8 *(Released Thu, 12 Apr 2012)*
10 Incompatible/important changes and bugfixes
11 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
13 - The default of the ``/2/instances/[instance_name]/rename`` RAPI
14 resource's ``ip_check`` parameter changed from ``True`` to ``False``
15 to match the underlying LUXI interface.
16 - The ``/2/nodes/[node_name]/evacuate`` RAPI resource was changed to use
17 body parameters, see :doc:`RAPI documentation <rapi>`. The server does
18 not maintain backwards-compatibility as the underlying operation
19 changed in an incompatible way. The RAPI client can talk to old
20 servers, but it needs to be told so as the return value changed.
21 - When creating file-based instances via RAPI, the ``file_driver``
22 parameter no longer defaults to ``loop`` and must be specified.
23 - The deprecated ``bridge`` NIC parameter is no longer supported. Use
25 - Support for the undocumented and deprecated RAPI instance creation
26 request format version 0 has been dropped. Use version 1, supported
27 since Ganeti 2.1.3 and :doc:`documented <rapi>`, instead.
28 - Pyparsing 1.4.6 or above is required, see :doc:`installation
29 documentation <install>`.
30 - The "cluster-verify" hooks are now executed per group by the
31 ``OP_CLUSTER_VERIFY_GROUP`` opcode. This maintains the same behavior
32 if you just run ``gnt-cluster verify``, which generates one opcode per
34 - The environment as passed to the OS scripts is cleared, and thus no
35 environment variables defined in the node daemon's environment will be
36 inherited by the scripts.
37 - The :doc:`iallocator <iallocator>` mode ``multi-evacuate`` has been
39 - :doc:`New iallocator modes <design-multi-reloc>` have been added to
40 support operations involving multiple node groups.
41 - Offline nodes are ignored when failing over an instance.
42 - Support for KVM version 1.0, which changed the version reporting format
44 - TCP/IP ports used by DRBD disks are returned to a pool upon instance
46 - ``Makefile`` is now compatible with Automake 1.11.2
47 - Includes all bugfixes made in the 2.4 series
52 - The ganeti-htools project has been merged into the ganeti-core source
53 tree and will be built as part of Ganeti (see :doc:`install-quick`).
54 - Implemented support for :doc:`shared storage <design-shared-storage>`.
55 - Add support for disks larger than 2 TB in ``lvmstrap`` by supporting
56 GPT-style partition tables (requires `parted
57 <http://www.gnu.org/s/parted/>`_).
58 - Added support for floppy drive and 2nd CD-ROM drive in KVM hypervisor.
59 - Allowed adding tags on instance creation.
60 - Export instance tags to hooks (``INSTANCE_TAGS``, see :doc:`hooks`)
61 - Allow instances to be started in a paused state, enabling the user to
62 see the complete console output on boot using the console.
63 - Added new hypervisor flag to control default reboot behaviour
64 (``reboot_behavior``).
65 - Added support for KVM keymaps (hypervisor parameter ``keymap``).
66 - Improved out-of-band management support:
68 - Added ``gnt-node health`` command reporting the health status of
70 - Added ``gnt-node power`` command to manage power status of nodes.
71 - Added command for emergency power-off (EPO), ``gnt-cluster epo``.
73 - Instance migration can fall back to failover if instance is not
75 - Filters can be used when listing nodes, instances, groups and locks;
76 see *ganeti(7)* manpage.
77 - Added post-execution status as variables to :doc:`hooks <hooks>`
79 - Instance tags are exported/imported together with the instance.
80 - When given an explicit job ID, ``gnt-job info`` will work for archived
82 - Jobs can define dependencies on other jobs (not yet supported via
83 RAPI or command line, but used by internal commands and usable via
86 - Lock monitor (``gnt-debug locks``) shows jobs waiting for
89 - Instance failover is now available as a RAPI resource
90 (``/2/instances/[instance_name]/failover``).
91 - ``gnt-instance info`` defaults to static information if primary node
93 - Opcodes have a new ``comment`` attribute.
94 - Added basic SPICE support to KVM hypervisor.
95 - ``tools/ganeti-listrunner`` allows passing of arguments to executable.
97 Node group improvements
98 ~~~~~~~~~~~~~~~~~~~~~~~
100 - ``gnt-cluster verify`` has been modified to check groups separately,
101 thereby improving performance.
102 - Node group support has been added to ``gnt-cluster verify-disks``,
103 which now operates per node group.
104 - Watcher has been changed to work better with node groups.
106 - One process and state file per node group.
107 - Slow watcher in one group doesn't block other group's watcher.
109 - Added new command, ``gnt-group evacuate``, to move all instances in a
110 node group to other groups.
111 - Added ``gnt-instance change-group`` to move an instance to another
113 - ``gnt-cluster command`` and ``gnt-cluster copyfile`` now support
114 per-group operations.
115 - Node groups can be tagged.
116 - Some operations switch from an exclusive to a shared lock as soon as
118 - Instance's primary and secondary nodes' groups are now available as
119 query fields (``pnode.group``, ``pnode.group.uuid``, ``snodes.group``
120 and ``snodes.group.uuid``).
125 - Numerous updates to documentation and manpages.
127 - :doc:`RAPI <rapi>` documentation now has detailed parameter
129 - Some opcode/job results are now also documented, see :doc:`RAPI
132 - A lockset's internal lock is now also visible in lock monitor.
133 - Log messages from job queue workers now contain information about the
134 opcode they're processing.
135 - ``gnt-instance console`` no longer requires the instance lock.
136 - A short delay when waiting for job changes reduces the number of LUXI
137 requests significantly.
138 - DRBD metadata volumes are overwritten with zeros during disk creation.
139 - Out-of-band commands no longer acquire the cluster lock in exclusive
141 - ``devel/upload`` now uses correct permissions for directories.
