10 - ``gnt-instance batch-create`` has been changed to use the bulk create
11 opcode from Ganeti. This lead to incompatible changes in the format of
12 the JSON file. It's now not a custom dict anymore but a dict
13 compatible with the ``OpInstanceCreate`` opcode.
14 - Parent directories for file storage need now to be listed in
15 ``$sysconfdir/ganeti/file-storage-paths``. ``cfgupgrade`` will write
16 the file automatically based on old configuration values, but it can
17 not distribute it across all nodes and the file contents should be
18 verified. Use ``gnt-cluster copyfile
19 $sysconfdir/ganeti/file-storage-paths`` once the cluster has been
20 upgraded. The reason for requiring this list of paths now is that
21 before it would have been possible to inject new paths via RPC,
22 allowing files to be created in arbitrary locations. The RPC protocol
23 is protected using SSL/X.509 certificates, but as a design principle
24 Ganeti does not permit arbitrary paths to be passed.
25 - The parsing of the variants file for OSes (see
26 :manpage:`ganeti-os-interface(8)` has been slightly changed: now empty
27 lines and comment lines are ignored for better readability.
28 - The ``setup-ssh`` tool added in Ganeti 2.2 has been replaced.
29 ``gnt-node add`` now invokes a new tool on the destination node, named
30 ``prepare-node-join``, to configure the SSH daemon. Paramiko is no
31 longer necessary to configure nodes' SSH daemons via ``gnt-node add``.
32 - A new user option, :pyeval:`rapi.RAPI_ACCESS_READ`, has been added
33 for RAPI users. It allows granting permissions to query for
34 information to a specific user without giving
35 :pyeval:`rapi.RAPI_ACCESS_WRITE` permissions.
41 *(Released Fri, 12 Oct 2012)*
43 A small bugfix release. Among the bugs fixed:
45 - Fixed double use of ``PRIORITY_OPT`` in ``gnt-node migrate``, that
46 made the command unusable.
47 - Commands that issue many jobs don't fail anymore just because some jobs
48 take so long that other jobs are archived.
49 - Failures during ``gnt-instance reinstall`` are reflected by the exit
51 - Issue 190 fixed. Check for DRBD in cluster verify is enabled only when
53 - When ``always_failover`` is set, ``--allow-failover`` is not required
54 in migrate commands anymore.
55 - ``bash_completion`` works even if extglob is disabled.
56 - Fixed bug with locks that made failover for RDB-based instances fail.
57 - Fixed bug in non-mirrored instance allocation that made Ganeti choose
58 a random node instead of one based on the allocator metric.
59 - Support for newer versions of pylint and pep8.
60 - Hail doesn't fail anymore when trying to add an instance of type
61 ``file``, ``sharedfile`` or ``rbd``.
62 - Added new Makefile target to rebuild the whole distribution, so that
63 all files are included.
69 *(Released Fri, 27 Jul 2012)*
72 .. attention:: The ``LUXI`` protocol has been made more consistent
73 regarding its handling of command arguments. This, however, leads to
74 incompatibility issues with previous versions. Please ensure that you
75 restart Ganeti daemons soon after the upgrade, otherwise most
76 ``LUXI`` calls (job submission, setting/resetting the drain flag,
77 pausing/resuming the watcher, cancelling and archiving jobs, querying
78 the cluster configuration) will fail.
87 The current ``admin_up`` field, which used to denote whether an instance
88 should be running or not, has been removed. Instead, ``admin_state`` is
89 introduced, with 3 possible values -- ``up``, ``down`` and ``offline``.
91 The rational behind this is that an instance being “down” can have
94 - it could be down during a reboot
95 - it could be temporarily be down for a reinstall
96 - or it could be down because it is deprecated and kept just for its
99 The previous Boolean state was making it difficult to do capacity
100 calculations: should Ganeti reserve memory for a down instance? Now, the
101 tri-state field makes it clear:
103 - in ``up`` and ``down`` state, all resources are reserved for the
104 instance, and it can be at any time brought up if it is down
105 - in ``offline`` state, only disk space is reserved for it, but not
108 The field can have an extra use: since the transition between ``up`` and
109 ``down`` and vice-versus is done via ``gnt-instance start/stop``, but
110 transition between ``offline`` and ``down`` is done via ``gnt-instance
111 modify``, it is possible to given different rights to users. For
112 example, owners of an instance could be allowed to start/stop it, but
113 not transition it out of the offline state.
