10 - Deprecated ``admin_up`` field. Instead, ``admin_state`` is introduced,
11 with 3 possible values -- ``up``, ``down`` and ``offline``.
12 - Replaced ``--disks`` option of ``gnt-instance replace-disks`` with a
13 more flexible ``--disk`` option. Now disk size and mode can be changed
15 - Removed deprecated ``QueryLocks`` LUXI request. Use
16 ``Query(what=QR_LOCK, ...)`` instead.
22 *(Released Mon, 9 Jan 2012)*
24 Incompatible/important changes and bugfixes
25 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
27 - The default of the ``/2/instances/[instance_name]/rename`` RAPI
28 resource's ``ip_check`` parameter changed from ``True`` to ``False``
29 to match the underlying LUXI interface.
30 - The ``/2/nodes/[node_name]/evacuate`` RAPI resource was changed to use
31 body parameters, see :doc:`RAPI documentation <rapi>`. The server does
32 not maintain backwards-compatibility as the underlying operation
33 changed in an incompatible way. The RAPI client can talk to old
34 servers, but it needs to be told so as the return value changed.
35 - When creating file-based instances via RAPI, the ``file_driver``
36 parameter no longer defaults to ``loop`` and must be specified.
37 - The deprecated ``bridge`` NIC parameter is no longer supported. Use
39 - Support for the undocumented and deprecated RAPI instance creation
40 request format version 0 has been dropped. Use version 1, supported
41 since Ganeti 2.1.3 and :doc:`documented <rapi>`, instead.
42 - Pyparsing 1.4.6 or above is required, see :doc:`installation
43 documentation <install>`.
44 - The "cluster-verify" hooks are now executed per group by the
45 ``OP_CLUSTER_VERIFY_GROUP`` opcode. This maintains the same behavior
46 if you just run ``gnt-cluster verify``, which generates one opcode per
48 - The environment as passed to the OS scripts is cleared, and thus no
49 environment variables defined in the node daemon's environment will be
50 inherited by the scripts.
51 - The :doc:`iallocator <iallocator>` mode ``multi-evacuate`` has been
53 - :doc:`New iallocator modes <design-multi-reloc>` have been added to
54 support operations involving multiple node groups.
55 - Offline nodes are ignored when failing over an instance.
56 - Support for KVM version 1.0, which changed the version reporting format
58 - Includes all bugfixes made in the 2.4 series
63 - The ganeti-htools project has been merged into the ganeti-core source
64 tree and will be built as part of Ganeti (see :doc:`install-quick`).
65 - Implemented support for :doc:`shared storage <design-shared-storage>`.
66 - Add support for disks larger than 2 TB in ``lvmstrap`` by supporting
67 GPT-style partition tables (requires `parted
68 <http://www.gnu.org/s/parted/>`_).
69 - Added support for floppy drive and 2nd CD-ROM drive in KVM hypervisor.
70 - Allowed adding tags on instance creation.
71 - Export instance tags to hooks (``INSTANCE_TAGS``, see :doc:`hooks`)
72 - Allow instances to be started in a paused state, enabling the user to
73 see the complete console output on boot using the console.
74 - Added new hypervisor flag to control default reboot behaviour
75 (``reboot_behavior``).
76 - Added support for KVM keymaps (hypervisor parameter ``keymap``).
77 - Improved out-of-band management support:
79 - Added ``gnt-node health`` command reporting the health status of
81 - Added ``gnt-node power`` command to manage power status of nodes.
82 - Added command for emergency power-off (EPO), ``gnt-cluster epo``.
84 - Instance migration can fall back to failover if instance is not
86 - Filters can be used when listing nodes, instances, groups and locks;
87 see *ganeti(7)* manpage.
88 - Added post-execution status as variables to :doc:`hooks <hooks>`
90 - Instance tags are exported/imported together with the instance.
91 - When given an explicit job ID, ``gnt-job info`` will work for archived
93 - Jobs can define dependencies on other jobs (not yet supported via
94 RAPI or command line, but used by internal commands and usable via
97 - Lock monitor (``gnt-debug locks``) shows jobs waiting for
100 - Instance failover is now available as a RAPI resource
101 (``/2/instances/[instance_name]/failover``).
102 - ``gnt-instance info`` defaults to static information if primary node
104 - Opcodes have a new ``comment`` attribute.
105 - Added basic SPICE support to KVM hypervisor.
106 - ``tools/ganeti-listrunner`` allows passing of arguments to executable.
108 Node group improvements
109 ~~~~~~~~~~~~~~~~~~~~~~~
111 - ``gnt-cluster verify`` has been modified to check groups separately,
112 thereby improving performance.
113 - Node group support has been added to ``gnt-cluster verify-disks``,
114 which now operates per node group.
115 - Watcher has been changed to work better with node groups.
117 - One process and state file per node group.
118 - Slow watcher in one group doesn't block other group's watcher.
120 - Added new command, ``gnt-group evacuate``, to move all instances in a
121 node group to other groups.
122 - Added ``gnt-instance change-group`` to move an instance to another
124 - ``gnt-cluster command`` and ``gnt-cluster copyfile`` now support
125 per-group operations.
126 - Node groups can be tagged.
127 - Some operations switch from an exclusive to a shared lock as soon as
129 - Instance's primary and secondary nodes' groups are now available as
130 query fields (``pnode.group``, ``pnode.group.uuid``, ``snodes.group``
131 and ``snodes.group.uuid``).
136 - Numerous updates to documentation and manpages.
138 - :doc:`RAPI <rapi>` documentation now has detailed parameter
140 - Some opcode/job results are now also documented, see :doc:`RAPI
143 - A lockset's internal lock is now also visible in lock monitor.
144 - Log messages from job queue workers now contain information about the
145 opcode they're processing.
146 - ``gnt-instance console`` no longer requires the instance lock.
147 - A short delay when waiting for job changes reduces the number of LUXI
148 requests significantly.
149 - DRBD metadata volumes are overwritten with zeros during disk creation.