147 *(Released Fri, 23 Mar 2012)*
149 This was the sixth release candidate of the 2.5 series.
155 *(Released Mon, 9 Jan 2012)*
157 This was the fifth release candidate of the 2.5 series.
163 *(Released Thu, 27 Oct 2011)*
165 This was the fourth release candidate of the 2.5 series.
171 *(Released Wed, 26 Oct 2011)*
173 This was the third release candidate of the 2.5 series.
179 *(Released Tue, 18 Oct 2011)*
181 This was the second release candidate of the 2.5 series.
187 *(Released Tue, 4 Oct 2011)*
189 This was the first release candidate of the 2.5 series.
195 *(Released Wed, 31 Aug 2011)*
197 This was the third beta release of the 2.5 series.
203 *(Released Mon, 22 Aug 2011)*
205 This was the second beta release of the 2.5 series.
211 *(Released Fri, 12 Aug 2011)*
213 This was the first beta release of the 2.5 series.
219 *(Released Thu, 27 Oct 2011)*
221 - Fixed bug when parsing command line parameter values ending in
223 - Fixed assertion error after unclean master shutdown
224 - Disable HTTP client pool for RPC, significantly reducing memory usage
226 - Fixed queue archive creation with wrong permissions
232 *(Released Tue, 23 Aug 2011)*
236 - Fixed documentation for importing with ``--src-dir`` option
237 - Fixed a bug in ``ensure-dirs`` with queue/archive permissions
238 - Fixed a parsing issue with DRBD 8.3.11 in the Linux kernel
244 *(Released Fri, 5 Aug 2011)*
246 Many bug-fixes and a few small features:
248 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
249 issues when you tried to add an instance with two MAC addresses in one
251 - KVM: fixed per-instance stored UID value
252 - KVM: configure bridged NICs at migration start
253 - KVM: Fix a bug where instance will not start with never KVM versions
255 - Added OS search path to ``gnt-cluster info``
256 - Fixed an issue with ``file_storage_dir`` where you were forced to
257 provide an absolute path, but the documentation states it is a
258 relative path, the documentation was right
259 - Added a new parameter to instance stop/start called ``--no-remember``
260 that will make the state change to not be remembered
261 - Implemented ``no_remember`` at RAPI level
262 - Improved the documentation
263 - Node evacuation: don't call IAllocator if node is already empty
264 - Fixed bug in DRBD8 replace disks on current nodes
265 - Fixed bug in recreate-disks for DRBD instances
266 - Moved assertion checking locks in ``gnt-instance replace-disks``
267 causing it to abort with not owning the right locks for some situation
268 - Job queue: Fixed potential race condition when cancelling queued jobs
269 - Fixed off-by-one bug in job serial generation
270 - ``gnt-node volumes``: Fix instance names
271 - Fixed aliases in bash completion
272 - Fixed a bug in reopening log files after being sent a SIGHUP
273 - Added a flag to burnin to allow specifying VCPU count
274 - Bugfixes to non-root Ganeti configuration
280 *(Released Thu, 12 May 2011)*
282 Many bug-fixes and a few new small features:
284 - Fixed a bug related to log opening failures
285 - Fixed a bug in instance listing with orphan instances
286 - Fixed a bug which prevented resetting the cluster-level node parameter
287 ``oob_program`` to the default
288 - Many fixes related to the ``cluster-merge`` tool
289 - Fixed a race condition in the lock monitor, which caused failures
290 during (at least) creation of many instances in parallel
291 - Improved output for gnt-job info
292 - Removed the quiet flag on some ssh calls which prevented debugging
294 - Improved the N+1 failure messages in cluster verify by actually
295 showing the memory values (needed and available)
296 - Increased lock attempt timeouts so that when executing long operations
297 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
298 too early and thus prevent the use of the 'fair' mechanism
299 - Changed instance query data (``gnt-instance info``) to not acquire
300 locks unless needed, thus allowing its use on locked instance if only
301 static information is asked for
302 - Improved behaviour with filesystems that do not support rename on an
304 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
305 kept locks on all nodes during the wipe, which is unneeded
306 - Fixed ``gnt-watcher`` handling of errors during hooks execution
307 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
308 10GiB) which caused the wipe to fail right at the end in some cases
309 - Fixed master IP activation when doing master failover with no-voting
310 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
311 the master node itself
312 - Fixed potential data-loss in under disk full conditions, where Ganeti
313 wouldn't check correctly the return code and would consider
314 partially-written files 'correct'
315 - Fixed bug related to multiple VGs and DRBD disk replacing
316 - Added new disk parameter ``metavg`` that allows placement of the meta
317 device for DRBD in a different volume group
318 - Fixed error handling in the node daemon when the system libc doesn't
319 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
320 - Fixed lock release during replace-disks, which kept cluster-wide locks
321 when doing disk replaces with an iallocator script
322 - Added check for missing bridges in cluster verify
323 - Handle EPIPE errors while writing to the terminal better, so that
324 piping the output to e.g. ``less`` doesn't cause a backtrace
325 - Fixed rare case where a ^C during Luxi calls could have been
326 interpreted as server errors, instead of simply terminating
327 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
329 - Added a few more parameters to the KVM hypervisor, allowing a second
330 CDROM, custom disk type for CDROMs and a floppy image
331 - Removed redundant message in instance rename when the name is given
333 - Added option to ``gnt-instance recreate-disks`` to allow creating the
334 disks on new nodes, allowing recreation when the original instance
335 nodes are completely gone
336 - Added option when converting disk templates to DRBD to skip waiting
337 for the resync, in order to make the instance available sooner
338 - Added two new variables to the OS scripts environment (containing the
340 - Made the root_path and optional parameter for the xen-pvm hypervisor,
341 to allow use of ``pvgrub`` as bootloader
342 - Changed the instance memory modifications to only check out-of-memory
343 conditions on memory increases, and turned the secondary node warnings
344 into errors (they can still be overridden via ``--force``)
345 - Fixed the handling of a corner case when the Python installation gets
346 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
347 to execute a command that doesn't exist
348 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
349 node of the instance returned failures during instance shutdown; this
350 adds the option ``--ignore-consistency`` to gnt-instance move
352 And as usual, various improvements to the error messages, documentation
358 *(Released Wed, 09 Mar 2011)*
360 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
361 the RAPI users file if run twice (even with ``--dry-run``).
363 The release fixes that bug (nothing else changed).
369 *(Released Mon, 07 Mar 2011)*
371 Final 2.4.0 release. Just a few small fixes:
373 - Fixed RAPI node evacuate
374 - Fixed the kvm-ifup script
375 - Fixed internal error handling for special job cases
376 - Updated man page to specify the escaping feature for options
382 *(Released Mon, 28 Feb 2011)*
384 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
385 that this feature wiped the disks of the wrong instance, leading to loss
390 - Fixed title of query field containing instance name
391 - Expanded the glossary in the documentation
392 - Fixed one unittest (internal issue)
398 *(Released Mon, 21 Feb 2011)*
400 A number of bug fixes plus just a couple functionality changes.
402 On the user-visible side, the ``gnt-* list`` command output has changed
403 with respect to "special" field states. The current rc1 style of display
404 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
405 the default output mode special fields states are displayed as follows:
407 - Offline resource: ``*``
408 - Unavailable/not applicable: ``-``
409 - Data missing (RPC failure): ``?``
410 - Unknown field: ``??``
412 Another user-visible change is the addition of ``--force-join`` to
417 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
418 - Fixed regression in RAPI/instance reinstall where all parameters were
419 required (instead of optional)
420 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
421 - Fixed iallocator usage (offline nodes were not considered offline)
422 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
423 - Fixed hypervisor and OS parameter validation with respect to
425 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
427 - Fixed ``tools/listrunner`` with respect to agent-based usage
433 *(Released Fri, 4 Feb 2011)*
435 Many changes and fixes since the beta1 release. While there were some
436 internal changes, the code has been mostly stabilised for the RC
439 Note: the dumb allocator was removed in this release, as it was not kept
440 up-to-date with the IAllocator protocol changes. It is recommended to
441 use the ``hail`` command from the ganeti-htools package.
443 Note: the 2.4 and up versions of Ganeti are not compatible with the
444 0.2.x branch of ganeti-htools. You need to upgrade to
445 ganeti-htools-0.3.0 (or later).