115 Instance policies and specs
116 +++++++++++++++++++++++++++
118 In previous Ganeti versions, an instance creation request was not
119 limited on the minimum size and on the maximum size just by the cluster
120 resources. As such, any policy could be implemented only in third-party
121 clients (RAPI clients, or shell wrappers over ``gnt-*``
122 tools). Furthermore, calculating cluster capacity via ``hspace`` again
123 required external input with regards to instance sizes.
125 In order to improve these workflows and to allow for example better
126 per-node group differentiation, we introduced instance specs, which
129 - minimum instance disk size, disk count, memory size, cpu count
130 - maximum values for the above metrics
131 - and “standard” values (used in ``hspace`` to calculate the standard
134 The minimum/maximum values can be also customised at node-group level,
135 for example allowing more powerful hardware to support bigger instance
138 Beside the instance specs, there are a few other settings belonging to
139 the instance policy framework. It is possible now to customise, per
140 cluster and node-group:
142 - the list of allowed disk templates
143 - the maximum ratio of VCPUs per PCPUs (to control CPU oversubscription)
144 - the maximum ratio of instance to spindles (see below for more
145 information) for local storage
147 All these together should allow all tools that talk to Ganeti to know
148 what are the ranges of allowed values for instances and the
149 over-subscription that is allowed.
151 For the VCPU/PCPU ratio, we already have the VCPU configuration from the
152 instance configuration, and the physical CPU configuration from the
153 node. For the spindle ratios however, we didn't track before these
154 values, so new parameters have been added:
156 - a new node parameter ``spindle_count``, defaults to 1, customisable at
157 node group or node level
158 - at new backend parameter (for instances), ``spindle_use`` defaults to 1
160 Note that spindles in this context doesn't need to mean actual
161 mechanical hard-drives; it's just a relative number for both the node
162 I/O capacity and instance I/O consumption.
164 Instance migration behaviour
165 ++++++++++++++++++++++++++++
167 While live-migration is in general desirable over failover, it is
168 possible that for some workloads it is actually worse, due to the
169 variable time of the “suspend” phase during live migration.
171 To allow the tools to work consistently over such instances (without
172 having to hard-code instance names), a new backend parameter
173 ``always_failover`` has been added to control the migration/failover
174 behaviour. When set to True, all migration requests for an instance will
175 instead fall-back to failover.
177 Instance memory ballooning
178 ++++++++++++++++++++++++++
180 Initial support for memory ballooning has been added. The memory for an
181 instance is no longer fixed (backend parameter ``memory``), but instead
182 can vary between minimum and maximum values (backend parameters
183 ``minmem`` and ``maxmem``). Currently we only change an instance's
186 - live migrating or failing over and instance and the target node
187 doesn't have enough memory
188 - user requests changing the memory via ``gnt-instance modify
194 In order to control the use of specific CPUs by instance, support for
195 controlling CPU pinning has been added for the Xen, HVM and LXC
196 hypervisors. This is controlled by a new hypervisor parameter
197 ``cpu_mask``; details about possible values for this are in the
198 :manpage:`gnt-instance(8)`. Note that use of the most specific (precise
199 VCPU-to-CPU mapping) form will work well only when all nodes in your
200 cluster have the same amount of CPUs.
205 Another area in which Ganeti was not customisable were the parameters
206 used for storage configuration, e.g. how many stripes to use for LVM,
207 DRBD resync configuration, etc.
209 To improve this area, we've added disks parameters, which are
210 customisable at cluster and node group level, and which allow to
211 specify various parameters for disks (DRBD has the most parameters
212 currently), for example:
214 - DRBD resync algorithm and parameters (e.g. speed)
215 - the default VG for meta-data volumes for DRBD
216 - number of stripes for LVM (plain disk template)
219 These parameters can be modified via ``gnt-cluster modify -D …`` and
220 ``gnt-group modify -D …``, and are used at either instance creation (in
221 case of LVM stripes, for example) or at disk “activation” time
224 Rados block device support
225 ++++++++++++++++++++++++++
227 A Rados (http://ceph.com/wiki/Rbd) storage backend has been added,
228 denoted by the ``rbd`` disk template type. This is considered
229 experimental, feedback is welcome. For details on configuring it, see
230 the :doc:`install` document and the :manpage:`gnt-cluster(8)` man page.