155 *(Released Thu, 27 Oct 2011)*
157 This was the fourth release candidate of the 2.5 series.
163 *(Released Wed, 26 Oct 2011)*
165 This was the third release candidate of the 2.5 series.
171 *(Released Tue, 18 Oct 2011)*
173 This was the second release candidate of the 2.5 series.
179 *(Released Tue, 4 Oct 2011)*
181 This was the first release candidate of the 2.5 series.
187 *(Released Wed, 31 Aug 2011)*
189 This was the third beta release of the 2.5 series.
195 *(Released Mon, 22 Aug 2011)*
197 This was the second beta release of the 2.5 series.
203 *(Released Fri, 12 Aug 2011)*
205 This was the first beta release of the 2.5 series.
211 *(Released Thu, 27 Oct 2011)*
213 - Fixed bug when parsing command line parameter values ending in
215 - Fixed assertion error after unclean master shutdown
216 - Disable HTTP client pool for RPC, significantly reducing memory usage
218 - Fixed queue archive creation with wrong permissions
224 *(Released Tue, 23 Aug 2011)*
228 - Fixed documentation for importing with ``--src-dir`` option
229 - Fixed a bug in ``ensure-dirs`` with queue/archive permissions
230 - Fixed a parsing issue with DRBD 8.3.11 in the Linux kernel
236 *(Released Fri, 5 Aug 2011)*
238 Many bug-fixes and a few small features:
240 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
241 issues when you tried to add an instance with two MAC addresses in one
243 - KVM: fixed per-instance stored UID value
244 - KVM: configure bridged NICs at migration start
245 - KVM: Fix a bug where instance will not start with never KVM versions
247 - Added OS search path to ``gnt-cluster info``
248 - Fixed an issue with ``file_storage_dir`` where you were forced to
249 provide an absolute path, but the documentation states it is a
250 relative path, the documentation was right
251 - Added a new parameter to instance stop/start called ``--no-remember``
252 that will make the state change to not be remembered
253 - Implemented ``no_remember`` at RAPI level
254 - Improved the documentation
255 - Node evacuation: don't call IAllocator if node is already empty
256 - Fixed bug in DRBD8 replace disks on current nodes
257 - Fixed bug in recreate-disks for DRBD instances
258 - Moved assertion checking locks in ``gnt-instance replace-disks``
259 causing it to abort with not owning the right locks for some situation
260 - Job queue: Fixed potential race condition when cancelling queued jobs
261 - Fixed off-by-one bug in job serial generation
262 - ``gnt-node volumes``: Fix instance names
263 - Fixed aliases in bash completion
264 - Fixed a bug in reopening log files after being sent a SIGHUP
265 - Added a flag to burnin to allow specifying VCPU count
266 - Bugfixes to non-root Ganeti configuration
272 *(Released Thu, 12 May 2011)*
274 Many bug-fixes and a few new small features:
276 - Fixed a bug related to log opening failures
277 - Fixed a bug in instance listing with orphan instances
278 - Fixed a bug which prevented resetting the cluster-level node parameter
279 ``oob_program`` to the default
280 - Many fixes related to the ``cluster-merge`` tool
281 - Fixed a race condition in the lock monitor, which caused failures
282 during (at least) creation of many instances in parallel
283 - Improved output for gnt-job info
284 - Removed the quiet flag on some ssh calls which prevented debugging
286 - Improved the N+1 failure messages in cluster verify by actually
287 showing the memory values (needed and available)
288 - Increased lock attempt timeouts so that when executing long operations
289 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
290 too early and thus prevent the use of the 'fair' mechanism
291 - Changed instance query data (``gnt-instance info``) to not acquire
292 locks unless needed, thus allowing its use on locked instance if only
293 static information is asked for
294 - Improved behaviour with filesystems that do not support rename on an
296 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
297 kept locks on all nodes during the wipe, which is unneeded
298 - Fixed ``gnt-watcher`` handling of errors during hooks execution
299 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
300 10GiB) which caused the wipe to fail right at the end in some cases
301 - Fixed master IP activation when doing master failover with no-voting
302 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
303 the master node itself
304 - Fixed potential data-loss in under disk full conditions, where Ganeti
305 wouldn't check correctly the return code and would consider
306 partially-written files 'correct'
307 - Fixed bug related to multiple VGs and DRBD disk replacing
308 - Added new disk parameter ``metavg`` that allows placement of the meta
309 device for DRBD in a different volume group
310 - Fixed error handling in the node daemon when the system libc doesn't
311 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
312 - Fixed lock release during replace-disks, which kept cluster-wide locks
313 when doing disk replaces with an iallocator script
314 - Added check for missing bridges in cluster verify
315 - Handle EPIPE errors while writing to the terminal better, so that
316 piping the output to e.g. ``less`` doesn't cause a backtrace
317 - Fixed rare case where a ^C during Luxi calls could have been
318 interpreted as server errors, instead of simply terminating
319 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
321 - Added a few more parameters to the KVM hypervisor, allowing a second
322 CDROM, custom disk type for CDROMs and a floppy image
323 - Removed redundant message in instance rename when the name is given
325 - Added option to ``gnt-instance recreate-disks`` to allow creating the
326 disks on new nodes, allowing recreation when the original instance
327 nodes are completely gone
328 - Added option when converting disk templates to DRBD to skip waiting
329 for the resync, in order to make the instance available sooner
330 - Added two new variables to the OS scripts environment (containing the
332 - Made the root_path and optional parameter for the xen-pvm hypervisor,
333 to allow use of ``pvgrub`` as bootloader
334 - Changed the instance memory modifications to only check out-of-memory
335 conditions on memory increases, and turned the secondary node warnings
336 into errors (they can still be overridden via ``--force``)
337 - Fixed the handling of a corner case when the Python installation gets
338 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
339 to execute a command that doesn't exist
340 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
341 node of the instance returned failures during instance shutdown; this
342 adds the option ``--ignore-consistency`` to gnt-instance move
344 And as usual, various improvements to the error messages, documentation
351 *(Released Wed, 09 Mar 2011)*
353 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
354 the RAPI users file if run twice (even with ``--dry-run``).
356 The release fixes that bug (nothing else changed).
362 *(Released Mon, 07 Mar 2011)*
364 Final 2.4.0 release. Just a few small fixes:
366 - Fixed RAPI node evacuate
367 - Fixed the kvm-ifup script
368 - Fixed internal error handling for special job cases
369 - Updated man page to specify the escaping feature for options
375 *(Released Mon, 28 Feb 2011)*
377 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
378 that this feature wiped the disks of the wrong instance, leading to loss
383 - Fixed title of query field containing instance name
384 - Expanded the glossary in the documentation
385 - Fixed one unittest (internal issue)
391 *(Released Mon, 21 Feb 2011)*
393 A number of bug fixes plus just a couple functionality changes.
395 On the user-visible side, the ``gnt-* list`` command output has changed
396 with respect to "special" field states. The current rc1 style of display
397 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
398 the default output mode special fields states are displayed as follows:
400 - Offline resource: ``*``
401 - Unavailable/not applicable: ``-``
402 - Data missing (RPC failure): ``?``
403 - Unknown field: ``??``
405 Another user-visible change is the addition of ``--force-join`` to
410 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
411 - Fixed regression in RAPI/instance reinstall where all parameters were
412 required (instead of optional)
413 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
414 - Fixed iallocator usage (offline nodes were not considered offline)
415 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
416 - Fixed hypervisor and OS parameter validation with respect to
418 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
420 - Fixed ``tools/listrunner`` with respect to agent-based usage
426 *(Released Fri, 4 Feb 2011)*
428 Many changes and fixes since the beta1 release. While there were some
429 internal changes, the code has been mostly stabilised for the RC
432 Note: the dumb allocator was removed in this release, as it was not kept
433 up-to-date with the IAllocator protocol changes. It is recommended to
434 use the ``hail`` command from the ganeti-htools package.
436 Note: the 2.4 and up versions of Ganeti are not compatible with the
437 0.2.x branch of ganeti-htools. You need to upgrade to
438 ganeti-htools-0.3.0 (or later).