447 Regressions fixed from 2.3
448 ~~~~~~~~~~~~~~~~~~~~~~~~~~
450 - Fixed the ``gnt-cluster verify-disks`` command
451 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
453 - Fixed disk adoption breakage
454 - Fixed wrong headers in instance listing for field aliases
459 - Fixed corner case in KVM handling of NICs
460 - Fixed many cases of wrong handling of non-vm_capable nodes
461 - Fixed a bug where a missing instance symlink was not possible to
462 recreate with any ``gnt-*`` command (now ``gnt-instance
463 activate-disks`` does it)
464 - Fixed the volume group name as reported by ``gnt-cluster
466 - Increased timeouts for the import-export code, hopefully leading to
467 fewer aborts due network or instance timeouts
468 - Fixed bug in ``gnt-node list-storage``
469 - Fixed bug where not all daemons were started on cluster
470 initialisation, but only at the first watcher run
471 - Fixed many bugs in the OOB implementation
472 - Fixed watcher behaviour in presence of instances with offline
474 - Fixed instance list output for instances running on the wrong node
475 - a few fixes to the cluster-merge tool, but it still cannot merge
476 multi-node groups (currently it is not recommended to use this tool)
482 - Improved network configuration for the KVM hypervisor
483 - Added e1000 as a supported NIC for Xen-HVM
484 - Improved the lvmstrap tool to also be able to use partitions, as
485 opposed to full disks
486 - Improved speed of disk wiping (the cluster parameter
487 ``prealloc_wipe_disks``, so that it has a low impact on the total time
488 of instance creations
489 - Added documentation for the OS parameters
490 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
491 hypervisor is not responding
492 - Added display of blacklisted and hidden OS information in
494 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
495 NIC and node parameters, which might create problems with currently
496 invalid (but undetected) configuration files, but prevents validation
497 failures when unrelated parameters are modified
498 - Changed cluster initialisation to wait for the master daemon to become
500 - Expanded the RAPI interface:
502 - Added config redistribution resource
503 - Added activation/deactivation of instance disks
504 - Added export of console information
506 - Implemented log file reopening on SIGHUP, which allows using
507 logrotate(8) for the Ganeti log files
508 - Added a basic OOB helper script as an example
514 *(Released Fri, 14 Jan 2011)*
519 - Fixed timezone issues when formatting timestamps
520 - Added support for node groups, available via ``gnt-group`` and other
522 - Added out-of-band framework and management, see :doc:`design
523 document <design-oob>`
524 - Removed support for roman numbers from ``gnt-node list`` and
525 ``gnt-instance list``.
526 - Allowed modification of master network interface via ``gnt-cluster
527 modify --master-netdev``
528 - Accept offline secondaries while shutting down instance disks
529 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
530 - Added support for multiple LVM volume groups
531 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
533 - Added commands to list available fields:
535 - ``gnt-node list-fields``
536 - ``gnt-group list-fields``
537 - ``gnt-instance list-fields``
539 - Updated documentation and man pages
544 - Moved ``rapi_users`` file into separate directory, now named
545 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
547 - Added new tool for running commands on many machines,
548 ``tools/ganeti-listrunner``
549 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
550 improving the ``gnt-instance console`` output
551 - Allowed customisation of disk index separator at ``configure`` time
552 - Export node group allocation policy to :doc:`iallocator <iallocator>`
553 - Added support for non-partitioned md disks in ``lvmstrap``
554 - Added script to gracefully power off KVM instances
555 - Split ``utils`` module into smaller parts
556 - Changed query operations to return more detailed information, e.g.
557 whether an information is unavailable due to an offline node. To use
558 this new functionality, the LUXI call ``Query`` must be used. Field
559 information is now stored by the master daemon and can be retrieved
560 using ``QueryFields``. Instances, nodes and groups can also be queried
561 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
562 exposed via RAPI). The following commands make use of this
563 infrastructure change:
566 - ``gnt-group list-fields``
568 - ``gnt-node list-fields``
569 - ``gnt-instance list``
570 - ``gnt-instance list-fields``
571 - ``gnt-debug locks``
576 - New RAPI resources (see :doc:`rapi`):
580 - ``/2/groups/[group_name]``
581 - ``/2/groups/[group_name]/assign-nodes``
582 - ``/2/groups/[group_name]/modify``
583 - ``/2/groups/[group_name]/rename``
584 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
588 - Implemented ``no_install`` for instance creation
589 - Implemented OS parameters for instance reinstallation, allowing
590 use of special settings on reinstallation (e.g. for preserving data)
595 - Added IPv6 support in import/export
596 - Pause DRBD synchronization while wiping disks on instance creation
597 - Updated unittests and QA scripts
598 - Improved network parameters passed to KVM
599 - Converted man pages from docbook to reStructuredText
605 *(Released Mon, 20 Dec 2010)*
607 Released version 2.3.1~rc1 without any changes.
613 *(Released Wed, 1 Dec 2010)*
615 - impexpd: Disable OpenSSL compression in socat if possible (backport
616 from master, commit e90739d625b, see :doc:`installation guide
617 <install-quick>` for details)
618 - Changed unittest coverage report to exclude test scripts
619 - Added script to check version format
625 *(Released Wed, 1 Dec 2010)*
627 Released version 2.3.0~rc1 without any changes.
633 *(Released Fri, 19 Nov 2010)*
635 A number of bugfixes and documentation updates:
637 - Update ganeti-os-interface documentation
638 - Fixed a bug related to duplicate MACs or similar items which should be
640 - Fix breakage in OS state modify
641 - Reinstall instance: disallow offline secondaries (fixes bug related to
642 OS changing but reinstall failing)
643 - plus all the other fixes between 2.2.1 and 2.2.2
649 *(Released Tue, 2 Nov 2010)*
651 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
652 - Fixed master failover race with watcher
653 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
655 - Accept previously stopped instance for export with instance removal
656 - Simplify and extend the environment variables for instance OS scripts
657 - Added new node flags, ``master_capable`` and ``vm_capable``
658 - Added optional instance disk wiping prior during allocation. This is a
659 cluster-wide option and can be set/modified using
660 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
661 - Added IPv6 support, see :doc:`design document <design-2.3>` and
663 - Added a new watcher option (``--ignore-pause``)
664 - Added option to ignore offline node on instance start/stop
665 (``--ignore-offline``)
666 - Allow overriding OS parameters with ``gnt-instance reinstall``
667 - Added ability to change node's secondary IP address using ``gnt-node
669 - Implemented privilege separation for all daemons except
670 ``ganeti-noded``, see ``configure`` options
671 - Complain if an instance's disk is marked faulty in ``gnt-cluster
673 - Implemented job priorities (see ``ganeti(7)`` manpage)
674 - Ignore failures while shutting down instances during failover from
676 - Exit daemon's bootstrap process only once daemon is ready
677 - Export more information via ``LUInstanceQuery``/remote API
678 - Improved documentation, QA and unittests
679 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
680 restart if the file was created or changed
681 - Added LUXI protocol version sent with each request and response,
682 allowing detection of server/client mismatches
683 - Moved the Python scripts among gnt-* and ganeti-* into modules
684 - Moved all code related to setting up SSH to an external script,
686 - Infrastructure changes for node group support in future versions
692 *(Released Fri, 19 Nov 2010)*
694 A few small bugs fixed, and some improvements to the build system:
696 - Fix documentation regarding conversion to drbd
697 - Fix validation of parameters in cluster modify (``gnt-cluster modify
699 - Fix error handling in node modify with multiple changes
700 - Allow remote imports without checked names
706 *(Released Tue, 19 Oct 2010)*
708 - Disable SSL session ID cache in RPC client
714 *(Released Thu, 14 Oct 2010)*
716 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
717 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
719 - Fix problems with interaction between Curl and Python's HTTP server,
720 resulting in increased speed in many RPC calls
721 - Improve our release script to prevent breakage with older aclocal and
728 *(Released Thu, 7 Oct 2010)*
730 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
731 value retrieved from master
732 - Added support for blacklisted or hidden OS definitions
733 - Added simple lock monitor (accessible via (``gnt-debug locks``)
734 - Added support for -mem-path in KVM hypervisor abstraction layer
735 - Allow overriding instance parameters in tool for inter-cluster
736 instance moves (``tools/move-instance``)
737 - Improved opcode summaries (e.g. in ``gnt-job list``)
738 - Improve consistency of OS listing by sorting it
