8 *(Released Tue, 03 Jul 2012)*
16 The current ``admin_up`` field, which used to denote whether an instance
17 should be running or not, has been removed. Instead, ``admin_state`` is
18 introduced, with 3 possible values -- ``up``, ``down`` and ``offline``.
20 The rational behind this is that an instance being “down” can have
23 - it could be down during a reboot
24 - it could be temporarily be down for a reinstall
25 - or it could be down because it is deprecated and kept just for its
28 The previous Boolean state was making it difficult to do capacity
29 calculations: should Ganeti reserve memory for a down instance? Now, the
30 tri-state field makes it clear:
32 - in ``up`` and ``down`` state, all resources are reserved for the
33 instance, and it can be at any time brought up if it is down
34 - in ``offline`` state, only disk space is reserved for it, but not
37 The field can have an extra use: since the transition between ``up`` and
38 ``down`` and vice-versus is done via ``gnt-instance start/stop``, but
39 transition between ``offline`` and ``down`` is done via ``gnt-instance
40 modify``, it is possible to given different rights to users. For
41 example, owners of an instance could be allowed to start/stop it, but
42 not transition it out of the offline state.
44 Instance policies and specs
45 +++++++++++++++++++++++++++
47 In previous Ganeti versions, an instance creation request was not
48 limited on the minimum size and on the maximum size just by the cluster
49 resources. As such, any policy could be implemented only in third-party
50 clients (RAPI clients, or shell wrappers over ``gnt-*``
51 tools). Furthermore, calculating cluster capacity via ``hspace`` again
52 required external input with regards to instance sizes.
54 In order to improve these workflows and to allow for example better
55 per-node group differentiation, we introduced instance specs, which
58 - minimum instance disk size, disk count, memory size, cpu count
59 - maximum values for the above metrics
60 - and “standard” values (used in ``hspace`` to calculate the standard
63 The minimum/maximum values can be also customised at node-group level,
64 for example allowing more powerful hardware to support bigger instance
67 Beside the instance specs, there are a few other settings belonging to
68 the instance policy framework. It is possible now to customise, per
69 cluster and node-group:
71 - the list of allowed disk templates
72 - the maximum ratio of VCPUs per PCPUs (to control CPU oversubscription)
73 - the maximum ratio of instance to spindles (see below for more
74 information) for local storage
76 All these together should allow all tools that talk to Ganeti to know
77 what are the ranges of allowed values for instances and the
78 over-subscription that is allowed.
80 For the VCPU/PCPU ratio, we already have the VCPU configuration from the
81 instance configuration, and the physical CPU configuration from the
82 node. For the spindle ratios however, we didn't track before these
83 values, so new parameters have been added:
85 - a new node parameter ``spindle_count``, defaults to 1, customisable at
86 node group or node level
87 - at new backend parameter (for instances), ``spindle_use`` defaults to 1
89 Note that spindles in this context doesn't need to mean actual
90 mechanical hard-drives; it's just a relative number for both the node
91 I/O capacity and instance I/O consumption.
93 Instance migration behaviour
94 ++++++++++++++++++++++++++++
96 While live-migration is in general desirable over failover, it is
97 possible that for some workloads it is actually worse, due to the
98 variable time of the “suspend” phase during live migration.
100 To allow the tools to work consistently over such instances (without
101 having to hard-code instance names), a new backend parameter
102 ``always_failover`` has been added to control the migration/failover
103 behaviour. When set to True, all migration requests for an instance will
104 instead fall-back to failover.
106 Instance memory ballooning
107 ++++++++++++++++++++++++++
109 Initial support for memory ballooning has been added. The memory for an
110 instance is no longer fixed (backend parameter ``memory``), but instead
111 can vary between minimum and maximum values (backend parameters
112 ``minmem`` and ``maxmem``). Currently we only change an instance's
115 - live migrating or failing over and instance and the target node
116 doesn't have enough memory
117 - user requests changing the memory via ``gnt-instance modify
123 In order to control the use of specific CPUs by instance, support for
124 controlling CPU pinning has been added for the Xen, HVM and LXC
125 hypervisors. This is controlled by a new hypervisor parameter
126 ``cpu_mask``; details about possible values for this are in the
127 :manpage:`gnt-instance(8)`. Note that use of the most specific (precise
128 VCPU-to-CPU mapping) form will work well only when all nodes in your
129 cluster have the same amount of CPUs.
134 Another area in which Ganeti was not customisable were the parameters
135 used for storage configuration, e.g. how many stripes to use for LVM,
136 DRBD resync configuration, etc.
138 To improve this area, we've added disks parameters, which are
139 customisable at cluster and node group level, and which allow to
140 specify various parameters for disks (DRBD has the most parameters
141 currently), for example:
143 - DRBD resync algorithm and parameters (e.g. speed)
144 - the default VG for meta-data volumes for DRBD
145 - number of stripes for LVM (plain disk template)
148 These parameters can be modified via ``gnt-cluster modify -D …`` and
149 ``gnt-group modify -D …``, and are used at either instance creation (in
150 case of LVM stripes, for example) or at disk “activation” time
153 Rados block device support
154 ++++++++++++++++++++++++++
156 A Rados (http://ceph.com/wiki/Rbd) storage backend has been added,
157 denoted by the ``rbd`` disk template type. This is considered
158 experimental, feedback is welcome. For details on configuring it, see
159 the :doc:`install` document and the :manpage:`gnt-cluster(8)` man page.
164 The existing master IP functionality works well only in simple setups (a
165 single network shared by all nodes); however, if nodes belong to
166 different networks, then the ``/32`` setup and lack of routing
167 information is not enough.
169 To allow the master IP to function well in more complex cases, the
170 system was reworked as follows:
172 - a master IP netmask setting has been added
173 - the master IP activation/turn-down code was moved from the node daemon
175 - whether to run the Ganeti-supplied master IP script or a user-supplied
176 on is a ``gnt-cluster init`` setting
178 Details about the location of the standard and custom setup scripts are
179 in the man page :manpage:`gnt-cluster(8)`; for information about the
180 setup script protocol, look at the Ganeti-supplied script.
185 The `SPICE <http://www.linux-kvm.org/page/SPICE>`_ support has been
188 It is now possible to use TLS-protected connections, and when renewing
189 or changing the cluster certificates (via ``gnt-cluster renew-crypto``,
190 it is now possible to specify spice or spice CA certificates. Also, it
191 is possible to configure a password for SPICE sessions via the
192 hypervisor parameter ``spice_password_file``.
194 There are also new parameters to control the compression and streaming
195 options (e.g. ``spice_image_compression``, ``spice_streaming_video``,
196 etc.). For details, see the man page :manpage:`gnt-instance(8)` and look
197 for the spice parameters.
199 Lastly, it is now possible to see the SPICE connection information via
200 ``gnt-instance console``.
205 A new tool (``tools/ovfconverter``) has been added that supports
206 conversion between Ganeti and the `Open Virtualization Format
207 <http://en.wikipedia.org/wiki/Open_Virtualization_Format>`_ (both to and
210 This relies on the ``qemu-img`` tool to convert the disk formats, so the
211 actual compatibility with other virtualization solutions depends on it.
216 The configuration query daemon (``ganeti-confd``) is now optional, and
217 has been rewritten in Haskell; whether to use the daemon at all, use the
218 Python (default) or the Haskell version is selectable at configure time
219 via the ``--enable-confd`` parameter, which can take one of the
220 ``haskell``, ``python`` or ``no`` values. If not used, disabling the
221 daemon will result in a smaller footprint; for larger systems, we
222 welcome feedback on the Haskell version which might become the default
226 User interface changes
227 ~~~~~~~~~~~~~~~~~~~~~~
229 We have replaced the ``--disks`` option of ``gnt-instance
230 replace-disks`` with a more flexible ``--disk`` option, which allows
231 adding and removing disks at arbitrary indices (Issue 188). Furthermore,
232 disk size and mode can be changed upon recreation (via ``gnt-instance
233 recreate-disks``, which accepts the same ``--disk`` option).
235 As many people are used to a ``show`` command, we have added that as an
236 alias to ``info`` on all ``gnt-*`` commands.
238 The ``gnt-instance grow-disk`` command has a new mode in which it can
239 accept the target size of the disk, instead of the delta; this can be
240 more safe since two runs in absolute mode will be idempotent, and
241 sometimes it's also easier to specify the desired size directly.
246 RAPI coverage has improved, with (for example) new resources for
247 recreate-disks, node power-cycle, etc.
252 There is partial support for ``xl`` in the Xen hypervisor; feedback is
255 Python 2.7 is better supported, and after Ganeti 2.6 we will investigate
256 whether to still support Python 2.4 or move to Python 2.6 as minimum
259 Support for Fedora has been slightly improved; the provided example
260 init.d script should work better on it and the INSTALL file should
261 document the needed dependencies.
266 The deprecated ``QueryLocks`` LUXI request has been removed. Use
267 ``Query(what=QR_LOCK, ...)`` instead.
269 The LUXI requests :pyeval:`luxi.REQ_QUERY_JOBS`,
270 :pyeval:`luxi.REQ_QUERY_INSTANCES`, :pyeval:`luxi.REQ_QUERY_NODES`,
271 :pyeval:`luxi.REQ_QUERY_GROUPS`, :pyeval:`luxi.REQ_QUERY_EXPORTS` and
272 :pyeval:`luxi.REQ_QUERY_TAGS` are deprecated and will be removed in a
273 future version. :pyeval:`luxi.REQ_QUERY` should be used instead.