440 Regressions fixed from 2.3
441 ~~~~~~~~~~~~~~~~~~~~~~~~~~
443 - Fixed the ``gnt-cluster verify-disks`` command
444 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
446 - Fixed disk adoption breakage
447 - Fixed wrong headers in instance listing for field aliases
452 - Fixed corner case in KVM handling of NICs
453 - Fixed many cases of wrong handling of non-vm_capable nodes
454 - Fixed a bug where a missing instance symlink was not possible to
455 recreate with any ``gnt-*`` command (now ``gnt-instance
456 activate-disks`` does it)
457 - Fixed the volume group name as reported by ``gnt-cluster
459 - Increased timeouts for the import-export code, hopefully leading to
460 fewer aborts due network or instance timeouts
461 - Fixed bug in ``gnt-node list-storage``
462 - Fixed bug where not all daemons were started on cluster
463 initialisation, but only at the first watcher run
464 - Fixed many bugs in the OOB implementation
465 - Fixed watcher behaviour in presence of instances with offline
467 - Fixed instance list output for instances running on the wrong node
468 - a few fixes to the cluster-merge tool, but it still cannot merge
469 multi-node groups (currently it is not recommended to use this tool)
475 - Improved network configuration for the KVM hypervisor
476 - Added e1000 as a supported NIC for Xen-HVM
477 - Improved the lvmstrap tool to also be able to use partitions, as
478 opposed to full disks
479 - Improved speed of disk wiping (the cluster parameter
480 ``prealloc_wipe_disks``, so that it has a low impact on the total time
481 of instance creations
482 - Added documentation for the OS parameters
483 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
484 hypervisor is not responding
485 - Added display of blacklisted and hidden OS information in
487 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
488 NIC and node parameters, which might create problems with currently
489 invalid (but undetected) configuration files, but prevents validation
490 failures when unrelated parameters are modified
491 - Changed cluster initialisation to wait for the master daemon to become
493 - Expanded the RAPI interface:
495 - Added config redistribution resource
496 - Added activation/deactivation of instance disks
497 - Added export of console information
499 - Implemented log file reopening on SIGHUP, which allows using
500 logrotate(8) for the Ganeti log files
501 - Added a basic OOB helper script as an example
507 *(Released Fri, 14 Jan 2011)*
512 - Fixed timezone issues when formatting timestamps
513 - Added support for node groups, available via ``gnt-group`` and other
515 - Added out-of-band framework and management, see :doc:`design
516 document <design-oob>`
517 - Removed support for roman numbers from ``gnt-node list`` and
518 ``gnt-instance list``.
519 - Allowed modification of master network interface via ``gnt-cluster
520 modify --master-netdev``
521 - Accept offline secondaries while shutting down instance disks
522 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
523 - Added support for multiple LVM volume groups
524 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
526 - Added commands to list available fields:
528 - ``gnt-node list-fields``
529 - ``gnt-group list-fields``
530 - ``gnt-instance list-fields``
532 - Updated documentation and man pages
537 - Moved ``rapi_users`` file into separate directory, now named
538 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
540 - Added new tool for running commands on many machines,
541 ``tools/ganeti-listrunner``
542 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
543 improving the ``gnt-instance console`` output
544 - Allowed customisation of disk index separator at ``configure`` time
545 - Export node group allocation policy to :doc:`iallocator <iallocator>`
546 - Added support for non-partitioned md disks in ``lvmstrap``
547 - Added script to gracefully power off KVM instances
548 - Split ``utils`` module into smaller parts
549 - Changed query operations to return more detailed information, e.g.
550 whether an information is unavailable due to an offline node. To use
551 this new functionality, the LUXI call ``Query`` must be used. Field
552 information is now stored by the master daemon and can be retrieved
553 using ``QueryFields``. Instances, nodes and groups can also be queried
554 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
555 exposed via RAPI). The following commands make use of this
556 infrastructure change:
559 - ``gnt-group list-fields``
561 - ``gnt-node list-fields``
562 - ``gnt-instance list``
563 - ``gnt-instance list-fields``
564 - ``gnt-debug locks``
569 - New RAPI resources (see :doc:`rapi`):
573 - ``/2/groups/[group_name]``
574 - ``/2/groups/[group_name]/assign-nodes``
575 - ``/2/groups/[group_name]/modify``
576 - ``/2/groups/[group_name]/rename``
577 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
581 - Implemented ``no_install`` for instance creation
582 - Implemented OS parameters for instance reinstallation, allowing
583 use of special settings on reinstallation (e.g. for preserving data)
588 - Added IPv6 support in import/export
589 - Pause DRBD synchronization while wiping disks on instance creation
590 - Updated unittests and QA scripts
591 - Improved network parameters passed to KVM
592 - Converted man pages from docbook to reStructuredText
598 *(Released Mon, 20 Dec 2010)*
600 Released version 2.3.1~rc1 without any changes.
606 *(Released Wed, 1 Dec 2010)*
608 - impexpd: Disable OpenSSL compression in socat if possible (backport
609 from master, commit e90739d625b, see :doc:`installation guide
610 <install-quick>` for details)
611 - Changed unittest coverage report to exclude test scripts
612 - Added script to check version format
618 *(Released Wed, 1 Dec 2010)*
620 Released version 2.3.0~rc1 without any changes.
626 *(Released Fri, 19 Nov 2010)*
628 A number of bugfixes and documentation updates:
630 - Update ganeti-os-interface documentation
631 - Fixed a bug related to duplicate MACs or similar items which should be
633 - Fix breakage in OS state modify
634 - Reinstall instance: disallow offline secondaries (fixes bug related to
635 OS changing but reinstall failing)
636 - plus all the other fixes between 2.2.1 and 2.2.2
642 *(Released Tue, 2 Nov 2010)*
644 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
645 - Fixed master failover race with watcher
646 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
648 - Accept previously stopped instance for export with instance removal
649 - Simplify and extend the environment variables for instance OS scripts
650 - Added new node flags, ``master_capable`` and ``vm_capable``
651 - Added optional instance disk wiping prior during allocation. This is a
652 cluster-wide option and can be set/modified using
653 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
654 - Added IPv6 support, see :doc:`design document <design-2.3>` and
656 - Added a new watcher option (``--ignore-pause``)
657 - Added option to ignore offline node on instance start/stop
658 (``--ignore-offline``)
659 - Allow overriding OS parameters with ``gnt-instance reinstall``
660 - Added ability to change node's secondary IP address using ``gnt-node
662 - Implemented privilege separation for all daemons except
663 ``ganeti-noded``, see ``configure`` options
664 - Complain if an instance's disk is marked faulty in ``gnt-cluster
666 - Implemented job priorities (see ``ganeti(7)`` manpage)
667 - Ignore failures while shutting down instances during failover from
669 - Exit daemon's bootstrap process only once daemon is ready
670 - Export more information via ``LUInstanceQuery``/remote API
671 - Improved documentation, QA and unittests
672 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
673 restart if the file was created or changed
674 - Added LUXI protocol version sent with each request and response,
675 allowing detection of server/client mismatches
676 - Moved the Python scripts among gnt-* and ganeti-* into modules
677 - Moved all code related to setting up SSH to an external script,
679 - Infrastructure changes for node group support in future versions
685 *(Released Fri, 19 Nov 2010)*
687 A few small bugs fixed, and some improvements to the build system:
689 - Fix documentation regarding conversion to drbd
690 - Fix validation of parameters in cluster modify (``gnt-cluster modify
692 - Fix error handling in node modify with multiple changes
693 - Allow remote imports without checked names
699 *(Released Tue, 19 Oct 2010)*
701 - Disable SSL session ID cache in RPC client
707 *(Released Thu, 14 Oct 2010)*
709 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
710 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
712 - Fix problems with interaction between Curl and Python's HTTP server,
713 resulting in increased speed in many RPC calls
714 - Improve our release script to prevent breakage with older aclocal and
721 *(Released Thu, 7 Oct 2010)*
723 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
724 value retrieved from master
725 - Added support for blacklisted or hidden OS definitions
726 - Added simple lock monitor (accessible via (``gnt-debug locks``)