739 - Documentation updates
745 *(Released Fri, 8 Oct 2010)*
747 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
753 *(Released Mon, 4 Oct 2010)*
755 - Fixed regression in ``gnt-instance rename``
761 *(Released Wed, 22 Sep 2010)*
763 - Fixed OS_VARIANT variable for OS scripts
764 - Fixed cluster tag operations via RAPI
765 - Made ``setup-ssh`` exit with non-zero code if an error occurred
766 - Disabled RAPI CA checks in watcher
772 *(Released Mon, 23 Aug 2010)*
774 - Support DRBD versions of the format "a.b.c.d"
776 - Re-introduce support for usage from multiple threads in RAPI client
777 - Instance renames and modify via RAPI
778 - Work around race condition between processing and archival in job
780 - Mark opcodes following failed one as failed, too
781 - Job field ``lock_status`` was removed due to difficulties making it
782 work with the changed job queue in Ganeti 2.2; a better way to monitor
783 locks is expected for a later 2.2.x release
784 - Fixed dry-run behaviour with many commands
785 - Support ``ssh-agent`` again when adding nodes
786 - Many additional bugfixes
792 *(Released Fri, 30 Jul 2010)*
794 Important change: the internal RPC mechanism between Ganeti nodes has
795 changed from using a home-grown http library (based on the Python base
796 libraries) to use the PycURL library. This requires that PycURL is
797 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
798 against GnuTLS by default. cURL's support for GnuTLS had known issues
799 before cURL 7.21.0 and we recommend using the latest cURL release or
800 linking against OpenSSL. Most other distributions already link PycURL
801 and cURL against OpenSSL. The command::
803 python -c 'import pycurl; print pycurl.version'
805 can be used to determine the libraries PycURL and cURL are linked
808 Other significant changes:
810 - Rewrote much of the internals of the job queue, in order to achieve
811 better parallelism; this decouples job query operations from the job
812 processing, and it should allow much nicer behaviour of the master
813 daemon under load, and it also has uncovered some long-standing bugs
814 related to the job serialisation (now fixed)
815 - Added a default iallocator setting to the cluster parameters,
816 eliminating the need to always pass nodes or an iallocator for
817 operations that require selection of new node(s)
818 - Added experimental support for the LXC virtualization method
819 - Added support for OS parameters, which allows the installation of
820 instances to pass parameter to OS scripts in order to customise the
822 - Added a hypervisor parameter controlling the migration type (live or
823 non-live), since hypervisors have various levels of reliability; this
824 has renamed the 'live' parameter to 'mode'
825 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
826 logical volumes, meaning that cluster verify will ignore them and not
827 flag their presence as errors
828 - The watcher will now reset the error count for failed instances after
829 8 hours, thus allowing self-healing if the problem that caused the
830 instances to be down/fail to start has cleared in the meantime
831 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
832 check for, and warn, if the drbd module parameter ``usermode_helper``
833 is not consistent with the cluster-wide setting; this is needed to
834 make diagnose easier of failed drbd creations
835 - Started adding base IPv6 support, but this is not yet
836 enabled/available for use
837 - Rename operations (cluster, instance) will now return the new name,
838 which is especially useful if a short name was passed in
839 - Added support for instance migration in RAPI
840 - Added a tool to pre-configure nodes for the SSH setup, before joining
841 them to the cluster; this will allow in the future a simplified model
842 for node joining (but not yet fully enabled in 2.2); this needs the
843 paramiko python library
844 - Fixed handling of name-resolving errors
845 - Fixed consistency of job results on the error path
846 - Fixed master-failover race condition when executed multiple times in
848 - Fixed many bugs related to the job queue (mostly introduced during the
849 2.2 development cycle, so not all are impacting 2.1)
850 - Fixed instance migration with missing disk symlinks
851 - Fixed handling of unknown jobs in ``gnt-job archive``
852 - And many other small fixes/improvements
856 - Enhanced both the unittest and the QA coverage
857 - Switched the opcode validation to a generic model, and extended the
858 validation to all opcode parameters
859 - Changed more parts of the code that write shell scripts to use the
861 - Switched the master daemon to use the asyncore library for the Luxi
868 *(Released Thu, 17 Jun 2010)*
870 - Added tool (``move-instance``) and infrastructure to move instances
871 between separate clusters (see :doc:`separate documentation
872 <move-instance>` and :doc:`design document <design-2.2>`)
873 - Added per-request RPC timeout
874 - RAPI now requires a Content-Type header for requests with a body (e.g.
875 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
876 :rfc:`2616` (HTTP/1.1), section 7.2.1)
877 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
879 - Implemented initial support for running Ganeti daemons as separate
880 users, see configure-time flags ``--with-user-prefix`` and
881 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
883 - Instances can be removed after export (``gnt-backup export
885 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
886 key (instead of 1024 bit)
887 - Added new cluster configuration file for cluster domain secret
888 - Import/export now use SSL instead of SSH
889 - Added support for showing estimated time when exporting an instance,
890 see the ``ganeti-os-interface(7)`` manpage and look for
897 *(Released Tue, 16 Nov 2010)*
899 Some more bugfixes. Unless critical bugs occur, this will be the last
902 - Fix case of MAC special-values
903 - Fix mac checker regex
904 - backend: Fix typo causing "out of range" error
905 - Add missing --units in gnt-instance list man page
911 *(Released Tue, 24 Aug 2010)*
914 - Don't ignore secondary node silently on non-mirrored disk templates
916 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
917 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
918 - Properly document the usb_mouse parameter
919 - Fix path in ganeti-rapi(8) (issue 116)
920 - Adjust error message when the ganeti user's .ssh directory is
922 - Add same-node-check when changing the disk template to drbd
928 *(Released Fri, 16 Jul 2010)*
931 - Add an option to only select some reboot types during qa/burnin.
932 (on some hypervisors consequent reboots are not supported)
933 - Fix infrequent race condition in master failover. Sometimes the old
934 master ip address would be still detected as up for a short time
935 after it was removed, causing failover to fail.
936 - Decrease mlockall warnings when the ctypes module is missing. On
937 Python 2.4 we support running even if no ctypes module is installed,
938 but we were too verbose about this issue.
939 - Fix building on old distributions, on which man doesn't have a
941 - Fix RAPI not to ignore the MAC address on instance creation
942 - Implement the old instance creation format in the RAPI client.
948 *(Released Thu, 01 Jul 2010)*
950 A small bugfix release:
951 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
952 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
953 a non-existing option
954 - Fix instance create: the --force-variant option was ignored
955 - Improve pylint 0.21 compatibility and warnings with Python 2.6
956 - Fix modify node storage with non-FQDN arguments
957 - Fix RAPI client to authenticate under Python 2.6 when used
958 for more than 5 requests needing authentication
959 - Fix gnt-instance modify -t (storage) giving a wrong error message
960 when converting a non-shutdown drbd instance to plain
966 *(Released Fri, 18 Jun 2010)*
968 A small bugfix release:
970 - Fix live migration of KVM instances started with older Ganeti
971 versions which had fewer hypervisor parameters
972 - Fix gnt-instance grow-disk on down instances
973 - Fix an error-reporting bug during instance migration
974 - Better checking of the ``--net`` and ``--disk`` values, to avoid
975 silently ignoring broken ones
976 - Fix an RPC error reporting bug affecting, for example, RAPI client
978 - Fix bug triggered by different API version os-es on different nodes
979 - Fix a bug in instance startup with custom hvparams: OS level
980 parameters would fail to be applied.
981 - Fix the RAPI client under Python 2.6 (but more work is needed to
982 make it work completely well with OpenSSL)
983 - Fix handling of errors when resolving names from DNS
989 *(Released Thu, 3 Jun 2010)*
991 A medium sized development cycle. Some new features, and some
992 fixes/small improvements/cleanups.
997 The node deamon now tries to mlock itself into memory, unless the
998 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
999 its logs, and falls back to console logging. This allows emergency
1000 features such as ``gnt-node powercycle`` to work even in the event of a
1001 broken node disk (tested offlining the disk hosting the node's
1002 filesystem and dropping its memory caches; don't try this at home)
1004 KVM: add vhost-net acceleration support. It can be tested with a new
1005 enough version of the kernel and of qemu-kvm.