275 RAPI client: ``CertificateError`` now derives from
276 ``GanetiApiError``. This should make it more easy to handle Ganeti
279 Deprecation warnings due to PyCrypto/paramiko import in
280 ``tools/setup-ssh`` have been silenced, as usually they are safe; please
281 make sure to run an up-to-date paramiko version, if you use this tool.
283 The QA scripts now depend on Python 2.5 or above (the main code base
284 still works with Python 2.4).
286 The configuration file (``config.data``) is now written without
287 indentation for performance reasons; if you want to edit it, it can be
288 re-formatted via ``tools/fmtjson``.
290 A number of bugs has been fixed in the cluster merge tool.
292 ``x509`` certification verification (used in import-export) has been
293 changed to allow the same clock skew as permitted by the cluster
294 verification. This will remove some rare but hard to diagnose errors in
301 *(Released Mon, 25 Jun 2012)*
303 First release candidate for 2.6. The following changes were done from
306 - Fixed bugs with disk parameters and ``rbd`` templates as well as
308 - Made ``gnt-instance modify`` more consistent regarding new NIC/Disk
309 behaviour. It supports now the modify operation
310 - ``hcheck`` implemented to analyze cluster health and possibility of
311 improving health by rebalance
312 - ``hbal`` has been improved in dealing with split instances
318 *(Released Mon, 11 Jun 2012)*
320 Second beta release of 2.6. The following changes were done from beta2
323 - Fixed ``daemon-util`` with non-root user models
324 - Fixed creation of plain instances with ``--no-wait-for-sync``
325 - Fix wrong iv_names when running ``cfgupgrade``
326 - Export more information in RAPI group queries
327 - Fixed bug when changing instance network interfaces
328 - Extended burnin to do NIC changes
329 - query: Added ``<``, ``>``, ``<=``, ``>=`` comparison operators
330 - Changed default for DRBD barriers
331 - Fixed DRBD error reporting for syncer rate
332 - Verify the options on disk parameters
334 And of course various fixes to documentation and improved unittests and
341 *(Released Wed, 23 May 2012)*
343 First beta release of 2.6. The following changes were done from beta1 to
346 - integrated patch for distributions without ``start-stop-daemon``
347 - adapted example init.d script to work on Fedora
348 - fixed log handling in Haskell daemons
349 - adapted checks in the watcher for pycurl linked against libnss
350 - add partial support for ``xl`` instead of ``xm`` for Xen
351 - fixed a type issue in cluster verification
352 - fixed ssconf handling in the Haskell code (was breaking confd in IPv6
355 Plus integrated fixes from the 2.5 branch:
357 - fixed ``kvm-ifup`` to use ``/bin/bash``
358 - fixed parallel build failures
359 - KVM live migration when using a custom keymap
365 *(Released Fri, 11 May 2012)*
367 A small bugfix release.
369 The main issues solved are on the topic of compatibility with newer LVM
372 - fixed parsing of ``lv_attr`` field
373 - adapted to new ``vgreduce --removemissing`` behaviour where sometimes
374 the ``--force`` flag is needed
376 Also on the topic of compatibility, ``tools/lvmstrap`` has been changed
377 to accept kernel 3.x too (was hardcoded to 2.6.*).
379 A regression present in 2.5.0 that broke handling (in the gnt-* scripts)
380 of hook results and that also made display of other errors suboptimal
381 was fixed; the code behaves now like 2.4 and earlier.
383 Another change in 2.5, the cleanup of the OS scripts environment, is too
384 aggressive: it removed even the ``PATH`` variable, which requires the OS
385 scripts to *always* need to export it. Since this is a bit too strict,
386 we now export a minimal PATH, the same that we export for hooks.
388 The fix for issue 201 (Preserve bridge MTU in KVM ifup script) was
389 integrated into this release.
391 Finally, a few other miscellaneous changes were done (no new features,
392 just small improvements):
394 - Fix ``gnt-group --help`` display
395 - Fix hardcoded Xen kernel path
396 - Fix grow-disk handling of invalid units
397 - Update synopsis for ``gnt-cluster repair-disk-sizes``
398 - Accept both PUT and POST in noded (makes future upgrade to 2.6 easier)
404 *(Released Thu, 12 Apr 2012)*
406 Incompatible/important changes and bugfixes
407 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
409 - The default of the ``/2/instances/[instance_name]/rename`` RAPI
410 resource's ``ip_check`` parameter changed from ``True`` to ``False``
411 to match the underlying LUXI interface.
412 - The ``/2/nodes/[node_name]/evacuate`` RAPI resource was changed to use
413 body parameters, see :doc:`RAPI documentation <rapi>`. The server does
414 not maintain backwards-compatibility as the underlying operation
415 changed in an incompatible way. The RAPI client can talk to old
416 servers, but it needs to be told so as the return value changed.
417 - When creating file-based instances via RAPI, the ``file_driver``
418 parameter no longer defaults to ``loop`` and must be specified.
419 - The deprecated ``bridge`` NIC parameter is no longer supported. Use
421 - Support for the undocumented and deprecated RAPI instance creation
422 request format version 0 has been dropped. Use version 1, supported
423 since Ganeti 2.1.3 and :doc:`documented <rapi>`, instead.
424 - Pyparsing 1.4.6 or above is required, see :doc:`installation
425 documentation <install>`.
426 - The "cluster-verify" hooks are now executed per group by the
427 ``OP_CLUSTER_VERIFY_GROUP`` opcode. This maintains the same behavior
428 if you just run ``gnt-cluster verify``, which generates one opcode per
430 - The environment as passed to the OS scripts is cleared, and thus no
431 environment variables defined in the node daemon's environment will be
432 inherited by the scripts.
433 - The :doc:`iallocator <iallocator>` mode ``multi-evacuate`` has been
435 - :doc:`New iallocator modes <design-multi-reloc>` have been added to
436 support operations involving multiple node groups.
437 - Offline nodes are ignored when failing over an instance.
438 - Support for KVM version 1.0, which changed the version reporting format
440 - TCP/IP ports used by DRBD disks are returned to a pool upon instance
442 - ``Makefile`` is now compatible with Automake 1.11.2
443 - Includes all bugfixes made in the 2.4 series
448 - The ganeti-htools project has been merged into the ganeti-core source
449 tree and will be built as part of Ganeti (see :doc:`install-quick`).
450 - Implemented support for :doc:`shared storage <design-shared-storage>`.
451 - Add support for disks larger than 2 TB in ``lvmstrap`` by supporting
452 GPT-style partition tables (requires `parted
453 <http://www.gnu.org/s/parted/>`_).
454 - Added support for floppy drive and 2nd CD-ROM drive in KVM hypervisor.
455 - Allowed adding tags on instance creation.
456 - Export instance tags to hooks (``INSTANCE_TAGS``, see :doc:`hooks`)
457 - Allow instances to be started in a paused state, enabling the user to
458 see the complete console output on boot using the console.
459 - Added new hypervisor flag to control default reboot behaviour
460 (``reboot_behavior``).
461 - Added support for KVM keymaps (hypervisor parameter ``keymap``).
462 - Improved out-of-band management support:
464 - Added ``gnt-node health`` command reporting the health status of
466 - Added ``gnt-node power`` command to manage power status of nodes.
467 - Added command for emergency power-off (EPO), ``gnt-cluster epo``.
469 - Instance migration can fall back to failover if instance is not
471 - Filters can be used when listing nodes, instances, groups and locks;
472 see *ganeti(7)* manpage.
473 - Added post-execution status as variables to :doc:`hooks <hooks>`
475 - Instance tags are exported/imported together with the instance.
476 - When given an explicit job ID, ``gnt-job info`` will work for archived
478 - Jobs can define dependencies on other jobs (not yet supported via
479 RAPI or command line, but used by internal commands and usable via
482 - Lock monitor (``gnt-debug locks``) shows jobs waiting for
485 - Instance failover is now available as a RAPI resource
486 (``/2/instances/[instance_name]/failover``).
487 - ``gnt-instance info`` defaults to static information if primary node
489 - Opcodes have a new ``comment`` attribute.
490 - Added basic SPICE support to KVM hypervisor.
491 - ``tools/ganeti-listrunner`` allows passing of arguments to executable.
493 Node group improvements
494 ~~~~~~~~~~~~~~~~~~~~~~~
496 - ``gnt-cluster verify`` has been modified to check groups separately,
497 thereby improving performance.
498 - Node group support has been added to ``gnt-cluster verify-disks``,
499 which now operates per node group.
500 - Watcher has been changed to work better with node groups.
502 - One process and state file per node group.
503 - Slow watcher in one group doesn't block other group's watcher.
505 - Added new command, ``gnt-group evacuate``, to move all instances in a
506 node group to other groups.
507 - Added ``gnt-instance change-group`` to move an instance to another
509 - ``gnt-cluster command`` and ``gnt-cluster copyfile`` now support
510 per-group operations.
511 - Node groups can be tagged.
512 - Some operations switch from an exclusive to a shared lock as soon as
514 - Instance's primary and secondary nodes' groups are now available as
515 query fields (``pnode.group``, ``pnode.group.uuid``, ``snodes.group``
516 and ``snodes.group.uuid``).