727 - Added support for -mem-path in KVM hypervisor abstraction layer
728 - Allow overriding instance parameters in tool for inter-cluster
729 instance moves (``tools/move-instance``)
730 - Improved opcode summaries (e.g. in ``gnt-job list``)
731 - Improve consistency of OS listing by sorting it
732 - Documentation updates
738 *(Released Fri, 8 Oct 2010)*
740 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
746 *(Released Mon, 4 Oct 2010)*
748 - Fixed regression in ``gnt-instance rename``
754 *(Released Wed, 22 Sep 2010)*
756 - Fixed OS_VARIANT variable for OS scripts
757 - Fixed cluster tag operations via RAPI
758 - Made ``setup-ssh`` exit with non-zero code if an error occurred
759 - Disabled RAPI CA checks in watcher
765 *(Released Mon, 23 Aug 2010)*
767 - Support DRBD versions of the format "a.b.c.d"
769 - Re-introduce support for usage from multiple threads in RAPI client
770 - Instance renames and modify via RAPI
771 - Work around race condition between processing and archival in job
773 - Mark opcodes following failed one as failed, too
774 - Job field ``lock_status`` was removed due to difficulties making it
775 work with the changed job queue in Ganeti 2.2; a better way to monitor
776 locks is expected for a later 2.2.x release
777 - Fixed dry-run behaviour with many commands
778 - Support ``ssh-agent`` again when adding nodes
779 - Many additional bugfixes
785 *(Released Fri, 30 Jul 2010)*
787 Important change: the internal RPC mechanism between Ganeti nodes has
788 changed from using a home-grown http library (based on the Python base
789 libraries) to use the PycURL library. This requires that PycURL is
790 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
791 against GnuTLS by default. cURL's support for GnuTLS had known issues
792 before cURL 7.21.0 and we recommend using the latest cURL release or
793 linking against OpenSSL. Most other distributions already link PycURL
794 and cURL against OpenSSL. The command::
796 python -c 'import pycurl; print pycurl.version'
798 can be used to determine the libraries PycURL and cURL are linked
801 Other significant changes:
803 - Rewrote much of the internals of the job queue, in order to achieve
804 better parallelism; this decouples job query operations from the job
805 processing, and it should allow much nicer behaviour of the master
806 daemon under load, and it also has uncovered some long-standing bugs
807 related to the job serialisation (now fixed)
808 - Added a default iallocator setting to the cluster parameters,
809 eliminating the need to always pass nodes or an iallocator for
810 operations that require selection of new node(s)
811 - Added experimental support for the LXC virtualization method
812 - Added support for OS parameters, which allows the installation of
813 instances to pass parameter to OS scripts in order to customise the
815 - Added a hypervisor parameter controlling the migration type (live or
816 non-live), since hypervisors have various levels of reliability; this
817 has renamed the 'live' parameter to 'mode'
818 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
819 logical volumes, meaning that cluster verify will ignore them and not
820 flag their presence as errors
821 - The watcher will now reset the error count for failed instances after
822 8 hours, thus allowing self-healing if the problem that caused the
823 instances to be down/fail to start has cleared in the meantime
824 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
825 check for, and warn, if the drbd module parameter ``usermode_helper``
826 is not consistent with the cluster-wide setting; this is needed to
827 make diagnose easier of failed drbd creations
828 - Started adding base IPv6 support, but this is not yet
829 enabled/available for use
830 - Rename operations (cluster, instance) will now return the new name,
831 which is especially useful if a short name was passed in
832 - Added support for instance migration in RAPI
833 - Added a tool to pre-configure nodes for the SSH setup, before joining
834 them to the cluster; this will allow in the future a simplified model
835 for node joining (but not yet fully enabled in 2.2); this needs the
836 paramiko python library
837 - Fixed handling of name-resolving errors
838 - Fixed consistency of job results on the error path
839 - Fixed master-failover race condition when executed multiple times in
841 - Fixed many bugs related to the job queue (mostly introduced during the
842 2.2 development cycle, so not all are impacting 2.1)
843 - Fixed instance migration with missing disk symlinks
844 - Fixed handling of unknown jobs in ``gnt-job archive``
845 - And many other small fixes/improvements
849 - Enhanced both the unittest and the QA coverage
850 - Switched the opcode validation to a generic model, and extended the
851 validation to all opcode parameters
852 - Changed more parts of the code that write shell scripts to use the
854 - Switched the master daemon to use the asyncore library for the Luxi
861 *(Released Thu, 17 Jun 2010)*
863 - Added tool (``move-instance``) and infrastructure to move instances
864 between separate clusters (see :doc:`separate documentation
865 <move-instance>` and :doc:`design document <design-2.2>`)
866 - Added per-request RPC timeout
867 - RAPI now requires a Content-Type header for requests with a body (e.g.
868 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
869 :rfc:`2616` (HTTP/1.1), section 7.2.1)
870 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
872 - Implemented initial support for running Ganeti daemons as separate
873 users, see configure-time flags ``--with-user-prefix`` and
874 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
876 - Instances can be removed after export (``gnt-backup export
878 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
879 key (instead of 1024 bit)
880 - Added new cluster configuration file for cluster domain secret
881 - Import/export now use SSL instead of SSH
882 - Added support for showing estimated time when exporting an instance,
883 see the ``ganeti-os-interface(7)`` manpage and look for
890 *(Released Tue, 16 Nov 2010)*
892 Some more bugfixes. Unless critical bugs occur, this will be the last
895 - Fix case of MAC special-values
896 - Fix mac checker regex
897 - backend: Fix typo causing "out of range" error
898 - Add missing --units in gnt-instance list man page
904 *(Released Tue, 24 Aug 2010)*
907 - Don't ignore secondary node silently on non-mirrored disk templates
909 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
910 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
911 - Properly document the usb_mouse parameter
912 - Fix path in ganeti-rapi(8) (issue 116)
913 - Adjust error message when the ganeti user's .ssh directory is
915 - Add same-node-check when changing the disk template to drbd
921 *(Released Fri, 16 Jul 2010)*
924 - Add an option to only select some reboot types during qa/burnin.
925 (on some hypervisors consequent reboots are not supported)
926 - Fix infrequent race condition in master failover. Sometimes the old
927 master ip address would be still detected as up for a short time
928 after it was removed, causing failover to fail.
929 - Decrease mlockall warnings when the ctypes module is missing. On
930 Python 2.4 we support running even if no ctypes module is installed,
931 but we were too verbose about this issue.
932 - Fix building on old distributions, on which man doesn't have a
934 - Fix RAPI not to ignore the MAC address on instance creation
935 - Implement the old instance creation format in the RAPI client.
941 *(Released Thu, 01 Jul 2010)*
943 A small bugfix release:
944 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
945 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
946 a non-existing option
947 - Fix instance create: the --force-variant option was ignored
948 - Improve pylint 0.21 compatibility and warnings with Python 2.6
949 - Fix modify node storage with non-FQDN arguments
950 - Fix RAPI client to authenticate under Python 2.6 when used
951 for more than 5 requests needing authentication
952 - Fix gnt-instance modify -t (storage) giving a wrong error message
953 when converting a non-shutdown drbd instance to plain
959 *(Released Fri, 18 Jun 2010)*
961 A small bugfix release:
963 - Fix live migration of KVM instances started with older Ganeti
964 versions which had fewer hypervisor parameters
965 - Fix gnt-instance grow-disk on down instances
966 - Fix an error-reporting bug during instance migration
967 - Better checking of the ``--net`` and ``--disk`` values, to avoid
968 silently ignoring broken ones
969 - Fix an RPC error reporting bug affecting, for example, RAPI client
971 - Fix bug triggered by different API version os-es on different nodes
972 - Fix a bug in instance startup with custom hvparams: OS level
973 parameters would fail to be applied.
974 - Fix the RAPI client under Python 2.6 (but more work is needed to
975 make it work completely well with OpenSSL)
976 - Fix handling of errors when resolving names from DNS
982 *(Released Thu, 3 Jun 2010)*
984 A medium sized development cycle. Some new features, and some
985 fixes/small improvements/cleanups.