1007 KVM: Add instance chrooting feature. If you use privilege dropping for
1008 your VMs you can also now force them to chroot to an empty directory,
1009 before starting the emulated guest.
1011 KVM: Add maximum migration bandwith and maximum downtime tweaking
1012 support (requires a new-enough version of qemu-kvm).
1014 Cluster verify will now warn if the master node doesn't have the master
1015 ip configured on it.
1017 Add a new (incompatible) instance creation request format to RAPI which
1018 supports all parameters (previously only a subset was supported, and it
1019 wasn't possible to extend the old format to accomodate all the new
1020 features. The old format is still supported, and a client can check for
1021 this feature, before using it, by checking for its presence in the
1022 ``features`` RAPI resource.
1024 Now with ancient latin support. Try it passing the ``--roman`` option to
1025 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
1026 (requires the python-roman module to be installed, in order to work).
1031 As usual many internal code refactorings, documentation updates, and
1034 - Lots of improvements and cleanups to the experimental Remote API
1035 (RAPI) client library.
1036 - A new unit test suite for the core daemon libraries.
1037 - A fix to creating missing directories makes sure the umask is not
1038 applied anymore. This enforces the same directory permissions
1040 - Better handling terminating daemons with ctrl+c (used when running
1041 them in debugging mode).
1042 - Fix a race condition in live migrating a KVM instance, when stat()
1043 on the old proc status file returned EINVAL, which is an unexpected
1045 - Fixed manpage checking with newer man and utf-8 charachters. But now
1046 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
1052 *(Released Fri, 7 May 2010)*
1054 Fix a bug which prevented untagged KVM instances from starting.
1060 *(Released Fri, 7 May 2010)*
1062 Another release with a long development cycle, during which many
1063 different features were added.
1065 Significant features
1066 ~~~~~~~~~~~~~~~~~~~~
1068 The KVM hypervisor now can run the individual instances as non-root, to
1069 reduce the impact of a VM being hijacked due to bugs in the
1070 hypervisor. It is possible to run all instances as a single (non-root)
1071 user, to manually specify a user for each instance, or to dynamically
1072 allocate a user out of a cluster-wide pool to each instance, with the
1073 guarantee that no two instances will run under the same user ID on any
1076 An experimental RAPI client library, that can be used standalone
1077 (without the other Ganeti libraries), is provided in the source tree as
1078 ``lib/rapi/client.py``. Note this client might change its interface in
1079 the future, as we iterate on its capabilities.
1081 A new command, ``gnt-cluster renew-crypto`` has been added to easily
1082 replace the cluster's certificates and crypto keys. This might help in
1083 case they have been compromised, or have simply expired.
1085 A new disk option for instance creation has been added that allows one
1086 to "adopt" currently existing logical volumes, with data
1087 preservation. This should allow easier migration to Ganeti from
1088 unmanaged (or managed via other software) instances.
1090 Another disk improvement is the possibility to convert between redundant
1091 (DRBD) and plain (LVM) disk configuration for an instance. This should
1092 allow better scalability (starting with one node and growing the
1093 cluster, or shrinking a two-node cluster to one node).
1095 A new feature that could help with automated node failovers has been
1096 implemented: if a node sees itself as offline (by querying the master
1097 candidates), it will try to shutdown (hard) all instances and any active
1098 DRBD devices. This reduces the risk of duplicate instances if an
1099 external script automatically failovers the instances on such nodes. To
1100 enable this, the cluster parameter ``maintain_node_health`` should be
1101 enabled; in the future this option (per the name) will enable other
1102 automatic maintenance features.
1104 Instance export/import now will reuse the original instance
1105 specifications for all parameters; that means exporting an instance,
1106 deleting it and the importing it back should give an almost identical
1107 instance. Note that the default import behaviour has changed from
1108 before, where it created only one NIC; now it recreates the original
1111 Cluster verify has added a few new checks: SSL certificates validity,
1112 /etc/hosts consistency across the cluster, etc.
1117 As usual, many internal changes were done, documentation fixes,
1120 - Fixed cluster initialization with disabled cluster storage (regression
1121 introduced in 2.1.1)
1122 - File-based storage supports growing the disks
1123 - Fixed behaviour of node role changes
1124 - Fixed cluster verify for some corner cases, plus a general rewrite of
1125 cluster verify to allow future extension with more checks
1126 - Fixed log spamming by watcher and node daemon (regression introduced
1128 - Fixed possible validation issues when changing the list of enabled
1130 - Fixed cleanup of /etc/hosts during node removal
1131 - Fixed RAPI response for invalid methods
1132 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
1133 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
1134 from ide disks, etc.)
1135 - Allow OS changes without re-installation (to record a changed OS
1136 outside of Ganeti, or to allow OS renames)
1137 - Allow instance creation without OS installation (useful for example if
1138 the OS will be installed manually, or restored from a backup not in
1140 - Implemented option to make cluster ``copyfile`` use the replication
1142 - Added list of enabled hypervisors to ssconf (possibly useful for
1144 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
1146 - A partial form of node re-IP is possible via node readd, which now
1147 allows changed node primary IP
1148 - Command line utilities now show an informational message if the job is
1150 - The logs of the master daemon now show the PID/UID/GID of the
1157 *(Released Fri, 12 Mar 2010)*
1159 During the 2.1.0 long release candidate cycle, a lot of improvements and
1160 changes have accumulated with were released later as 2.1.1.
1165 The node evacuate command (``gnt-node evacuate``) was significantly
1166 rewritten, and as such the IAllocator protocol was changed - a new
1167 request type has been added. This unfortunate change during a stable
1168 series is designed to improve performance of node evacuations; on
1169 clusters with more than about five nodes and which are well-balanced,
1170 evacuation should proceed in parallel for all instances of the node
1171 being evacuated. As such, any existing IAllocator scripts need to be
1172 updated, otherwise the above command will fail due to the unknown
1173 request. The provided "dumb" allocator has not been updated; but the
1174 ganeti-htools package supports the new protocol since version 0.2.4.
1176 Another important change is increased validation of node and instance
1177 names. This might create problems in special cases, if invalid host
1178 names are being used.
1180 Also, a new layer of hypervisor parameters has been added, that sits at
1181 OS level between the cluster defaults and the instance ones. This allows
1182 customisation of virtualization parameters depending on the installed
1183 OS. For example instances with OS 'X' may have a different KVM kernel
1184 (or any other parameter) than the cluster defaults. This is intended to
1185 help managing a multiple OSes on the same cluster, without manual
1186 modification of each instance's parameters.
1188 A tool for merging clusters, ``cluster-merge``, has been added in the
1189 tools sub-directory.