521 - Numerous updates to documentation and manpages.
523 - :doc:`RAPI <rapi>` documentation now has detailed parameter
525 - Some opcode/job results are now also documented, see :doc:`RAPI
528 - A lockset's internal lock is now also visible in lock monitor.
529 - Log messages from job queue workers now contain information about the
530 opcode they're processing.
531 - ``gnt-instance console`` no longer requires the instance lock.
532 - A short delay when waiting for job changes reduces the number of LUXI
533 requests significantly.
534 - DRBD metadata volumes are overwritten with zeros during disk creation.
535 - Out-of-band commands no longer acquire the cluster lock in exclusive
537 - ``devel/upload`` now uses correct permissions for directories.
543 *(Released Fri, 23 Mar 2012)*
545 This was the sixth release candidate of the 2.5 series.
551 *(Released Mon, 9 Jan 2012)*
553 This was the fifth release candidate of the 2.5 series.
559 *(Released Thu, 27 Oct 2011)*
561 This was the fourth release candidate of the 2.5 series.
567 *(Released Wed, 26 Oct 2011)*
569 This was the third release candidate of the 2.5 series.
575 *(Released Tue, 18 Oct 2011)*
577 This was the second release candidate of the 2.5 series.
583 *(Released Tue, 4 Oct 2011)*
585 This was the first release candidate of the 2.5 series.
591 *(Released Wed, 31 Aug 2011)*
593 This was the third beta release of the 2.5 series.
599 *(Released Mon, 22 Aug 2011)*
601 This was the second beta release of the 2.5 series.
607 *(Released Fri, 12 Aug 2011)*
609 This was the first beta release of the 2.5 series.
615 *(Released Thu, 27 Oct 2011)*
617 - Fixed bug when parsing command line parameter values ending in
619 - Fixed assertion error after unclean master shutdown
620 - Disable HTTP client pool for RPC, significantly reducing memory usage
622 - Fixed queue archive creation with wrong permissions
628 *(Released Tue, 23 Aug 2011)*
632 - Fixed documentation for importing with ``--src-dir`` option
633 - Fixed a bug in ``ensure-dirs`` with queue/archive permissions
634 - Fixed a parsing issue with DRBD 8.3.11 in the Linux kernel
640 *(Released Fri, 5 Aug 2011)*
642 Many bug-fixes and a few small features:
644 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
645 issues when you tried to add an instance with two MAC addresses in one
647 - KVM: fixed per-instance stored UID value
648 - KVM: configure bridged NICs at migration start
649 - KVM: Fix a bug where instance will not start with never KVM versions
651 - Added OS search path to ``gnt-cluster info``
652 - Fixed an issue with ``file_storage_dir`` where you were forced to
653 provide an absolute path, but the documentation states it is a
654 relative path, the documentation was right
655 - Added a new parameter to instance stop/start called ``--no-remember``
656 that will make the state change to not be remembered
657 - Implemented ``no_remember`` at RAPI level
658 - Improved the documentation
659 - Node evacuation: don't call IAllocator if node is already empty
660 - Fixed bug in DRBD8 replace disks on current nodes
661 - Fixed bug in recreate-disks for DRBD instances
662 - Moved assertion checking locks in ``gnt-instance replace-disks``
663 causing it to abort with not owning the right locks for some situation
664 - Job queue: Fixed potential race condition when cancelling queued jobs
665 - Fixed off-by-one bug in job serial generation
666 - ``gnt-node volumes``: Fix instance names
667 - Fixed aliases in bash completion
668 - Fixed a bug in reopening log files after being sent a SIGHUP
669 - Added a flag to burnin to allow specifying VCPU count
670 - Bugfixes to non-root Ganeti configuration
676 *(Released Thu, 12 May 2011)*
678 Many bug-fixes and a few new small features:
680 - Fixed a bug related to log opening failures
681 - Fixed a bug in instance listing with orphan instances
682 - Fixed a bug which prevented resetting the cluster-level node parameter
683 ``oob_program`` to the default
684 - Many fixes related to the ``cluster-merge`` tool
685 - Fixed a race condition in the lock monitor, which caused failures
686 during (at least) creation of many instances in parallel
687 - Improved output for gnt-job info
688 - Removed the quiet flag on some ssh calls which prevented debugging
690 - Improved the N+1 failure messages in cluster verify by actually
691 showing the memory values (needed and available)
692 - Increased lock attempt timeouts so that when executing long operations
693 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
694 too early and thus prevent the use of the 'fair' mechanism
695 - Changed instance query data (``gnt-instance info``) to not acquire
696 locks unless needed, thus allowing its use on locked instance if only
697 static information is asked for
698 - Improved behaviour with filesystems that do not support rename on an
700 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
701 kept locks on all nodes during the wipe, which is unneeded
702 - Fixed ``gnt-watcher`` handling of errors during hooks execution
703 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
704 10GiB) which caused the wipe to fail right at the end in some cases
705 - Fixed master IP activation when doing master failover with no-voting
706 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
707 the master node itself
708 - Fixed potential data-loss in under disk full conditions, where Ganeti
709 wouldn't check correctly the return code and would consider
710 partially-written files 'correct'
711 - Fixed bug related to multiple VGs and DRBD disk replacing
712 - Added new disk parameter ``metavg`` that allows placement of the meta
713 device for DRBD in a different volume group
714 - Fixed error handling in the node daemon when the system libc doesn't
715 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
716 - Fixed lock release during replace-disks, which kept cluster-wide locks
717 when doing disk replaces with an iallocator script
718 - Added check for missing bridges in cluster verify
719 - Handle EPIPE errors while writing to the terminal better, so that
720 piping the output to e.g. ``less`` doesn't cause a backtrace
721 - Fixed rare case where a ^C during Luxi calls could have been
722 interpreted as server errors, instead of simply terminating
723 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
725 - Added a few more parameters to the KVM hypervisor, allowing a second
726 CDROM, custom disk type for CDROMs and a floppy image
727 - Removed redundant message in instance rename when the name is given
729 - Added option to ``gnt-instance recreate-disks`` to allow creating the
730 disks on new nodes, allowing recreation when the original instance
731 nodes are completely gone
732 - Added option when converting disk templates to DRBD to skip waiting
733 for the resync, in order to make the instance available sooner
734 - Added two new variables to the OS scripts environment (containing the
736 - Made the root_path and optional parameter for the xen-pvm hypervisor,
737 to allow use of ``pvgrub`` as bootloader
738 - Changed the instance memory modifications to only check out-of-memory
739 conditions on memory increases, and turned the secondary node warnings
740 into errors (they can still be overridden via ``--force``)
741 - Fixed the handling of a corner case when the Python installation gets
742 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
743 to execute a command that doesn't exist
744 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
745 node of the instance returned failures during instance shutdown; this
746 adds the option ``--ignore-consistency`` to gnt-instance move
748 And as usual, various improvements to the error messages, documentation
755 *(Released Wed, 09 Mar 2011)*
757 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
758 the RAPI users file if run twice (even with ``--dry-run``).
760 The release fixes that bug (nothing else changed).
766 *(Released Mon, 07 Mar 2011)*
768 Final 2.4.0 release. Just a few small fixes:
770 - Fixed RAPI node evacuate
771 - Fixed the kvm-ifup script
772 - Fixed internal error handling for special job cases
773 - Updated man page to specify the escaping feature for options
779 *(Released Mon, 28 Feb 2011)*
781 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
782 that this feature wiped the disks of the wrong instance, leading to loss
787 - Fixed title of query field containing instance name
788 - Expanded the glossary in the documentation
789 - Fixed one unittest (internal issue)
795 *(Released Mon, 21 Feb 2011)*
797 A number of bug fixes plus just a couple functionality changes.
799 On the user-visible side, the ``gnt-* list`` command output has changed
800 with respect to "special" field states. The current rc1 style of display
801 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
802 the default output mode special fields states are displayed as follows:
804 - Offline resource: ``*``
805 - Unavailable/not applicable: ``-``
806 - Data missing (RPC failure): ``?``
807 - Unknown field: ``??``
809 Another user-visible change is the addition of ``--force-join`` to
814 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
815 - Fixed regression in RAPI/instance reinstall where all parameters were
816 required (instead of optional)
817 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
818 - Fixed iallocator usage (offline nodes were not considered offline)
819 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
820 - Fixed hypervisor and OS parameter validation with respect to
822 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
824 - Fixed ``tools/listrunner`` with respect to agent-based usage
830 *(Released Fri, 4 Feb 2011)*
832 Many changes and fixes since the beta1 release. While there were some
833 internal changes, the code has been mostly stabilised for the RC
836 Note: the dumb allocator was removed in this release, as it was not kept
837 up-to-date with the IAllocator protocol changes. It is recommended to
838 use the ``hail`` command from the ganeti-htools package.
840 Note: the 2.4 and up versions of Ganeti are not compatible with the
841 0.2.x branch of ganeti-htools. You need to upgrade to
842 ganeti-htools-0.3.0 (or later).
