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
1853 - Removed dependency on debian's patched fping that uses the
1854 non-standard ``-S`` option
1855 - Now the OS definitions are searched for in multiple, configurable
1856 paths (easier for distros to package)
1857 - Some changes to the hooks infrastructure (especially the new
1858 post-configuration update hook)
1859 - Other small bugfixes
1861 .. vim: set textwidth=72 syntax=rst :