1194 - Improved the int/float conversions that should make the code more
1195 robust in face of errors from the node daemons
1196 - Fixed the remove node code in case of internal configuration errors
1197 - Fixed the node daemon behaviour in face of inconsistent queue
1198 directory (e.g. read-only file-system where we can't open the files
1200 - Fixed the behaviour of gnt-node modify for master candidate demotion;
1201 now it either aborts cleanly or, if given the new "auto_promote"
1202 parameter, will automatically promote other nodes as needed
1203 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
1204 completely prevent Ganeti from working
1205 - Fixed bug for instance export when not all disks were successfully
1207 - Fixed behaviour of node add when the new node is slow in starting up
1209 - Fixed handling of signals in the LUXI client, which should improve
1210 behaviour of command-line scripts
1211 - Added checks for invalid node/instance names in the configuration (now
1212 flagged during cluster verify)
1213 - Fixed watcher behaviour for disk activation errors
1214 - Fixed two potentially endless loops in http library, which led to the
1215 RAPI daemon hanging and consuming 100% CPU in some cases
1216 - Fixed bug in RAPI daemon related to hashed passwords
1217 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
1219 - Enhanced compatibility with non-Debian OSes, but not using absolute
1220 path in some commands and allowing customisation of the ssh
1221 configuration directory
1222 - Fixed possible future issue with new Python versions by abiding to the
1223 proper use of ``__slots__`` attribute on classes
1224 - Added checks that should prevent directory traversal attacks
1225 - Many documentation fixes based on feedback from users
1230 - Added an "early_release" more for instance replace disks and node
1231 evacuate, where we release locks earlier and thus allow higher
1232 parallelism within the cluster
1233 - Added watcher hooks, intended to allow the watcher to restart other
1234 daemons (e.g. from the ganeti-nbma project), but they can be used of
1235 course for any other purpose
1236 - Added a compile-time disable for DRBD barriers, to increase
1237 performance if the administrator trusts the power supply or the
1238 storage system to not lose writes
1239 - Added the option of using syslog for logging instead of, or in
1240 addition to, Ganeti's own log files
1241 - Removed boot restriction for paravirtual NICs for KVM, recent versions
1242 can indeed boot from a paravirtual NIC
1243 - Added a generic debug level for many operations; while this is not
1244 used widely yet, it allows one to pass the debug value all the way to
1246 - Enhanced the hooks environment for instance moves (failovers,
1247 migrations) where the primary/secondary nodes changed during the
1248 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
1249 - Enhanced data validations for many user-supplied values; one important
1250 item is the restrictions imposed on instance and node names, which
1251 might reject some (invalid) host names
1252 - Add a configure-time option to disable file-based storage, if it's not
1253 needed; this allows greater security separation between the master
1254 node and the other nodes from the point of view of the inter-node RPC
1256 - Added user notification in interactive tools if job is waiting in the
1257 job queue or trying to acquire locks
1258 - Added log messages when a job is waiting for locks
1259 - Added filtering by node tags in instance operations which admit
1260 multiple instances (start, stop, reboot, reinstall)
1261 - Added a new tool for cluster mergers, ``cluster-merge``
1262 - Parameters from command line which are of the form ``a=b,c=d`` can now
1263 use backslash escapes to pass in values which contain commas,
1264 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
1266 - For KVM, the instance name is the first parameter passed to KVM, so
1267 that it's more visible in the process list
1273 *(Released Tue, 2 Mar 2010)*
1275 Ganeti 2.1 brings many improvements with it. Major changes:
1277 - Added infrastructure to ease automated disk repairs
1278 - Added new daemon to export configuration data in a cheaper way than
1279 using the remote API
1280 - Instance NICs can now be routed instead of being associated with a
1282 - Improved job locking logic to reduce impact of jobs acquiring multiple
1283 locks waiting for other long-running jobs
1285 In-depth implementation details can be found in the Ganeti 2.1 design
1291 - Added chroot hypervisor
1292 - Added more options to xen-hvm hypervisor (``kernel_path`` and
1294 - Added more options to xen-pvm hypervisor (``use_bootloader``,
1295 ``bootloader_path`` and ``bootloader_args``)
1296 - Added the ``use_localtime`` option for the xen-hvm and kvm
1297 hypervisors, and the default value for this has changed to false (in
1298 2.0 xen-hvm always enabled it)
1299 - Added luxi call to submit multiple jobs in one go
1300 - Added cluster initialization option to not modify ``/etc/hosts``
1302 - Added network interface parameters
1303 - Added dry run mode to some LUs
1304 - Added RAPI resources:
1306 - ``/2/instances/[instance_name]/info``
1307 - ``/2/instances/[instance_name]/replace-disks``
1308 - ``/2/nodes/[node_name]/evacuate``
1309 - ``/2/nodes/[node_name]/migrate``
1310 - ``/2/nodes/[node_name]/role``
1311 - ``/2/nodes/[node_name]/storage``
1312 - ``/2/nodes/[node_name]/storage/modify``
1313 - ``/2/nodes/[node_name]/storage/repair``
1315 - Added OpCodes to evacuate or migrate all instances on a node
1316 - Added new command to list storage elements on nodes (``gnt-node
1317 list-storage``) and modify them (``gnt-node modify-storage``)
1318 - Added new ssconf files with master candidate IP address
1319 (``ssconf_master_candidates_ips``), node primary IP address
1320 (``ssconf_node_primary_ips``) and node secondary IP address
1321 (``ssconf_node_secondary_ips``)
1322 - Added ``ganeti-confd`` and a client library to query the Ganeti
1323 configuration via UDP
1324 - Added ability to run hooks after cluster initialization and before
1326 - Added automatic mode for disk replace (``gnt-instance replace-disks
1328 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
1329 after catastrophic data-loss
1330 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
1332 - Added ``gnt-instance move`` command to move instances
1333 - Added ``gnt-cluster watcher`` command to control watcher
1334 - Added ``gnt-node powercycle`` command to powercycle nodes
1335 - Added new job status field ``lock_status``
1336 - Added parseable error codes to cluster verification (``gnt-cluster
1337 verify --error-codes``) and made output less verbose (use
1338 ``--verbose`` to restore previous behaviour)
1339 - Added UUIDs to the main config entities (cluster, nodes, instances)
1340 - Added support for OS variants
1341 - Added support for hashed passwords in the Ganeti remote API users file
1343 - Added option to specify maximum timeout on instance shutdown
1344 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
1345 - Added new helper script to start and stop Ganeti daemons
1346 (``daemon-util``), with the intent to reduce the work necessary to
1347 adjust Ganeti for non-Debian distributions and to start/stop daemons
1349 - Added more unittests
1350 - Fixed critical bug in ganeti-masterd startup
1351 - Removed the configure-time ``kvm-migration-port`` parameter, this is
1352 now customisable at the cluster level for both the KVM and Xen
1353 hypervisors using the new ``migration_port`` parameter
1354 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
1355 reinstalling an instance
1356 - Allowed ``@`` in tag names
1357 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
1358 - Many documentation updates
1359 - Distribute hypervisor files on ``gnt-cluster redist-conf``
1360 - ``gnt-instance reinstall`` can now reinstall multiple instances
1361 - Updated many command line parameters
1362 - Introduced new OS API version 15
1363 - No longer support a default hypervisor
1364 - Treat virtual LVs as inexistent
1365 - Improved job locking logic to reduce lock contention
1366 - Match instance and node names case insensitively
1367 - Reimplemented bash completion script to be more complete
1374 *(Released Thu, 4 Feb 2010)*
1376 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
1377 - Fix a string formatting bug
1378 - Improve safety of the code in some error paths
1379 - Improve data validation in the master of values returned from nodes
1385 *(Released Thu, 17 Dec 2009)*
1387 - Fix security issue due to missing validation of iallocator names; this
1388 allows local and remote execution of arbitrary executables
1389 - Fix failure of gnt-node list during instance removal
1390 - Ship the RAPI documentation in the archive
1396 *(Released Wed, 30 Sep 2009)*
1398 - Fixed many wrong messages
1399 - Fixed a few bugs related to the locking library
1400 - Fixed MAC checking at instance creation time
1401 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
1402 - Fixed a few issues related to signal handling in both daemons and
1404 - Fixed the example startup script provided
1405 - Fixed insserv dependencies in the example startup script (patch from
1407 - Fixed handling of drained nodes in the iallocator framework
1408 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
1410 - Fixed error related to invalid job IDs in job polling
1411 - Fixed job/opcode persistence on unclean master shutdown