844 Regressions fixed from 2.3
845 ~~~~~~~~~~~~~~~~~~~~~~~~~~
847 - Fixed the ``gnt-cluster verify-disks`` command
848 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
850 - Fixed disk adoption breakage
851 - Fixed wrong headers in instance listing for field aliases
856 - Fixed corner case in KVM handling of NICs
857 - Fixed many cases of wrong handling of non-vm_capable nodes
858 - Fixed a bug where a missing instance symlink was not possible to
859 recreate with any ``gnt-*`` command (now ``gnt-instance
860 activate-disks`` does it)
861 - Fixed the volume group name as reported by ``gnt-cluster
863 - Increased timeouts for the import-export code, hopefully leading to
864 fewer aborts due network or instance timeouts
865 - Fixed bug in ``gnt-node list-storage``
866 - Fixed bug where not all daemons were started on cluster
867 initialisation, but only at the first watcher run
868 - Fixed many bugs in the OOB implementation
869 - Fixed watcher behaviour in presence of instances with offline
871 - Fixed instance list output for instances running on the wrong node
872 - a few fixes to the cluster-merge tool, but it still cannot merge
873 multi-node groups (currently it is not recommended to use this tool)
879 - Improved network configuration for the KVM hypervisor
880 - Added e1000 as a supported NIC for Xen-HVM
881 - Improved the lvmstrap tool to also be able to use partitions, as
882 opposed to full disks
883 - Improved speed of disk wiping (the cluster parameter
884 ``prealloc_wipe_disks``, so that it has a low impact on the total time
885 of instance creations
886 - Added documentation for the OS parameters
887 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
888 hypervisor is not responding
889 - Added display of blacklisted and hidden OS information in
891 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
892 NIC and node parameters, which might create problems with currently
893 invalid (but undetected) configuration files, but prevents validation
894 failures when unrelated parameters are modified
895 - Changed cluster initialisation to wait for the master daemon to become
897 - Expanded the RAPI interface:
899 - Added config redistribution resource
900 - Added activation/deactivation of instance disks
901 - Added export of console information
903 - Implemented log file reopening on SIGHUP, which allows using
904 logrotate(8) for the Ganeti log files
905 - Added a basic OOB helper script as an example
911 *(Released Fri, 14 Jan 2011)*
916 - Fixed timezone issues when formatting timestamps
917 - Added support for node groups, available via ``gnt-group`` and other
919 - Added out-of-band framework and management, see :doc:`design
920 document <design-oob>`
921 - Removed support for roman numbers from ``gnt-node list`` and
922 ``gnt-instance list``.
923 - Allowed modification of master network interface via ``gnt-cluster
924 modify --master-netdev``
925 - Accept offline secondaries while shutting down instance disks
926 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
927 - Added support for multiple LVM volume groups
928 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
930 - Added commands to list available fields:
932 - ``gnt-node list-fields``
933 - ``gnt-group list-fields``
934 - ``gnt-instance list-fields``
936 - Updated documentation and man pages
941 - Moved ``rapi_users`` file into separate directory, now named
942 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
944 - Added new tool for running commands on many machines,
945 ``tools/ganeti-listrunner``
946 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
947 improving the ``gnt-instance console`` output
948 - Allowed customisation of disk index separator at ``configure`` time
949 - Export node group allocation policy to :doc:`iallocator <iallocator>`
950 - Added support for non-partitioned md disks in ``lvmstrap``
951 - Added script to gracefully power off KVM instances
952 - Split ``utils`` module into smaller parts
953 - Changed query operations to return more detailed information, e.g.
954 whether an information is unavailable due to an offline node. To use
955 this new functionality, the LUXI call ``Query`` must be used. Field
956 information is now stored by the master daemon and can be retrieved
957 using ``QueryFields``. Instances, nodes and groups can also be queried
958 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
959 exposed via RAPI). The following commands make use of this
960 infrastructure change:
963 - ``gnt-group list-fields``
965 - ``gnt-node list-fields``
966 - ``gnt-instance list``
967 - ``gnt-instance list-fields``
968 - ``gnt-debug locks``
973 - New RAPI resources (see :doc:`rapi`):
977 - ``/2/groups/[group_name]``
978 - ``/2/groups/[group_name]/assign-nodes``
979 - ``/2/groups/[group_name]/modify``
980 - ``/2/groups/[group_name]/rename``
981 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
985 - Implemented ``no_install`` for instance creation
986 - Implemented OS parameters for instance reinstallation, allowing
987 use of special settings on reinstallation (e.g. for preserving data)
992 - Added IPv6 support in import/export
993 - Pause DRBD synchronization while wiping disks on instance creation
994 - Updated unittests and QA scripts
995 - Improved network parameters passed to KVM
996 - Converted man pages from docbook to reStructuredText
1002 *(Released Mon, 20 Dec 2010)*
1004 Released version 2.3.1~rc1 without any changes.
1010 *(Released Wed, 1 Dec 2010)*
1012 - impexpd: Disable OpenSSL compression in socat if possible (backport
1013 from master, commit e90739d625b, see :doc:`installation guide
1014 <install-quick>` for details)
1015 - Changed unittest coverage report to exclude test scripts
1016 - Added script to check version format
1022 *(Released Wed, 1 Dec 2010)*
1024 Released version 2.3.0~rc1 without any changes.
1030 *(Released Fri, 19 Nov 2010)*
1032 A number of bugfixes and documentation updates:
1034 - Update ganeti-os-interface documentation
1035 - Fixed a bug related to duplicate MACs or similar items which should be
1037 - Fix breakage in OS state modify
1038 - Reinstall instance: disallow offline secondaries (fixes bug related to
1039 OS changing but reinstall failing)
1040 - plus all the other fixes between 2.2.1 and 2.2.2
1046 *(Released Tue, 2 Nov 2010)*
1048 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
1049 - Fixed master failover race with watcher
1050 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
1052 - Accept previously stopped instance for export with instance removal
1053 - Simplify and extend the environment variables for instance OS scripts
1054 - Added new node flags, ``master_capable`` and ``vm_capable``
1055 - Added optional instance disk wiping prior during allocation. This is a
1056 cluster-wide option and can be set/modified using
1057 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
1058 - Added IPv6 support, see :doc:`design document <design-2.3>` and
1059 :doc:`install-quick`
1060 - Added a new watcher option (``--ignore-pause``)
1061 - Added option to ignore offline node on instance start/stop
1062 (``--ignore-offline``)
1063 - Allow overriding OS parameters with ``gnt-instance reinstall``
1064 - Added ability to change node's secondary IP address using ``gnt-node
1066 - Implemented privilege separation for all daemons except
1067 ``ganeti-noded``, see ``configure`` options
1068 - Complain if an instance's disk is marked faulty in ``gnt-cluster
1070 - Implemented job priorities (see ``ganeti(7)`` manpage)
1071 - Ignore failures while shutting down instances during failover from
1073 - Exit daemon's bootstrap process only once daemon is ready
1074 - Export more information via ``LUInstanceQuery``/remote API
1075 - Improved documentation, QA and unittests
1076 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
1077 restart if the file was created or changed
1078 - Added LUXI protocol version sent with each request and response,
1079 allowing detection of server/client mismatches
1080 - Moved the Python scripts among gnt-* and ganeti-* into modules
1081 - Moved all code related to setting up SSH to an external script,
1083 - Infrastructure changes for node group support in future versions
1089 *(Released Fri, 19 Nov 2010)*
1091 A few small bugs fixed, and some improvements to the build system:
1093 - Fix documentation regarding conversion to drbd
1094 - Fix validation of parameters in cluster modify (``gnt-cluster modify
1096 - Fix error handling in node modify with multiple changes
1097 - Allow remote imports without checked names
1103 *(Released Tue, 19 Oct 2010)*
1105 - Disable SSL session ID cache in RPC client
1111 *(Released Thu, 14 Oct 2010)*
1113 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
1114 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
1116 - Fix problems with interaction between Curl and Python's HTTP server,
1117 resulting in increased speed in many RPC calls
1118 - Improve our release script to prevent breakage with older aclocal and
1125 *(Released Thu, 7 Oct 2010)*
1127 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
1128 value retrieved from master
1129 - Added support for blacklisted or hidden OS definitions
1130 - Added simple lock monitor (accessible via (``gnt-debug locks``)
1131 - Added support for -mem-path in KVM hypervisor abstraction layer
1132 - Allow overriding instance parameters in tool for inter-cluster
1133 instance moves (``tools/move-instance``)
1134 - Improved opcode summaries (e.g. in ``gnt-job list``)
1135 - Improve consistency of OS listing by sorting it
1136 - Documentation updates
1142 *(Released Fri, 8 Oct 2010)*
1144 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
1150 *(Released Mon, 4 Oct 2010)*
1152 - Fixed regression in ``gnt-instance rename``
1158 *(Released Wed, 22 Sep 2010)*
1160 - Fixed OS_VARIANT variable for OS scripts
1161 - Fixed cluster tag operations via RAPI
1162 - Made ``setup-ssh`` exit with non-zero code if an error occurred
1163 - Disabled RAPI CA checks in watcher
1169 *(Released Mon, 23 Aug 2010)*
1171 - Support DRBD versions of the format "a.b.c.d"
1173 - Re-introduce support for usage from multiple threads in RAPI client
1174 - Instance renames and modify via RAPI
1175 - Work around race condition between processing and archival in job
1177 - Mark opcodes following failed one as failed, too
1178 - Job field ``lock_status`` was removed due to difficulties making it
1179 work with the changed job queue in Ganeti 2.2; a better way to monitor
1180 locks is expected for a later 2.2.x release
1181 - Fixed dry-run behaviour with many commands
1182 - Support ``ssh-agent`` again when adding nodes