990 The node deamon now tries to mlock itself into memory, unless the
991 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
992 its logs, and falls back to console logging. This allows emergency
993 features such as ``gnt-node powercycle`` to work even in the event of a
994 broken node disk (tested offlining the disk hosting the node's
995 filesystem and dropping its memory caches; don't try this at home)
997 KVM: add vhost-net acceleration support. It can be tested with a new
998 enough version of the kernel and of qemu-kvm.
1000 KVM: Add instance chrooting feature. If you use privilege dropping for
1001 your VMs you can also now force them to chroot to an empty directory,
1002 before starting the emulated guest.
1004 KVM: Add maximum migration bandwith and maximum downtime tweaking
1005 support (requires a new-enough version of qemu-kvm).
1007 Cluster verify will now warn if the master node doesn't have the master
1008 ip configured on it.
1010 Add a new (incompatible) instance creation request format to RAPI which
1011 supports all parameters (previously only a subset was supported, and it
1012 wasn't possible to extend the old format to accomodate all the new
1013 features. The old format is still supported, and a client can check for
1014 this feature, before using it, by checking for its presence in the
1015 ``features`` RAPI resource.
1017 Now with ancient latin support. Try it passing the ``--roman`` option to
1018 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
1019 (requires the python-roman module to be installed, in order to work).
1024 As usual many internal code refactorings, documentation updates, and
1027 - Lots of improvements and cleanups to the experimental Remote API
1028 (RAPI) client library.
1029 - A new unit test suite for the core daemon libraries.
1030 - A fix to creating missing directories makes sure the umask is not
1031 applied anymore. This enforces the same directory permissions
1033 - Better handling terminating daemons with ctrl+c (used when running
1034 them in debugging mode).
1035 - Fix a race condition in live migrating a KVM instance, when stat()
1036 on the old proc status file returned EINVAL, which is an unexpected
1038 - Fixed manpage checking with newer man and utf-8 charachters. But now
1039 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
1045 *(Released Fri, 7 May 2010)*
1047 Fix a bug which prevented untagged KVM instances from starting.
1053 *(Released Fri, 7 May 2010)*
1055 Another release with a long development cycle, during which many
1056 different features were added.
1058 Significant features
1059 ~~~~~~~~~~~~~~~~~~~~
1061 The KVM hypervisor now can run the individual instances as non-root, to
1062 reduce the impact of a VM being hijacked due to bugs in the
1063 hypervisor. It is possible to run all instances as a single (non-root)
1064 user, to manually specify a user for each instance, or to dynamically
1065 allocate a user out of a cluster-wide pool to each instance, with the
1066 guarantee that no two instances will run under the same user ID on any
1069 An experimental RAPI client library, that can be used standalone
1070 (without the other Ganeti libraries), is provided in the source tree as
1071 ``lib/rapi/client.py``. Note this client might change its interface in
1072 the future, as we iterate on its capabilities.
1074 A new command, ``gnt-cluster renew-crypto`` has been added to easily
1075 replace the cluster's certificates and crypto keys. This might help in
1076 case they have been compromised, or have simply expired.
1078 A new disk option for instance creation has been added that allows one
1079 to "adopt" currently existing logical volumes, with data
1080 preservation. This should allow easier migration to Ganeti from
1081 unmanaged (or managed via other software) instances.
1083 Another disk improvement is the possibility to convert between redundant
1084 (DRBD) and plain (LVM) disk configuration for an instance. This should
1085 allow better scalability (starting with one node and growing the
1086 cluster, or shrinking a two-node cluster to one node).
1088 A new feature that could help with automated node failovers has been
1089 implemented: if a node sees itself as offline (by querying the master
1090 candidates), it will try to shutdown (hard) all instances and any active
1091 DRBD devices. This reduces the risk of duplicate instances if an
1092 external script automatically failovers the instances on such nodes. To
1093 enable this, the cluster parameter ``maintain_node_health`` should be
1094 enabled; in the future this option (per the name) will enable other
1095 automatic maintenance features.
1097 Instance export/import now will reuse the original instance
1098 specifications for all parameters; that means exporting an instance,
1099 deleting it and the importing it back should give an almost identical
1100 instance. Note that the default import behaviour has changed from
1101 before, where it created only one NIC; now it recreates the original
1104 Cluster verify has added a few new checks: SSL certificates validity,
1105 /etc/hosts consistency across the cluster, etc.
1110 As usual, many internal changes were done, documentation fixes,
1113 - Fixed cluster initialization with disabled cluster storage (regression
1114 introduced in 2.1.1)
1115 - File-based storage supports growing the disks
1116 - Fixed behaviour of node role changes
1117 - Fixed cluster verify for some corner cases, plus a general rewrite of
1118 cluster verify to allow future extension with more checks
1119 - Fixed log spamming by watcher and node daemon (regression introduced
1121 - Fixed possible validation issues when changing the list of enabled
1123 - Fixed cleanup of /etc/hosts during node removal
1124 - Fixed RAPI response for invalid methods
1125 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
1126 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
1127 from ide disks, etc.)
1128 - Allow OS changes without re-installation (to record a changed OS
1129 outside of Ganeti, or to allow OS renames)
1130 - Allow instance creation without OS installation (useful for example if
1131 the OS will be installed manually, or restored from a backup not in
1133 - Implemented option to make cluster ``copyfile`` use the replication
1135 - Added list of enabled hypervisors to ssconf (possibly useful for
1137 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
1139 - A partial form of node re-IP is possible via node readd, which now
1140 allows changed node primary IP
1141 - Command line utilities now show an informational message if the job is
1143 - The logs of the master daemon now show the PID/UID/GID of the
1150 *(Released Fri, 12 Mar 2010)*
1152 During the 2.1.0 long release candidate cycle, a lot of improvements and
1153 changes have accumulated with were released later as 2.1.1.
1158 The node evacuate command (``gnt-node evacuate``) was significantly
1159 rewritten, and as such the IAllocator protocol was changed - a new
1160 request type has been added. This unfortunate change during a stable
1161 series is designed to improve performance of node evacuations; on
1162 clusters with more than about five nodes and which are well-balanced,
1163 evacuation should proceed in parallel for all instances of the node
1164 being evacuated. As such, any existing IAllocator scripts need to be
1165 updated, otherwise the above command will fail due to the unknown
1166 request. The provided "dumb" allocator has not been updated; but the
1167 ganeti-htools package supports the new protocol since version 0.2.4.
1169 Another important change is increased validation of node and instance
1170 names. This might create problems in special cases, if invalid host
1171 names are being used.
1173 Also, a new layer of hypervisor parameters has been added, that sits at
1174 OS level between the cluster defaults and the instance ones. This allows
1175 customisation of virtualization parameters depending on the installed
1176 OS. For example instances with OS 'X' may have a different KVM kernel
1177 (or any other parameter) than the cluster defaults. This is intended to
1178 help managing a multiple OSes on the same cluster, without manual
1179 modification of each instance's parameters.
1181 A tool for merging clusters, ``cluster-merge``, has been added in the
1182 tools sub-directory.