1412 - Fixed handling of partial job processing after unclean master
1414 - Fixed error reporting from LUs, previously all errors were converted
1415 into execution errors
1416 - Fixed error reporting from burnin
1417 - Decreased significantly the memory usage of the job queue
1418 - Optimised slightly multi-job submission
1419 - Optimised slightly opcode loading
1420 - Backported the multi-job submit framework from the development
1421 branch; multi-instance start and stop should be faster
1422 - Added script to clean archived jobs after 21 days; this will reduce
1423 the size of the queue directory
1424 - Added some extra checks in disk size tracking
1425 - Added an example ethers hook script
1426 - Added a cluster parameter that prevents Ganeti from modifying of
1428 - Added more node information to RAPI responses
1429 - Added a ``gnt-job watch`` command that allows following the ouput of a
1431 - Added a bind-address option to ganeti-rapi
1432 - Added more checks to the configuration verify
1433 - Enhanced the burnin script such that some operations can be retried
1435 - Converted instance reinstall to multi-instance model
1441 *(Released Fri, 7 Aug 2009)*
1443 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
1444 to allow using the pre-2.0.2 behaviour in activation, if any existing
1445 instances have mismatched disk sizes in the configuration
1446 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
1447 any configuration mismatches for disk sizes
1448 - Added ``gnt-master cluste-failover --no-voting`` to allow master
1449 failover to work on two-node clusters
1450 - Fixed the ``--net`` option of ``gnt-backup import``, which was
1452 - Fixed detection of OS script errors in ``gnt-backup export``
1453 - Fixed exit code of ``gnt-backup export``
1459 *(Released Fri, 17 Jul 2009)*
1461 - Added experimental support for stripped logical volumes; this should
1462 enhance performance but comes with a higher complexity in the block
1463 device handling; stripping is only enabled when passing
1464 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
1465 affected even in the non-stripped mode
1466 - Improved resiliency against transient failures at the end of DRBD
1467 resyncs, and in general of DRBD resync checks
1468 - Fixed a couple of issues with exports and snapshot errors
1469 - Fixed a couple of issues in instance listing
1470 - Added display of the disk size in ``gnt-instance info``
1471 - Fixed checking for valid OSes in instance creation
1472 - Fixed handling of the "vcpus" parameter in instance listing and in
1473 general of invalid parameters
1474 - Fixed http server library, and thus RAPI, to handle invalid
1475 username/password combinations correctly; this means that now they
1476 report unauthorized for queries too, not only for modifications,
1477 allowing earlier detect of configuration problems
1478 - Added a new "role" node list field, equivalent to the master/master
1479 candidate/drained/offline flags combinations
1480 - Fixed cluster modify and changes of candidate pool size
1481 - Fixed cluster verify error messages for wrong files on regular nodes
1482 - Fixed a couple of issues with node demotion from master candidate role
1483 - Fixed node readd issues
1484 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
1485 - Added a new ``--no-voting`` option for masterfailover to fix failover
1486 on two-nodes clusters when the former master node is unreachable
1487 - Added instance reinstall over RAPI
1493 *(Released Tue, 16 Jun 2009)*
1495 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
1496 allow re-adding the start in single-user option (regression from 1.2)
1497 - the watcher writes the instance status to a file, to allow monitoring
1498 to report the instance status (from the master) based on cached
1499 results of the watcher's queries; while this can get stale if the
1500 watcher is being locked due to other work on the cluster, this is
1501 still an improvement
1502 - the watcher now also restarts the node daemon and the rapi daemon if
1504 - fixed the watcher to handle full and drained queue cases
1505 - hooks export more instance data in the environment, which helps if
1506 hook scripts need to take action based on the instance's properties
1507 (no longer need to query back into ganeti)
1508 - instance failovers when the instance is stopped do not check for free
1509 RAM, so that failing over a stopped instance is possible in low memory
1511 - rapi uses queries for tags instead of jobs (for less job traffic), and
1512 for cluster tags it won't talk to masterd at all but read them from
1514 - a couple of error handling fixes in RAPI
1515 - drbd handling: improved the error handling of inconsistent disks after
1516 resync to reduce the frequency of "there are some degraded disks for
1517 this instance" messages
1518 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
1519 error handling path called a wrong function name)
1525 *(Released Wed, 27 May 2009)*
1527 - no changes from rc5
1530 Version 2.0 release candidate 5
1531 -------------------------------
1533 *(Released Wed, 20 May 2009)*
1535 - fix a couple of bugs (validation, argument checks)
1536 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
1537 - some small improvements to RAPI and IAllocator
1538 - make watcher automatically start the master daemon if down
1541 Version 2.0 release candidate 4
1542 -------------------------------
1544 *(Released Mon, 27 Apr 2009)*
1546 - change the OS list to not require locks; this helps with big clusters
1547 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
1548 volume group is broken
1549 - ``gnt-instance info``, without any arguments, doesn't run for all
1550 instances anymore; either pass ``--all`` or pass the desired
1551 instances; this helps against mistakes on big clusters where listing
1552 the information for all instances takes a long time
1553 - miscellaneous doc and man pages fixes
1556 Version 2.0 release candidate 3
1557 -------------------------------
1559 *(Released Wed, 8 Apr 2009)*
1561 - Change the internal locking model of some ``gnt-node`` commands, in
1562 order to reduce contention (and blocking of master daemon) when
1563 batching many creation/reinstall jobs
1564 - Fixes to Xen soft reboot
1565 - No longer build documentation at build time, instead distribute it in
1566 the archive, in order to reduce the need for the whole docbook/rst
1570 Version 2.0 release candidate 2
1571 -------------------------------
1573 *(Released Fri, 27 Mar 2009)*
1575 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
1576 - Fix watcher startup sequence, improves the behaviour of busy clusters
1577 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
1578 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
1579 bind address and other places
1580 - Some documentation fixes and updates
1583 Version 2.0 release candidate 1
1584 -------------------------------
1586 *(Released Mon, 2 Mar 2009)*
1588 - More documentation updates, now all docs should be more-or-less
1590 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
1592 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
1594 - Fix an issue related to $libdir/run/ganeti and cluster creation
1600 *(Released Thu, 19 Feb 2009)*
1602 - Xen PVM and KVM have switched the default value for the instance root
1603 disk to the first partition on the first drive, instead of the whole
1604 drive; this means that the OS installation scripts must be changed
1606 - Man pages have been updated
1607 - RAPI has been switched by default to HTTPS, and the exported functions
1608 should all work correctly
1609 - RAPI v1 has been removed
1610 - Many improvements to the KVM hypervisor
1611 - Block device errors are now better reported
1612 - Many other bugfixes and small improvements
1618 *(Released Mon, 26 Jan 2009)*
1620 - Version 2 is a general rewrite of the code and therefore the
1621 differences are too many to list, see the design document for 2.0 in
1622 the ``doc/`` subdirectory for more details
1623 - In this beta version there is not yet a migration path from 1.2 (there
1624 will be one in the final 2.0 release)
1625 - A few significant changes are:
1627 - all commands are executed by a daemon (``ganeti-masterd``) and the
1628 various ``gnt-*`` commands are just front-ends to it
1629 - all the commands are entered into, and executed from a job queue,
1630 see the ``gnt-job(8)`` manpage
1631 - the RAPI daemon supports read-write operations, secured by basic
1632 HTTP authentication on top of HTTPS
1633 - DRBD version 0.7 support has been removed, DRBD 8 is the only
1634 supported version (when migrating from Ganeti 1.2 to 2.0, you need
1635 to migrate to DRBD 8 first while still running Ganeti 1.2)
1636 - DRBD devices are using statically allocated minor numbers, which
1637 will be assigned to existing instances during the migration process
1638 - there is support for both Xen PVM and Xen HVM instances running on
1640 - KVM virtualization is supported too
1641 - file-based storage has been implemented, which means that it is
1642 possible to run the cluster without LVM and DRBD storage, for
1643 example using a shared filesystem exported from shared storage (and
1644 still have live migration)
1650 *(Released Tue, 13 Jan 2009)*
1652 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
1653 - Reuse the old instance mac address by default on instance import, if
1654 the instance name is the same.