1183 - Many additional bugfixes
1189 *(Released Fri, 30 Jul 2010)*
1191 Important change: the internal RPC mechanism between Ganeti nodes has
1192 changed from using a home-grown http library (based on the Python base
1193 libraries) to use the PycURL library. This requires that PycURL is
1194 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
1195 against GnuTLS by default. cURL's support for GnuTLS had known issues
1196 before cURL 7.21.0 and we recommend using the latest cURL release or
1197 linking against OpenSSL. Most other distributions already link PycURL
1198 and cURL against OpenSSL. The command::
1200 python -c 'import pycurl; print pycurl.version'
1202 can be used to determine the libraries PycURL and cURL are linked
1205 Other significant changes:
1207 - Rewrote much of the internals of the job queue, in order to achieve
1208 better parallelism; this decouples job query operations from the job
1209 processing, and it should allow much nicer behaviour of the master
1210 daemon under load, and it also has uncovered some long-standing bugs
1211 related to the job serialisation (now fixed)
1212 - Added a default iallocator setting to the cluster parameters,
1213 eliminating the need to always pass nodes or an iallocator for
1214 operations that require selection of new node(s)
1215 - Added experimental support for the LXC virtualization method
1216 - Added support for OS parameters, which allows the installation of
1217 instances to pass parameter to OS scripts in order to customise the
1219 - Added a hypervisor parameter controlling the migration type (live or
1220 non-live), since hypervisors have various levels of reliability; this
1221 has renamed the 'live' parameter to 'mode'
1222 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
1223 logical volumes, meaning that cluster verify will ignore them and not
1224 flag their presence as errors
1225 - The watcher will now reset the error count for failed instances after
1226 8 hours, thus allowing self-healing if the problem that caused the
1227 instances to be down/fail to start has cleared in the meantime
1228 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
1229 check for, and warn, if the drbd module parameter ``usermode_helper``
1230 is not consistent with the cluster-wide setting; this is needed to
1231 make diagnose easier of failed drbd creations
1232 - Started adding base IPv6 support, but this is not yet
1233 enabled/available for use
1234 - Rename operations (cluster, instance) will now return the new name,
1235 which is especially useful if a short name was passed in
1236 - Added support for instance migration in RAPI
1237 - Added a tool to pre-configure nodes for the SSH setup, before joining
1238 them to the cluster; this will allow in the future a simplified model
1239 for node joining (but not yet fully enabled in 2.2); this needs the
1240 paramiko python library
1241 - Fixed handling of name-resolving errors
1242 - Fixed consistency of job results on the error path
1243 - Fixed master-failover race condition when executed multiple times in
1245 - Fixed many bugs related to the job queue (mostly introduced during the
1246 2.2 development cycle, so not all are impacting 2.1)
1247 - Fixed instance migration with missing disk symlinks
1248 - Fixed handling of unknown jobs in ``gnt-job archive``
1249 - And many other small fixes/improvements
1253 - Enhanced both the unittest and the QA coverage
1254 - Switched the opcode validation to a generic model, and extended the
1255 validation to all opcode parameters
1256 - Changed more parts of the code that write shell scripts to use the
1258 - Switched the master daemon to use the asyncore library for the Luxi
1265 *(Released Thu, 17 Jun 2010)*
1267 - Added tool (``move-instance``) and infrastructure to move instances
1268 between separate clusters (see :doc:`separate documentation
1269 <move-instance>` and :doc:`design document <design-2.2>`)
1270 - Added per-request RPC timeout
1271 - RAPI now requires a Content-Type header for requests with a body (e.g.
1272 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
1273 :rfc:`2616` (HTTP/1.1), section 7.2.1)
1274 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
1276 - Implemented initial support for running Ganeti daemons as separate
1277 users, see configure-time flags ``--with-user-prefix`` and
1278 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
1280 - Instances can be removed after export (``gnt-backup export
1281 --remove-instance``)
1282 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
1283 key (instead of 1024 bit)
1284 - Added new cluster configuration file for cluster domain secret
1285 - Import/export now use SSL instead of SSH
1286 - Added support for showing estimated time when exporting an instance,
1287 see the ``ganeti-os-interface(7)`` manpage and look for
1294 *(Released Tue, 16 Nov 2010)*
1296 Some more bugfixes. Unless critical bugs occur, this will be the last
1299 - Fix case of MAC special-values
1300 - Fix mac checker regex
1301 - backend: Fix typo causing "out of range" error
1302 - Add missing --units in gnt-instance list man page
1308 *(Released Tue, 24 Aug 2010)*
1311 - Don't ignore secondary node silently on non-mirrored disk templates
1313 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
1314 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
1315 - Properly document the usb_mouse parameter
1316 - Fix path in ganeti-rapi(8) (issue 116)
1317 - Adjust error message when the ganeti user's .ssh directory is
1319 - Add same-node-check when changing the disk template to drbd
1325 *(Released Fri, 16 Jul 2010)*
1328 - Add an option to only select some reboot types during qa/burnin.
1329 (on some hypervisors consequent reboots are not supported)
1330 - Fix infrequent race condition in master failover. Sometimes the old
1331 master ip address would be still detected as up for a short time
1332 after it was removed, causing failover to fail.
1333 - Decrease mlockall warnings when the ctypes module is missing. On
1334 Python 2.4 we support running even if no ctypes module is installed,
1335 but we were too verbose about this issue.
1336 - Fix building on old distributions, on which man doesn't have a
1338 - Fix RAPI not to ignore the MAC address on instance creation
1339 - Implement the old instance creation format in the RAPI client.
1345 *(Released Thu, 01 Jul 2010)*
1347 A small bugfix release:
1348 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
1349 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
1350 a non-existing option
1351 - Fix instance create: the --force-variant option was ignored
1352 - Improve pylint 0.21 compatibility and warnings with Python 2.6
1353 - Fix modify node storage with non-FQDN arguments
1354 - Fix RAPI client to authenticate under Python 2.6 when used
1355 for more than 5 requests needing authentication
1356 - Fix gnt-instance modify -t (storage) giving a wrong error message
1357 when converting a non-shutdown drbd instance to plain
1363 *(Released Fri, 18 Jun 2010)*
1365 A small bugfix release:
1367 - Fix live migration of KVM instances started with older Ganeti
1368 versions which had fewer hypervisor parameters
1369 - Fix gnt-instance grow-disk on down instances
1370 - Fix an error-reporting bug during instance migration
1371 - Better checking of the ``--net`` and ``--disk`` values, to avoid
1372 silently ignoring broken ones
1373 - Fix an RPC error reporting bug affecting, for example, RAPI client
1375 - Fix bug triggered by different API version os-es on different nodes
1376 - Fix a bug in instance startup with custom hvparams: OS level
1377 parameters would fail to be applied.
1378 - Fix the RAPI client under Python 2.6 (but more work is needed to
1379 make it work completely well with OpenSSL)
1380 - Fix handling of errors when resolving names from DNS
1386 *(Released Thu, 3 Jun 2010)*
1388 A medium sized development cycle. Some new features, and some
1389 fixes/small improvements/cleanups.
1391 Significant features
1392 ~~~~~~~~~~~~~~~~~~~~
1394 The node deamon now tries to mlock itself into memory, unless the
1395 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
1396 its logs, and falls back to console logging. This allows emergency
1397 features such as ``gnt-node powercycle`` to work even in the event of a
1398 broken node disk (tested offlining the disk hosting the node's
1399 filesystem and dropping its memory caches; don't try this at home)
1401 KVM: add vhost-net acceleration support. It can be tested with a new
1402 enough version of the kernel and of qemu-kvm.
1404 KVM: Add instance chrooting feature. If you use privilege dropping for
1405 your VMs you can also now force them to chroot to an empty directory,
1406 before starting the emulated guest.
1408 KVM: Add maximum migration bandwith and maximum downtime tweaking
1409 support (requires a new-enough version of qemu-kvm).
1411 Cluster verify will now warn if the master node doesn't have the master
1412 ip configured on it.
1414 Add a new (incompatible) instance creation request format to RAPI which
1415 supports all parameters (previously only a subset was supported, and it
1416 wasn't possible to extend the old format to accomodate all the new
1417 features. The old format is still supported, and a client can check for
1418 this feature, before using it, by checking for its presence in the
1419 ``features`` RAPI resource.
1421 Now with ancient latin support. Try it passing the ``--roman`` option to
1422 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
1423 (requires the python-roman module to be installed, in order to work).
1428 As usual many internal code refactorings, documentation updates, and
1431 - Lots of improvements and cleanups to the experimental Remote API
1432 (RAPI) client library.
1433 - A new unit test suite for the core daemon libraries.
1434 - A fix to creating missing directories makes sure the umask is not
1435 applied anymore. This enforces the same directory permissions
1437 - Better handling terminating daemons with ctrl+c (used when running
1438 them in debugging mode).
1439 - Fix a race condition in live migrating a KVM instance, when stat()
1440 on the old proc status file returned EINVAL, which is an unexpected
1442 - Fixed manpage checking with newer man and utf-8 charachters. But now
1443 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
1449 *(Released Fri, 7 May 2010)*
1451 Fix a bug which prevented untagged KVM instances from starting.
1457 *(Released Fri, 7 May 2010)*
1459 Another release with a long development cycle, during which many
1460 different features were added.