1187 - Improved the int/float conversions that should make the code more
1188 robust in face of errors from the node daemons
1189 - Fixed the remove node code in case of internal configuration errors
1190 - Fixed the node daemon behaviour in face of inconsistent queue
1191 directory (e.g. read-only file-system where we can't open the files
1193 - Fixed the behaviour of gnt-node modify for master candidate demotion;
1194 now it either aborts cleanly or, if given the new "auto_promote"
1195 parameter, will automatically promote other nodes as needed
1196 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
1197 completely prevent Ganeti from working
1198 - Fixed bug for instance export when not all disks were successfully
1200 - Fixed behaviour of node add when the new node is slow in starting up
1202 - Fixed handling of signals in the LUXI client, which should improve
1203 behaviour of command-line scripts
1204 - Added checks for invalid node/instance names in the configuration (now
1205 flagged during cluster verify)
1206 - Fixed watcher behaviour for disk activation errors
1207 - Fixed two potentially endless loops in http library, which led to the
1208 RAPI daemon hanging and consuming 100% CPU in some cases
1209 - Fixed bug in RAPI daemon related to hashed passwords
1210 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
1212 - Enhanced compatibility with non-Debian OSes, but not using absolute
1213 path in some commands and allowing customisation of the ssh
1214 configuration directory
1215 - Fixed possible future issue with new Python versions by abiding to the
1216 proper use of ``__slots__`` attribute on classes
1217 - Added checks that should prevent directory traversal attacks
1218 - Many documentation fixes based on feedback from users
1223 - Added an "early_release" more for instance replace disks and node
1224 evacuate, where we release locks earlier and thus allow higher
1225 parallelism within the cluster
1226 - Added watcher hooks, intended to allow the watcher to restart other
1227 daemons (e.g. from the ganeti-nbma project), but they can be used of
1228 course for any other purpose
1229 - Added a compile-time disable for DRBD barriers, to increase
1230 performance if the administrator trusts the power supply or the
1231 storage system to not lose writes
1232 - Added the option of using syslog for logging instead of, or in
1233 addition to, Ganeti's own log files
1234 - Removed boot restriction for paravirtual NICs for KVM, recent versions
1235 can indeed boot from a paravirtual NIC
1236 - Added a generic debug level for many operations; while this is not
1237 used widely yet, it allows one to pass the debug value all the way to
1239 - Enhanced the hooks environment for instance moves (failovers,
1240 migrations) where the primary/secondary nodes changed during the
1241 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
1242 - Enhanced data validations for many user-supplied values; one important
1243 item is the restrictions imposed on instance and node names, which
1244 might reject some (invalid) host names
1245 - Add a configure-time option to disable file-based storage, if it's not
1246 needed; this allows greater security separation between the master
1247 node and the other nodes from the point of view of the inter-node RPC
1249 - Added user notification in interactive tools if job is waiting in the
1250 job queue or trying to acquire locks
1251 - Added log messages when a job is waiting for locks
1252 - Added filtering by node tags in instance operations which admit
1253 multiple instances (start, stop, reboot, reinstall)
1254 - Added a new tool for cluster mergers, ``cluster-merge``
1255 - Parameters from command line which are of the form ``a=b,c=d`` can now
1256 use backslash escapes to pass in values which contain commas,
1257 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
1259 - For KVM, the instance name is the first parameter passed to KVM, so
1260 that it's more visible in the process list
1266 *(Released Tue, 2 Mar 2010)*
1268 Ganeti 2.1 brings many improvements with it. Major changes:
1270 - Added infrastructure to ease automated disk repairs
1271 - Added new daemon to export configuration data in a cheaper way than
1272 using the remote API
1273 - Instance NICs can now be routed instead of being associated with a
1275 - Improved job locking logic to reduce impact of jobs acquiring multiple
1276 locks waiting for other long-running jobs
1278 In-depth implementation details can be found in the Ganeti 2.1 design
1284 - Added chroot hypervisor
1285 - Added more options to xen-hvm hypervisor (``kernel_path`` and
1287 - Added more options to xen-pvm hypervisor (``use_bootloader``,
1288 ``bootloader_path`` and ``bootloader_args``)
1289 - Added the ``use_localtime`` option for the xen-hvm and kvm
1290 hypervisors, and the default value for this has changed to false (in
1291 2.0 xen-hvm always enabled it)
1292 - Added luxi call to submit multiple jobs in one go
1293 - Added cluster initialization option to not modify ``/etc/hosts``
1295 - Added network interface parameters
1296 - Added dry run mode to some LUs
1297 - Added RAPI resources:
1299 - ``/2/instances/[instance_name]/info``
1300 - ``/2/instances/[instance_name]/replace-disks``
1301 - ``/2/nodes/[node_name]/evacuate``
1302 - ``/2/nodes/[node_name]/migrate``
1303 - ``/2/nodes/[node_name]/role``
1304 - ``/2/nodes/[node_name]/storage``
1305 - ``/2/nodes/[node_name]/storage/modify``
1306 - ``/2/nodes/[node_name]/storage/repair``
1308 - Added OpCodes to evacuate or migrate all instances on a node
1309 - Added new command to list storage elements on nodes (``gnt-node
1310 list-storage``) and modify them (``gnt-node modify-storage``)
1311 - Added new ssconf files with master candidate IP address
1312 (``ssconf_master_candidates_ips``), node primary IP address
1313 (``ssconf_node_primary_ips``) and node secondary IP address
1314 (``ssconf_node_secondary_ips``)
1315 - Added ``ganeti-confd`` and a client library to query the Ganeti
1316 configuration via UDP
1317 - Added ability to run hooks after cluster initialization and before
1319 - Added automatic mode for disk replace (``gnt-instance replace-disks
1321 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
1322 after catastrophic data-loss
1323 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
1325 - Added ``gnt-instance move`` command to move instances
1326 - Added ``gnt-cluster watcher`` command to control watcher
1327 - Added ``gnt-node powercycle`` command to powercycle nodes
1328 - Added new job status field ``lock_status``
1329 - Added parseable error codes to cluster verification (``gnt-cluster
1330 verify --error-codes``) and made output less verbose (use
1331 ``--verbose`` to restore previous behaviour)
1332 - Added UUIDs to the main config entities (cluster, nodes, instances)
1333 - Added support for OS variants
1334 - Added support for hashed passwords in the Ganeti remote API users file
1336 - Added option to specify maximum timeout on instance shutdown
1337 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
1338 - Added new helper script to start and stop Ganeti daemons
1339 (``daemon-util``), with the intent to reduce the work necessary to
1340 adjust Ganeti for non-Debian distributions and to start/stop daemons
1342 - Added more unittests
1343 - Fixed critical bug in ganeti-masterd startup
1344 - Removed the configure-time ``kvm-migration-port`` parameter, this is
1345 now customisable at the cluster level for both the KVM and Xen
1346 hypervisors using the new ``migration_port`` parameter
1347 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
1348 reinstalling an instance
1349 - Allowed ``@`` in tag names
1350 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
1351 - Many documentation updates
1352 - Distribute hypervisor files on ``gnt-cluster redist-conf``
1353 - ``gnt-instance reinstall`` can now reinstall multiple instances
1354 - Updated many command line parameters
1355 - Introduced new OS API version 15
1356 - No longer support a default hypervisor
1357 - Treat virtual LVs as inexistent
1358 - Improved job locking logic to reduce lock contention
1359 - Match instance and node names case insensitively
1360 - Reimplemented bash completion script to be more complete
1367 *(Released Thu, 4 Feb 2010)*
1369 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
1370 - Fix a string formatting bug
1371 - Improve safety of the code in some error paths
1372 - Improve data validation in the master of values returned from nodes
1378 *(Released Thu, 17 Dec 2009)*
1380 - Fix security issue due to missing validation of iallocator names; this
1381 allows local and remote execution of arbitrary executables
1382 - Fix failure of gnt-node list during instance removal
1383 - Ship the RAPI documentation in the archive
1389 *(Released Wed, 30 Sep 2009)*
1391 - Fixed many wrong messages
1392 - Fixed a few bugs related to the locking library
1393 - Fixed MAC checking at instance creation time
1394 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
1395 - Fixed a few issues related to signal handling in both daemons and
1397 - Fixed the example startup script provided
1398 - Fixed insserv dependencies in the example startup script (patch from
1400 - Fixed handling of drained nodes in the iallocator framework