1655 - Handle situations in which the node info rpc returns incomplete
1657 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
1658 - Improved version of batcher:
1660 - state file support
1661 - instance mac address support
1662 - support for HVM clusters/instances
1664 - Add an option to show the number of cpu sockets and nodes in
1666 - Support OSes that handle more than one version of the OS api (but do
1667 not change the current API in any other way)
1668 - Fix ``gnt-node migrate``
1669 - ``gnt-debug`` man page
1670 - Fixes various more typos and small issues
1671 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
1677 *(Released Wed, 24 Sep 2008)*
1679 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
1680 type of disk exported to fully virtualized instances.
1681 - provide access to the serial console of HVM instances
1682 - instance auto_balance flag, set by default. If turned off it will
1683 avoid warnings on cluster verify if there is not enough memory to fail
1684 over an instance. in the future it will prevent automatically failing
1685 it over when we will support that.
1686 - batcher tool for instance creation, see ``tools/README.batcher``
1687 - ``gnt-instance reinstall --select-os`` to interactively select a new
1688 operating system when reinstalling an instance.
1689 - when changing the memory amount on instance modify a check has been
1690 added that the instance will be able to start. also warnings are
1691 emitted if the instance will not be able to fail over, if auto_balance
1693 - documentation fixes
1694 - sync fields between ``gnt-instance list/modify/add/import``
1695 - fix a race condition in drbd when the sync speed was set after giving
1696 the device a remote peer.
1702 *(Released Tue, 22 Jul 2008)*
1704 - note: the allowed size and number of tags per object were reduced
1705 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
1706 - fixed twisted 8.x compatibility
1707 - fixed ``gnt-instance replace-disks`` with iallocator
1708 - add TCP keepalives on twisted connections to detect restarted nodes
1709 - disk increase support, see ``gnt-instance grow-disk``
1710 - implement bulk node/instance query for RAPI
1711 - add tags in node/instance listing (optional)
1712 - experimental migration (and live migration) support, read the man page
1713 for ``gnt-instance migrate``
1714 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
1715 has some small improvements in handling its state file
1721 *(Released Fri, 13 Jun 2008)*
1723 - Experimental readonly, REST-based remote API implementation;
1724 automatically started on master node, TCP port 5080, if enabled by
1725 ``--enable-rapi`` parameter to configure script.
1726 - Instance allocator support. Add and import instance accept a
1727 ``--iallocator`` parameter, and call that instance allocator to decide
1728 which node to use for the instance. The iallocator document describes
1729 what's expected from an allocator script.
1730 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
1731 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
1732 node is lost there is still enough memory to fail over the instances
1734 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
1735 ``gnt-cluster verify``, to check for site-specific compliance. All the
1736 hooks will run, and their output, if any, will be displayed. Any
1737 failing hook will make the verification return an error value.
1738 - ``gnt-cluster verify`` now checks that its peers are reachable on the
1739 primary and secondary interfaces
1740 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
1741 that is still declared as part of the cluster and has failed.
1742 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
1743 specifying output fields, that just adds the chosen fields to the
1745 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
1746 export from the filesystem.
1747 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
1748 have been added. Using them you can enable/disable acpi and pae
1749 support, and specify a path for a cd image to be exported to the
1750 instance. These parameters as the name suggest only work on HVM
1752 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
1753 PAE support will be set to the previously hardcoded values, but the
1754 (previously hardcoded) path to the CDROM ISO image will be unset and
1755 if required, needs to be set manually with ``gnt-instance modify``
1757 - The address to which an instance's VNC console is bound is now
1758 selectable per-instance, rather than being cluster wide. Of course
1759 this only applies to instances controlled via VNC, so currently just
1760 applies to HVM clusters.
1766 *(Released Mon, 18 Feb 2008)*
1768 - more tweaks to the disk activation code (especially helpful for DRBD)
1769 - change the default ``gnt-instance list`` output format, now there is
1770 one combined status field (see the manpage for the exact values this
1772 - some more fixes for the mac export to hooks change
1773 - make Ganeti not break with DRBD 8.2.x (which changed the version
1774 format in ``/proc/drbd``) (issue 24)
1775 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
1776 template, allowing migration from DRBD0.7+MD to DRBD8
1782 *(Released Wed, 30 Jan 2008)*
1784 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
1786 - add command aliases infrastructure and a few aliases
1787 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
1788 man pages and the ``--help`` option of ``gnt-node
1789 list``/``gnt-instance list``
1790 - fix ``gnt-backup list`` with down nodes (issue 21)
1791 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
1792 - fix the dist archive and add a check for including svn/git files in
1794 - some developer-related changes: improve the burnin and the QA suite,
1795 add an upload script for testing during development
1801 *(Released Wed, 16 Jan 2008)*
1803 - experimental HVM support, read the install document, section
1804 "Initializing the cluster"
1805 - allow for the PVM hypervisor per-instance kernel and initrd paths
1806 - add a new command ``gnt-cluster verify-disks`` which uses a new
1807 algorithm to improve the reconnection of the DRBD pairs if the device
1808 on the secondary node has gone away
1809 - make logical volume code auto-activate LVs at disk activation time
1810 - slightly improve the speed of activating disks
1811 - allow specification of the MAC address at instance creation time, and
1812 changing it later via ``gnt-instance modify``
1813 - fix handling of external commands that generate lots of output on
1815 - update documentation with regard to minimum version of DRBD8 supported
1821 *(Released Tue, 4 Dec 2007)*
1823 - Log the ``xm create`` output to the node daemon log on failure (to
1824 help diagnosing the error)
1825 - In debug mode, log all external commands output if failed to the logs
1826 - Change parsing of lvm commands to ignore stderr
1832 *(Released Wed, 28 Nov 2007)*
1834 - Another round of updates to the DRBD 8 code to deal with more failures
1835 in the replace secondary node operation
1836 - Some more logging of failures in disk operations (lvm, drbd)
1837 - A few documentation updates
1844 *(Released Tue, 13 Nov 2007)*
1846 - Change configuration file format from Python's Pickle to JSON.
1847 Upgrading is possible using the cfgupgrade utility.
1848 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
1849 faster replace disks and is more stable (DRBD 8 has many improvements
1850 compared to DRBD 0.7)
1851 - Added command line tags support (see man pages for ``gnt-instance``,
1852 ``gnt-node``, ``gnt-cluster``)
1853 - Added instance rename support
1854 - Added multi-instance startup/shutdown
1855 - Added cluster rename support
1856 - Added ``gnt-node evacuate`` to simplify some node operations
1857 - Added instance reboot operation that can speedup reboot as compared to
1859 - Soften the requirement that hostnames are in FQDN format
1860 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
1862 - Removed dependency on debian's patched fping that uses the
1863 non-standard ``-S`` option
1864 - Now the OS definitions are searched for in multiple, configurable
1865 paths (easier for distros to package)
1866 - Some changes to the hooks infrastructure (especially the new
1867 post-configuration update hook)
1868 - Other small bugfixes
1870 .. vim: set textwidth=72 syntax=rst :