1462 Significant features
1463 ~~~~~~~~~~~~~~~~~~~~
1465 The KVM hypervisor now can run the individual instances as non-root, to
1466 reduce the impact of a VM being hijacked due to bugs in the
1467 hypervisor. It is possible to run all instances as a single (non-root)
1468 user, to manually specify a user for each instance, or to dynamically
1469 allocate a user out of a cluster-wide pool to each instance, with the
1470 guarantee that no two instances will run under the same user ID on any
1473 An experimental RAPI client library, that can be used standalone
1474 (without the other Ganeti libraries), is provided in the source tree as
1475 ``lib/rapi/client.py``. Note this client might change its interface in
1476 the future, as we iterate on its capabilities.
1478 A new command, ``gnt-cluster renew-crypto`` has been added to easily
1479 replace the cluster's certificates and crypto keys. This might help in
1480 case they have been compromised, or have simply expired.
1482 A new disk option for instance creation has been added that allows one
1483 to "adopt" currently existing logical volumes, with data
1484 preservation. This should allow easier migration to Ganeti from
1485 unmanaged (or managed via other software) instances.
1487 Another disk improvement is the possibility to convert between redundant
1488 (DRBD) and plain (LVM) disk configuration for an instance. This should
1489 allow better scalability (starting with one node and growing the
1490 cluster, or shrinking a two-node cluster to one node).
1492 A new feature that could help with automated node failovers has been
1493 implemented: if a node sees itself as offline (by querying the master
1494 candidates), it will try to shutdown (hard) all instances and any active
1495 DRBD devices. This reduces the risk of duplicate instances if an
1496 external script automatically failovers the instances on such nodes. To
1497 enable this, the cluster parameter ``maintain_node_health`` should be
1498 enabled; in the future this option (per the name) will enable other
1499 automatic maintenance features.
1501 Instance export/import now will reuse the original instance
1502 specifications for all parameters; that means exporting an instance,
1503 deleting it and the importing it back should give an almost identical
1504 instance. Note that the default import behaviour has changed from
1505 before, where it created only one NIC; now it recreates the original
1508 Cluster verify has added a few new checks: SSL certificates validity,
1509 /etc/hosts consistency across the cluster, etc.
1514 As usual, many internal changes were done, documentation fixes,
1517 - Fixed cluster initialization with disabled cluster storage (regression
1518 introduced in 2.1.1)
1519 - File-based storage supports growing the disks
1520 - Fixed behaviour of node role changes
1521 - Fixed cluster verify for some corner cases, plus a general rewrite of
1522 cluster verify to allow future extension with more checks
1523 - Fixed log spamming by watcher and node daemon (regression introduced
1525 - Fixed possible validation issues when changing the list of enabled
1527 - Fixed cleanup of /etc/hosts during node removal
1528 - Fixed RAPI response for invalid methods
1529 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
1530 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
1531 from ide disks, etc.)
1532 - Allow OS changes without re-installation (to record a changed OS
1533 outside of Ganeti, or to allow OS renames)
1534 - Allow instance creation without OS installation (useful for example if
1535 the OS will be installed manually, or restored from a backup not in
1537 - Implemented option to make cluster ``copyfile`` use the replication
1539 - Added list of enabled hypervisors to ssconf (possibly useful for
1541 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
1543 - A partial form of node re-IP is possible via node readd, which now
1544 allows changed node primary IP
1545 - Command line utilities now show an informational message if the job is
1547 - The logs of the master daemon now show the PID/UID/GID of the
1554 *(Released Fri, 12 Mar 2010)*
1556 During the 2.1.0 long release candidate cycle, a lot of improvements and
1557 changes have accumulated with were released later as 2.1.1.
1562 The node evacuate command (``gnt-node evacuate``) was significantly
1563 rewritten, and as such the IAllocator protocol was changed - a new
1564 request type has been added. This unfortunate change during a stable
1565 series is designed to improve performance of node evacuations; on
1566 clusters with more than about five nodes and which are well-balanced,
1567 evacuation should proceed in parallel for all instances of the node
1568 being evacuated. As such, any existing IAllocator scripts need to be
1569 updated, otherwise the above command will fail due to the unknown
1570 request. The provided "dumb" allocator has not been updated; but the
1571 ganeti-htools package supports the new protocol since version 0.2.4.
1573 Another important change is increased validation of node and instance
1574 names. This might create problems in special cases, if invalid host
1575 names are being used.
1577 Also, a new layer of hypervisor parameters has been added, that sits at
1578 OS level between the cluster defaults and the instance ones. This allows
1579 customisation of virtualization parameters depending on the installed
1580 OS. For example instances with OS 'X' may have a different KVM kernel
1581 (or any other parameter) than the cluster defaults. This is intended to
1582 help managing a multiple OSes on the same cluster, without manual
1583 modification of each instance's parameters.
1585 A tool for merging clusters, ``cluster-merge``, has been added in the
1586 tools sub-directory.
1591 - Improved the int/float conversions that should make the code more
1592 robust in face of errors from the node daemons
1593 - Fixed the remove node code in case of internal configuration errors
1594 - Fixed the node daemon behaviour in face of inconsistent queue
1595 directory (e.g. read-only file-system where we can't open the files
1597 - Fixed the behaviour of gnt-node modify for master candidate demotion;
1598 now it either aborts cleanly or, if given the new "auto_promote"
1599 parameter, will automatically promote other nodes as needed
1600 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
1601 completely prevent Ganeti from working
1602 - Fixed bug for instance export when not all disks were successfully
1604 - Fixed behaviour of node add when the new node is slow in starting up
1606 - Fixed handling of signals in the LUXI client, which should improve
1607 behaviour of command-line scripts
1608 - Added checks for invalid node/instance names in the configuration (now
1609 flagged during cluster verify)
1610 - Fixed watcher behaviour for disk activation errors
1611 - Fixed two potentially endless loops in http library, which led to the
1612 RAPI daemon hanging and consuming 100% CPU in some cases
1613 - Fixed bug in RAPI daemon related to hashed passwords
1614 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
1616 - Enhanced compatibility with non-Debian OSes, but not using absolute
1617 path in some commands and allowing customisation of the ssh
1618 configuration directory
1619 - Fixed possible future issue with new Python versions by abiding to the
1620 proper use of ``__slots__`` attribute on classes
1621 - Added checks that should prevent directory traversal attacks
1622 - Many documentation fixes based on feedback from users
1627 - Added an "early_release" more for instance replace disks and node
1628 evacuate, where we release locks earlier and thus allow higher
1629 parallelism within the cluster
1630 - Added watcher hooks, intended to allow the watcher to restart other
1631 daemons (e.g. from the ganeti-nbma project), but they can be used of
1632 course for any other purpose
1633 - Added a compile-time disable for DRBD barriers, to increase
1634 performance if the administrator trusts the power supply or the
1635 storage system to not lose writes
1636 - Added the option of using syslog for logging instead of, or in
1637 addition to, Ganeti's own log files
1638 - Removed boot restriction for paravirtual NICs for KVM, recent versions
1639 can indeed boot from a paravirtual NIC
1640 - Added a generic debug level for many operations; while this is not
1641 used widely yet, it allows one to pass the debug value all the way to
1643 - Enhanced the hooks environment for instance moves (failovers,
1644 migrations) where the primary/secondary nodes changed during the
1645 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
1646 - Enhanced data validations for many user-supplied values; one important
1647 item is the restrictions imposed on instance and node names, which
1648 might reject some (invalid) host names
1649 - Add a configure-time option to disable file-based storage, if it's not
1650 needed; this allows greater security separation between the master
1651 node and the other nodes from the point of view of the inter-node RPC
1653 - Added user notification in interactive tools if job is waiting in the
1654 job queue or trying to acquire locks
1655 - Added log messages when a job is waiting for locks
1656 - Added filtering by node tags in instance operations which admit
1657 multiple instances (start, stop, reboot, reinstall)
1658 - Added a new tool for cluster mergers, ``cluster-merge``
1659 - Parameters from command line which are of the form ``a=b,c=d`` can now
1660 use backslash escapes to pass in values which contain commas,
1661 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
1663 - For KVM, the instance name is the first parameter passed to KVM, so
1664 that it's more visible in the process list
1670 *(Released Tue, 2 Mar 2010)*
1672 Ganeti 2.1 brings many improvements with it. Major changes:
1674 - Added infrastructure to ease automated disk repairs
1675 - Added new daemon to export configuration data in a cheaper way than
1676 using the remote API
1677 - Instance NICs can now be routed instead of being associated with a
1679 - Improved job locking logic to reduce impact of jobs acquiring multiple
1680 locks waiting for other long-running jobs
1682 In-depth implementation details can be found in the Ganeti 2.1 design
1688 - Added chroot hypervisor
1689 - Added more options to xen-hvm hypervisor (``kernel_path`` and
1691 - Added more options to xen-pvm hypervisor (``use_bootloader``,
1692 ``bootloader_path`` and ``bootloader_args``)
1693 - Added the ``use_localtime`` option for the xen-hvm and kvm
1694 hypervisors, and the default value for this has changed to false (in
1695 2.0 xen-hvm always enabled it)
1696 - Added luxi call to submit multiple jobs in one go
1697 - Added cluster initialization option to not modify ``/etc/hosts``
1699 - Added network interface parameters
1700 - Added dry run mode to some LUs
1701 - Added RAPI resources:
1703 - ``/2/instances/[instance_name]/info``
1704 - ``/2/instances/[instance_name]/replace-disks``
1705 - ``/2/nodes/[node_name]/evacuate``
1706 - ``/2/nodes/[node_name]/migrate``
1707 - ``/2/nodes/[node_name]/role``
1708 - ``/2/nodes/[node_name]/storage``
1709 - ``/2/nodes/[node_name]/storage/modify``
1710 - ``/2/nodes/[node_name]/storage/repair``
1712 - Added OpCodes to evacuate or migrate all instances on a node
1713 - Added new command to list storage elements on nodes (``gnt-node
1714 list-storage``) and modify them (``gnt-node modify-storage``)
1715 - Added new ssconf files with master candidate IP address
1716 (``ssconf_master_candidates_ips``), node primary IP address
1717 (``ssconf_node_primary_ips``) and node secondary IP address
1718 (``ssconf_node_secondary_ips``)
1719 - Added ``ganeti-confd`` and a client library to query the Ganeti
1720 configuration via UDP
1721 - Added ability to run hooks after cluster initialization and before
1723 - Added automatic mode for disk replace (``gnt-instance replace-disks
1725 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
1726 after catastrophic data-loss
1727 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
1729 - Added ``gnt-instance move`` command to move instances