1401 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
1403 - Fixed error related to invalid job IDs in job polling
1404 - Fixed job/opcode persistence on unclean master shutdown
1405 - Fixed handling of partial job processing after unclean master
1407 - Fixed error reporting from LUs, previously all errors were converted
1408 into execution errors
1409 - Fixed error reporting from burnin
1410 - Decreased significantly the memory usage of the job queue
1411 - Optimised slightly multi-job submission
1412 - Optimised slightly opcode loading
1413 - Backported the multi-job submit framework from the development
1414 branch; multi-instance start and stop should be faster
1415 - Added script to clean archived jobs after 21 days; this will reduce
1416 the size of the queue directory
1417 - Added some extra checks in disk size tracking
1418 - Added an example ethers hook script
1419 - Added a cluster parameter that prevents Ganeti from modifying of
1421 - Added more node information to RAPI responses
1422 - Added a ``gnt-job watch`` command that allows following the ouput of a
1424 - Added a bind-address option to ganeti-rapi
1425 - Added more checks to the configuration verify
1426 - Enhanced the burnin script such that some operations can be retried
1428 - Converted instance reinstall to multi-instance model
1434 *(Released Fri, 7 Aug 2009)*
1436 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
1437 to allow using the pre-2.0.2 behaviour in activation, if any existing
1438 instances have mismatched disk sizes in the configuration
1439 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
1440 any configuration mismatches for disk sizes
1441 - Added ``gnt-master cluste-failover --no-voting`` to allow master
1442 failover to work on two-node clusters
1443 - Fixed the ``--net`` option of ``gnt-backup import``, which was
1445 - Fixed detection of OS script errors in ``gnt-backup export``
1446 - Fixed exit code of ``gnt-backup export``
1452 *(Released Fri, 17 Jul 2009)*
1454 - Added experimental support for stripped logical volumes; this should
1455 enhance performance but comes with a higher complexity in the block
1456 device handling; stripping is only enabled when passing
1457 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
1458 affected even in the non-stripped mode
1459 - Improved resiliency against transient failures at the end of DRBD
1460 resyncs, and in general of DRBD resync checks
1461 - Fixed a couple of issues with exports and snapshot errors
1462 - Fixed a couple of issues in instance listing
1463 - Added display of the disk size in ``gnt-instance info``
1464 - Fixed checking for valid OSes in instance creation
1465 - Fixed handling of the "vcpus" parameter in instance listing and in
1466 general of invalid parameters
1467 - Fixed http server library, and thus RAPI, to handle invalid
1468 username/password combinations correctly; this means that now they
1469 report unauthorized for queries too, not only for modifications,
1470 allowing earlier detect of configuration problems
1471 - Added a new "role" node list field, equivalent to the master/master
1472 candidate/drained/offline flags combinations
1473 - Fixed cluster modify and changes of candidate pool size
1474 - Fixed cluster verify error messages for wrong files on regular nodes
1475 - Fixed a couple of issues with node demotion from master candidate role
1476 - Fixed node readd issues
1477 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
1478 - Added a new ``--no-voting`` option for masterfailover to fix failover
1479 on two-nodes clusters when the former master node is unreachable
1480 - Added instance reinstall over RAPI
1486 *(Released Tue, 16 Jun 2009)*
1488 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
1489 allow re-adding the start in single-user option (regression from 1.2)
1490 - the watcher writes the instance status to a file, to allow monitoring
1491 to report the instance status (from the master) based on cached
1492 results of the watcher's queries; while this can get stale if the
1493 watcher is being locked due to other work on the cluster, this is
1494 still an improvement
1495 - the watcher now also restarts the node daemon and the rapi daemon if
1497 - fixed the watcher to handle full and drained queue cases
1498 - hooks export more instance data in the environment, which helps if
1499 hook scripts need to take action based on the instance's properties
1500 (no longer need to query back into ganeti)
1501 - instance failovers when the instance is stopped do not check for free
1502 RAM, so that failing over a stopped instance is possible in low memory
1504 - rapi uses queries for tags instead of jobs (for less job traffic), and
1505 for cluster tags it won't talk to masterd at all but read them from
1507 - a couple of error handling fixes in RAPI
1508 - drbd handling: improved the error handling of inconsistent disks after
1509 resync to reduce the frequency of "there are some degraded disks for
1510 this instance" messages
1511 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
1512 error handling path called a wrong function name)
1518 *(Released Wed, 27 May 2009)*
1520 - no changes from rc5
1526 *(Released Wed, 20 May 2009)*
1528 - fix a couple of bugs (validation, argument checks)
1529 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
1530 - some small improvements to RAPI and IAllocator
1531 - make watcher automatically start the master daemon if down
1537 *(Released Mon, 27 Apr 2009)*
1539 - change the OS list to not require locks; this helps with big clusters
1540 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
1541 volume group is broken
1542 - ``gnt-instance info``, without any arguments, doesn't run for all
1543 instances anymore; either pass ``--all`` or pass the desired
1544 instances; this helps against mistakes on big clusters where listing
1545 the information for all instances takes a long time
1546 - miscellaneous doc and man pages fixes
1552 *(Released Wed, 8 Apr 2009)*
1554 - Change the internal locking model of some ``gnt-node`` commands, in
1555 order to reduce contention (and blocking of master daemon) when
1556 batching many creation/reinstall jobs
1557 - Fixes to Xen soft reboot
1558 - No longer build documentation at build time, instead distribute it in
1559 the archive, in order to reduce the need for the whole docbook/rst
1566 *(Released Fri, 27 Mar 2009)*
1568 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
1569 - Fix watcher startup sequence, improves the behaviour of busy clusters
1570 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
1571 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
1572 bind address and other places
1573 - Some documentation fixes and updates
1579 *(Released Mon, 2 Mar 2009)*
1581 - More documentation updates, now all docs should be more-or-less
1583 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
1585 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
1587 - Fix an issue related to $libdir/run/ganeti and cluster creation
1593 *(Released Thu, 19 Feb 2009)*
1595 - Xen PVM and KVM have switched the default value for the instance root
1596 disk to the first partition on the first drive, instead of the whole
1597 drive; this means that the OS installation scripts must be changed
1599 - Man pages have been updated
1600 - RAPI has been switched by default to HTTPS, and the exported functions
1601 should all work correctly
1602 - RAPI v1 has been removed
1603 - Many improvements to the KVM hypervisor
1604 - Block device errors are now better reported
1605 - Many other bugfixes and small improvements
1611 *(Released Mon, 26 Jan 2009)*
1613 - Version 2 is a general rewrite of the code and therefore the
1614 differences are too many to list, see the design document for 2.0 in
1615 the ``doc/`` subdirectory for more details
1616 - In this beta version there is not yet a migration path from 1.2 (there
1617 will be one in the final 2.0 release)
1618 - A few significant changes are:
1620 - all commands are executed by a daemon (``ganeti-masterd``) and the
1621 various ``gnt-*`` commands are just front-ends to it
1622 - all the commands are entered into, and executed from a job queue,
1623 see the ``gnt-job(8)`` manpage
1624 - the RAPI daemon supports read-write operations, secured by basic
1625 HTTP authentication on top of HTTPS
1626 - DRBD version 0.7 support has been removed, DRBD 8 is the only
1627 supported version (when migrating from Ganeti 1.2 to 2.0, you need
1628 to migrate to DRBD 8 first while still running Ganeti 1.2)
1629 - DRBD devices are using statically allocated minor numbers, which
1630 will be assigned to existing instances during the migration process
1631 - there is support for both Xen PVM and Xen HVM instances running on
1633 - KVM virtualization is supported too
1634 - file-based storage has been implemented, which means that it is
1635 possible to run the cluster without LVM and DRBD storage, for
1636 example using a shared filesystem exported from shared storage (and
1637 still have live migration)
1643 *(Released Tue, 13 Jan 2009)*
1645 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
1646 - Reuse the old instance mac address by default on instance import, if
1647 the instance name is the same.