1730 - Added ``gnt-cluster watcher`` command to control watcher
1731 - Added ``gnt-node powercycle`` command to powercycle nodes
1732 - Added new job status field ``lock_status``
1733 - Added parseable error codes to cluster verification (``gnt-cluster
1734 verify --error-codes``) and made output less verbose (use
1735 ``--verbose`` to restore previous behaviour)
1736 - Added UUIDs to the main config entities (cluster, nodes, instances)
1737 - Added support for OS variants
1738 - Added support for hashed passwords in the Ganeti remote API users file
1740 - Added option to specify maximum timeout on instance shutdown
1741 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
1742 - Added new helper script to start and stop Ganeti daemons
1743 (``daemon-util``), with the intent to reduce the work necessary to
1744 adjust Ganeti for non-Debian distributions and to start/stop daemons
1746 - Added more unittests
1747 - Fixed critical bug in ganeti-masterd startup
1748 - Removed the configure-time ``kvm-migration-port`` parameter, this is
1749 now customisable at the cluster level for both the KVM and Xen
1750 hypervisors using the new ``migration_port`` parameter
1751 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
1752 reinstalling an instance
1753 - Allowed ``@`` in tag names
1754 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
1755 - Many documentation updates
1756 - Distribute hypervisor files on ``gnt-cluster redist-conf``
1757 - ``gnt-instance reinstall`` can now reinstall multiple instances
1758 - Updated many command line parameters
1759 - Introduced new OS API version 15
1760 - No longer support a default hypervisor
1761 - Treat virtual LVs as inexistent
1762 - Improved job locking logic to reduce lock contention
1763 - Match instance and node names case insensitively
1764 - Reimplemented bash completion script to be more complete
1771 *(Released Thu, 4 Feb 2010)*
1773 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
1774 - Fix a string formatting bug
1775 - Improve safety of the code in some error paths
1776 - Improve data validation in the master of values returned from nodes
1782 *(Released Thu, 17 Dec 2009)*
1784 - Fix security issue due to missing validation of iallocator names; this
1785 allows local and remote execution of arbitrary executables
1786 - Fix failure of gnt-node list during instance removal
1787 - Ship the RAPI documentation in the archive
1793 *(Released Wed, 30 Sep 2009)*
1795 - Fixed many wrong messages
1796 - Fixed a few bugs related to the locking library
1797 - Fixed MAC checking at instance creation time
1798 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
1799 - Fixed a few issues related to signal handling in both daemons and
1801 - Fixed the example startup script provided
1802 - Fixed insserv dependencies in the example startup script (patch from
1804 - Fixed handling of drained nodes in the iallocator framework
1805 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
1807 - Fixed error related to invalid job IDs in job polling
1808 - Fixed job/opcode persistence on unclean master shutdown
1809 - Fixed handling of partial job processing after unclean master
1811 - Fixed error reporting from LUs, previously all errors were converted
1812 into execution errors
1813 - Fixed error reporting from burnin
1814 - Decreased significantly the memory usage of the job queue
1815 - Optimised slightly multi-job submission
1816 - Optimised slightly opcode loading
1817 - Backported the multi-job submit framework from the development
1818 branch; multi-instance start and stop should be faster
1819 - Added script to clean archived jobs after 21 days; this will reduce
1820 the size of the queue directory
1821 - Added some extra checks in disk size tracking
1822 - Added an example ethers hook script
1823 - Added a cluster parameter that prevents Ganeti from modifying of
1825 - Added more node information to RAPI responses
1826 - Added a ``gnt-job watch`` command that allows following the ouput of a
1828 - Added a bind-address option to ganeti-rapi
1829 - Added more checks to the configuration verify
1830 - Enhanced the burnin script such that some operations can be retried
1832 - Converted instance reinstall to multi-instance model
1838 *(Released Fri, 7 Aug 2009)*
1840 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
1841 to allow using the pre-2.0.2 behaviour in activation, if any existing
1842 instances have mismatched disk sizes in the configuration
1843 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
1844 any configuration mismatches for disk sizes
1845 - Added ``gnt-master cluste-failover --no-voting`` to allow master
1846 failover to work on two-node clusters
1847 - Fixed the ``--net`` option of ``gnt-backup import``, which was
1849 - Fixed detection of OS script errors in ``gnt-backup export``
1850 - Fixed exit code of ``gnt-backup export``
1856 *(Released Fri, 17 Jul 2009)*
1858 - Added experimental support for stripped logical volumes; this should
1859 enhance performance but comes with a higher complexity in the block
1860 device handling; stripping is only enabled when passing
1861 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
1862 affected even in the non-stripped mode
1863 - Improved resiliency against transient failures at the end of DRBD
1864 resyncs, and in general of DRBD resync checks
1865 - Fixed a couple of issues with exports and snapshot errors
1866 - Fixed a couple of issues in instance listing
1867 - Added display of the disk size in ``gnt-instance info``
1868 - Fixed checking for valid OSes in instance creation
1869 - Fixed handling of the "vcpus" parameter in instance listing and in
1870 general of invalid parameters
1871 - Fixed http server library, and thus RAPI, to handle invalid
1872 username/password combinations correctly; this means that now they
1873 report unauthorized for queries too, not only for modifications,
1874 allowing earlier detect of configuration problems
1875 - Added a new "role" node list field, equivalent to the master/master
1876 candidate/drained/offline flags combinations
1877 - Fixed cluster modify and changes of candidate pool size
1878 - Fixed cluster verify error messages for wrong files on regular nodes
1879 - Fixed a couple of issues with node demotion from master candidate role
1880 - Fixed node readd issues
1881 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
1882 - Added a new ``--no-voting`` option for masterfailover to fix failover
1883 on two-nodes clusters when the former master node is unreachable
1884 - Added instance reinstall over RAPI
1890 *(Released Tue, 16 Jun 2009)*
1892 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
1893 allow re-adding the start in single-user option (regression from 1.2)
1894 - the watcher writes the instance status to a file, to allow monitoring
1895 to report the instance status (from the master) based on cached
1896 results of the watcher's queries; while this can get stale if the
1897 watcher is being locked due to other work on the cluster, this is
1898 still an improvement
1899 - the watcher now also restarts the node daemon and the rapi daemon if
1901 - fixed the watcher to handle full and drained queue cases
1902 - hooks export more instance data in the environment, which helps if
1903 hook scripts need to take action based on the instance's properties
1904 (no longer need to query back into ganeti)
1905 - instance failovers when the instance is stopped do not check for free
1906 RAM, so that failing over a stopped instance is possible in low memory
1908 - rapi uses queries for tags instead of jobs (for less job traffic), and
1909 for cluster tags it won't talk to masterd at all but read them from
1911 - a couple of error handling fixes in RAPI
1912 - drbd handling: improved the error handling of inconsistent disks after
1913 resync to reduce the frequency of "there are some degraded disks for
1914 this instance" messages
1915 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
1916 error handling path called a wrong function name)
1922 *(Released Wed, 27 May 2009)*
1924 - no changes from rc5
1930 *(Released Wed, 20 May 2009)*
1932 - fix a couple of bugs (validation, argument checks)
1933 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
1934 - some small improvements to RAPI and IAllocator
1935 - make watcher automatically start the master daemon if down
1941 *(Released Mon, 27 Apr 2009)*
1943 - change the OS list to not require locks; this helps with big clusters
1944 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
1945 volume group is broken
1946 - ``gnt-instance info``, without any arguments, doesn't run for all
1947 instances anymore; either pass ``--all`` or pass the desired
1948 instances; this helps against mistakes on big clusters where listing
1949 the information for all instances takes a long time
1950 - miscellaneous doc and man pages fixes
1956 *(Released Wed, 8 Apr 2009)*
1958 - Change the internal locking model of some ``gnt-node`` commands, in
1959 order to reduce contention (and blocking of master daemon) when
1960 batching many creation/reinstall jobs
1961 - Fixes to Xen soft reboot
1962 - No longer build documentation at build time, instead distribute it in
1963 the archive, in order to reduce the need for the whole docbook/rst
1970 *(Released Fri, 27 Mar 2009)*
1972 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
1973 - Fix watcher startup sequence, improves the behaviour of busy clusters
1974 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
1975 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
1976 bind address and other places
1977 - Some documentation fixes and updates
1983 *(Released Mon, 2 Mar 2009)*
1985 - More documentation updates, now all docs should be more-or-less
1987 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
1989 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
1991 - Fix an issue related to $libdir/run/ganeti and cluster creation
1997 *(Released Thu, 19 Feb 2009)*
1999 - Xen PVM and KVM have switched the default value for the instance root
2000 disk to the first partition on the first drive, instead of the whole
2001 drive; this means that the OS installation scripts must be changed
2003 - Man pages have been updated
2004 - RAPI has been switched by default to HTTPS, and the exported functions
2005 should all work correctly
2006 - RAPI v1 has been removed
2007 - Many improvements to the KVM hypervisor
2008 - Block device errors are now better reported
2009 - Many other bugfixes and small improvements
2015 *(Released Mon, 26 Jan 2009)*
2017 - Version 2 is a general rewrite of the code and therefore the
2018 differences are too many to list, see the design document for 2.0 in
2019 the ``doc/`` subdirectory for more details
2020 - In this beta version there is not yet a migration path from 1.2 (there
2021 will be one in the final 2.0 release)
2022 - A few significant changes are:
2024 - all commands are executed by a daemon (``ganeti-masterd``) and the
2025 various ``gnt-*`` commands are just front-ends to it
2026 - all the commands are entered into, and executed from a job queue,
2027 see the ``gnt-job(8)`` manpage
2028 - the RAPI daemon supports read-write operations, secured by basic
2029 HTTP authentication on top of HTTPS
2030 - DRBD version 0.7 support has been removed, DRBD 8 is the only
2031 supported version (when migrating from Ganeti 1.2 to 2.0, you need
2032 to migrate to DRBD 8 first while still running Ganeti 1.2)
2033 - DRBD devices are using statically allocated minor numbers, which
2034 will be assigned to existing instances during the migration process
2035 - there is support for both Xen PVM and Xen HVM instances running on
2037 - KVM virtualization is supported too
2038 - file-based storage has been implemented, which means that it is
2039 possible to run the cluster without LVM and DRBD storage, for
2040 example using a shared filesystem exported from shared storage (and
2041 still have live migration)
2047 *(Released Tue, 13 Jan 2009)*
2049 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
2050 - Reuse the old instance mac address by default on instance import, if
2051 the instance name is the same.