1648 - Handle situations in which the node info rpc returns incomplete
1650 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
1651 - Improved version of batcher:
1653 - state file support
1654 - instance mac address support
1655 - support for HVM clusters/instances
1657 - Add an option to show the number of cpu sockets and nodes in
1659 - Support OSes that handle more than one version of the OS api (but do
1660 not change the current API in any other way)
1661 - Fix ``gnt-node migrate``
1662 - ``gnt-debug`` man page
1663 - Fixes various more typos and small issues
1664 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
1670 *(Released Wed, 24 Sep 2008)*
1672 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
1673 type of disk exported to fully virtualized instances.
1674 - provide access to the serial console of HVM instances
1675 - instance auto_balance flag, set by default. If turned off it will
1676 avoid warnings on cluster verify if there is not enough memory to fail
1677 over an instance. in the future it will prevent automatically failing
1678 it over when we will support that.
1679 - batcher tool for instance creation, see ``tools/README.batcher``
1680 - ``gnt-instance reinstall --select-os`` to interactively select a new
1681 operating system when reinstalling an instance.
1682 - when changing the memory amount on instance modify a check has been
1683 added that the instance will be able to start. also warnings are
1684 emitted if the instance will not be able to fail over, if auto_balance
1686 - documentation fixes
1687 - sync fields between ``gnt-instance list/modify/add/import``
1688 - fix a race condition in drbd when the sync speed was set after giving
1689 the device a remote peer.
1695 *(Released Tue, 22 Jul 2008)*
1697 - note: the allowed size and number of tags per object were reduced
1698 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
1699 - fixed twisted 8.x compatibility
1700 - fixed ``gnt-instance replace-disks`` with iallocator
1701 - add TCP keepalives on twisted connections to detect restarted nodes
1702 - disk increase support, see ``gnt-instance grow-disk``
1703 - implement bulk node/instance query for RAPI
1704 - add tags in node/instance listing (optional)
1705 - experimental migration (and live migration) support, read the man page
1706 for ``gnt-instance migrate``
1707 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
1708 has some small improvements in handling its state file
1714 *(Released Fri, 13 Jun 2008)*
1716 - Experimental readonly, REST-based remote API implementation;
1717 automatically started on master node, TCP port 5080, if enabled by
1718 ``--enable-rapi`` parameter to configure script.
1719 - Instance allocator support. Add and import instance accept a
1720 ``--iallocator`` parameter, and call that instance allocator to decide
1721 which node to use for the instance. The iallocator document describes
1722 what's expected from an allocator script.
1723 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
1724 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
1725 node is lost there is still enough memory to fail over the instances
1727 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
1728 ``gnt-cluster verify``, to check for site-specific compliance. All the
1729 hooks will run, and their output, if any, will be displayed. Any
1730 failing hook will make the verification return an error value.
1731 - ``gnt-cluster verify`` now checks that its peers are reachable on the
1732 primary and secondary interfaces
1733 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
1734 that is still declared as part of the cluster and has failed.
1735 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
1736 specifying output fields, that just adds the chosen fields to the
1738 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
1739 export from the filesystem.
1740 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
1741 have been added. Using them you can enable/disable acpi and pae
1742 support, and specify a path for a cd image to be exported to the
1743 instance. These parameters as the name suggest only work on HVM
1745 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
1746 PAE support will be set to the previously hardcoded values, but the
1747 (previously hardcoded) path to the CDROM ISO image will be unset and
1748 if required, needs to be set manually with ``gnt-instance modify``
1750 - The address to which an instance's VNC console is bound is now
1751 selectable per-instance, rather than being cluster wide. Of course
1752 this only applies to instances controlled via VNC, so currently just
1753 applies to HVM clusters.
1759 *(Released Mon, 18 Feb 2008)*
1761 - more tweaks to the disk activation code (especially helpful for DRBD)
1762 - change the default ``gnt-instance list`` output format, now there is
1763 one combined status field (see the manpage for the exact values this
1765 - some more fixes for the mac export to hooks change
1766 - make Ganeti not break with DRBD 8.2.x (which changed the version
1767 format in ``/proc/drbd``) (issue 24)
1768 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
1769 template, allowing migration from DRBD0.7+MD to DRBD8
1775 *(Released Wed, 30 Jan 2008)*
1777 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
1779 - add command aliases infrastructure and a few aliases
1780 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
1781 man pages and the ``--help`` option of ``gnt-node
1782 list``/``gnt-instance list``
1783 - fix ``gnt-backup list`` with down nodes (issue 21)
1784 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
1785 - fix the dist archive and add a check for including svn/git files in
1787 - some developer-related changes: improve the burnin and the QA suite,
1788 add an upload script for testing during development
1794 *(Released Wed, 16 Jan 2008)*
1796 - experimental HVM support, read the install document, section
1797 "Initializing the cluster"
1798 - allow for the PVM hypervisor per-instance kernel and initrd paths
1799 - add a new command ``gnt-cluster verify-disks`` which uses a new
1800 algorithm to improve the reconnection of the DRBD pairs if the device
1801 on the secondary node has gone away
1802 - make logical volume code auto-activate LVs at disk activation time
1803 - slightly improve the speed of activating disks
1804 - allow specification of the MAC address at instance creation time, and
1805 changing it later via ``gnt-instance modify``
1806 - fix handling of external commands that generate lots of output on
1808 - update documentation with regard to minimum version of DRBD8 supported
1814 *(Released Tue, 4 Dec 2007)*
1816 - Log the ``xm create`` output to the node daemon log on failure (to
1817 help diagnosing the error)
1818 - In debug mode, log all external commands output if failed to the logs
1819 - Change parsing of lvm commands to ignore stderr
1825 *(Released Wed, 28 Nov 2007)*
1827 - Another round of updates to the DRBD 8 code to deal with more failures
1828 in the replace secondary node operation
1829 - Some more logging of failures in disk operations (lvm, drbd)
1830 - A few documentation updates
1837 *(Released Tue, 13 Nov 2007)*
1839 - Change configuration file format from Python's Pickle to JSON.
1840 Upgrading is possible using the cfgupgrade utility.
1841 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
1842 faster replace disks and is more stable (DRBD 8 has many improvements
1843 compared to DRBD 0.7)
1844 - Added command line tags support (see man pages for ``gnt-instance``,
1845 ``gnt-node``, ``gnt-cluster``)
1846 - Added instance rename support
1847 - Added multi-instance startup/shutdown
1848 - Added cluster rename support
1849 - Added ``gnt-node evacuate`` to simplify some node operations
1850 - Added instance reboot operation that can speedup reboot as compared to
1852 - Soften the requirement that hostnames are in FQDN format
1853 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
1855 - Removed dependency on debian's patched fping that uses the
1856 non-standard ``-S`` option
1857 - Now the OS definitions are searched for in multiple, configurable
1858 paths (easier for distros to package)
1859 - Some changes to the hooks infrastructure (especially the new
1860 post-configuration update hook)
1861 - Other small bugfixes
1863 .. vim: set textwidth=72 syntax=rst :