2052 - Handle situations in which the node info rpc returns incomplete
2054 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
2055 - Improved version of batcher:
2057 - state file support
2058 - instance mac address support
2059 - support for HVM clusters/instances
2061 - Add an option to show the number of cpu sockets and nodes in
2063 - Support OSes that handle more than one version of the OS api (but do
2064 not change the current API in any other way)
2065 - Fix ``gnt-node migrate``
2066 - ``gnt-debug`` man page
2067 - Fixes various more typos and small issues
2068 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
2074 *(Released Wed, 24 Sep 2008)*
2076 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
2077 type of disk exported to fully virtualized instances.
2078 - provide access to the serial console of HVM instances
2079 - instance auto_balance flag, set by default. If turned off it will
2080 avoid warnings on cluster verify if there is not enough memory to fail
2081 over an instance. in the future it will prevent automatically failing
2082 it over when we will support that.
2083 - batcher tool for instance creation, see ``tools/README.batcher``
2084 - ``gnt-instance reinstall --select-os`` to interactively select a new
2085 operating system when reinstalling an instance.
2086 - when changing the memory amount on instance modify a check has been
2087 added that the instance will be able to start. also warnings are
2088 emitted if the instance will not be able to fail over, if auto_balance
2090 - documentation fixes
2091 - sync fields between ``gnt-instance list/modify/add/import``
2092 - fix a race condition in drbd when the sync speed was set after giving
2093 the device a remote peer.
2099 *(Released Tue, 22 Jul 2008)*
2101 - note: the allowed size and number of tags per object were reduced
2102 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
2103 - fixed twisted 8.x compatibility
2104 - fixed ``gnt-instance replace-disks`` with iallocator
2105 - add TCP keepalives on twisted connections to detect restarted nodes
2106 - disk increase support, see ``gnt-instance grow-disk``
2107 - implement bulk node/instance query for RAPI
2108 - add tags in node/instance listing (optional)
2109 - experimental migration (and live migration) support, read the man page
2110 for ``gnt-instance migrate``
2111 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
2112 has some small improvements in handling its state file
2118 *(Released Fri, 13 Jun 2008)*
2120 - Experimental readonly, REST-based remote API implementation;
2121 automatically started on master node, TCP port 5080, if enabled by
2122 ``--enable-rapi`` parameter to configure script.
2123 - Instance allocator support. Add and import instance accept a
2124 ``--iallocator`` parameter, and call that instance allocator to decide
2125 which node to use for the instance. The iallocator document describes
2126 what's expected from an allocator script.
2127 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
2128 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
2129 node is lost there is still enough memory to fail over the instances
2131 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
2132 ``gnt-cluster verify``, to check for site-specific compliance. All the
2133 hooks will run, and their output, if any, will be displayed. Any
2134 failing hook will make the verification return an error value.
2135 - ``gnt-cluster verify`` now checks that its peers are reachable on the
2136 primary and secondary interfaces
2137 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
2138 that is still declared as part of the cluster and has failed.
2139 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
2140 specifying output fields, that just adds the chosen fields to the
2142 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
2143 export from the filesystem.
2144 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
2145 have been added. Using them you can enable/disable acpi and pae
2146 support, and specify a path for a cd image to be exported to the
2147 instance. These parameters as the name suggest only work on HVM
2149 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
2150 PAE support will be set to the previously hardcoded values, but the
2151 (previously hardcoded) path to the CDROM ISO image will be unset and
2152 if required, needs to be set manually with ``gnt-instance modify``
2154 - The address to which an instance's VNC console is bound is now
2155 selectable per-instance, rather than being cluster wide. Of course
2156 this only applies to instances controlled via VNC, so currently just
2157 applies to HVM clusters.
2163 *(Released Mon, 18 Feb 2008)*
2165 - more tweaks to the disk activation code (especially helpful for DRBD)
2166 - change the default ``gnt-instance list`` output format, now there is
2167 one combined status field (see the manpage for the exact values this
2169 - some more fixes for the mac export to hooks change
2170 - make Ganeti not break with DRBD 8.2.x (which changed the version
2171 format in ``/proc/drbd``) (issue 24)
2172 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
2173 template, allowing migration from DRBD0.7+MD to DRBD8
2179 *(Released Wed, 30 Jan 2008)*
2181 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
2183 - add command aliases infrastructure and a few aliases
2184 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
2185 man pages and the ``--help`` option of ``gnt-node
2186 list``/``gnt-instance list``
2187 - fix ``gnt-backup list`` with down nodes (issue 21)
2188 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
2189 - fix the dist archive and add a check for including svn/git files in
2191 - some developer-related changes: improve the burnin and the QA suite,
2192 add an upload script for testing during development
2198 *(Released Wed, 16 Jan 2008)*
2200 - experimental HVM support, read the install document, section
2201 "Initializing the cluster"
2202 - allow for the PVM hypervisor per-instance kernel and initrd paths
2203 - add a new command ``gnt-cluster verify-disks`` which uses a new
2204 algorithm to improve the reconnection of the DRBD pairs if the device
2205 on the secondary node has gone away
2206 - make logical volume code auto-activate LVs at disk activation time
2207 - slightly improve the speed of activating disks
2208 - allow specification of the MAC address at instance creation time, and
2209 changing it later via ``gnt-instance modify``
2210 - fix handling of external commands that generate lots of output on
2212 - update documentation with regard to minimum version of DRBD8 supported
2218 *(Released Tue, 4 Dec 2007)*
2220 - Log the ``xm create`` output to the node daemon log on failure (to
2221 help diagnosing the error)
2222 - In debug mode, log all external commands output if failed to the logs
2223 - Change parsing of lvm commands to ignore stderr
2229 *(Released Wed, 28 Nov 2007)*
2231 - Another round of updates to the DRBD 8 code to deal with more failures
2232 in the replace secondary node operation
2233 - Some more logging of failures in disk operations (lvm, drbd)
2234 - A few documentation updates
2241 *(Released Tue, 13 Nov 2007)*
2243 - Change configuration file format from Python's Pickle to JSON.
2244 Upgrading is possible using the cfgupgrade utility.
2245 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
2246 faster replace disks and is more stable (DRBD 8 has many improvements
2247 compared to DRBD 0.7)
2248 - Added command line tags support (see man pages for ``gnt-instance``,
2249 ``gnt-node``, ``gnt-cluster``)
2250 - Added instance rename support
2251 - Added multi-instance startup/shutdown
2252 - Added cluster rename support
2253 - Added ``gnt-node evacuate`` to simplify some node operations
2254 - Added instance reboot operation that can speedup reboot as compared to
2256 - Soften the requirement that hostnames are in FQDN format
2257 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
2259 - Removed dependency on debian's patched fping that uses the
2260 non-standard ``-S`` option
2261 - Now the OS definitions are searched for in multiple, configurable
2262 paths (easier for distros to package)
2263 - Some changes to the hooks infrastructure (especially the new
2264 post-configuration update hook)
2265 - Other small bugfixes
2267 .. vim: set textwidth=72 syntax=rst :