235 The existing master IP functionality works well only in simple setups (a
236 single network shared by all nodes); however, if nodes belong to
237 different networks, then the ``/32`` setup and lack of routing
238 information is not enough.
240 To allow the master IP to function well in more complex cases, the
241 system was reworked as follows:
243 - a master IP netmask setting has been added
244 - the master IP activation/turn-down code was moved from the node daemon
246 - whether to run the Ganeti-supplied master IP script or a user-supplied
247 on is a ``gnt-cluster init`` setting
249 Details about the location of the standard and custom setup scripts are
250 in the man page :manpage:`gnt-cluster(8)`; for information about the
251 setup script protocol, look at the Ganeti-supplied script.
256 The `SPICE <http://www.linux-kvm.org/page/SPICE>`_ support has been
259 It is now possible to use TLS-protected connections, and when renewing
260 or changing the cluster certificates (via ``gnt-cluster renew-crypto``,
261 it is now possible to specify spice or spice CA certificates. Also, it
262 is possible to configure a password for SPICE sessions via the
263 hypervisor parameter ``spice_password_file``.
265 There are also new parameters to control the compression and streaming
266 options (e.g. ``spice_image_compression``, ``spice_streaming_video``,
267 etc.). For details, see the man page :manpage:`gnt-instance(8)` and look
268 for the spice parameters.
270 Lastly, it is now possible to see the SPICE connection information via
271 ``gnt-instance console``.
276 A new tool (``tools/ovfconverter``) has been added that supports
277 conversion between Ganeti and the `Open Virtualization Format
278 <http://en.wikipedia.org/wiki/Open_Virtualization_Format>`_ (both to and
281 This relies on the ``qemu-img`` tool to convert the disk formats, so the
282 actual compatibility with other virtualization solutions depends on it.
287 The configuration query daemon (``ganeti-confd``) is now optional, and
288 has been rewritten in Haskell; whether to use the daemon at all, use the
289 Python (default) or the Haskell version is selectable at configure time
290 via the ``--enable-confd`` parameter, which can take one of the
291 ``haskell``, ``python`` or ``no`` values. If not used, disabling the
292 daemon will result in a smaller footprint; for larger systems, we
293 welcome feedback on the Haskell version which might become the default
296 If you want to use ``gnt-node list-drbd`` you need to have the Haskell
297 daemon running. The Python version doesn't implement the new call.
300 User interface changes
301 ~~~~~~~~~~~~~~~~~~~~~~
303 We have replaced the ``--disks`` option of ``gnt-instance
304 replace-disks`` with a more flexible ``--disk`` option, which allows
305 adding and removing disks at arbitrary indices (Issue 188). Furthermore,
306 disk size and mode can be changed upon recreation (via ``gnt-instance
307 recreate-disks``, which accepts the same ``--disk`` option).
309 As many people are used to a ``show`` command, we have added that as an
310 alias to ``info`` on all ``gnt-*`` commands.
312 The ``gnt-instance grow-disk`` command has a new mode in which it can
313 accept the target size of the disk, instead of the delta; this can be
314 more safe since two runs in absolute mode will be idempotent, and
315 sometimes it's also easier to specify the desired size directly.
317 Also the handling of instances with regard to offline secondaries has
318 been improved. Instance operations should not fail because one of it's
319 secondary nodes is offline, even though it's safe to proceed.
321 A new command ``list-drbd`` has been added to the ``gnt-node`` script to
322 support debugging of DRBD issues on nodes. It provides a mapping of DRBD
323 minors to instance name.
328 RAPI coverage has improved, with (for example) new resources for
329 recreate-disks, node power-cycle, etc.
334 There is partial support for ``xl`` in the Xen hypervisor; feedback is
337 Python 2.7 is better supported, and after Ganeti 2.6 we will investigate
338 whether to still support Python 2.4 or move to Python 2.6 as minimum
341 Support for Fedora has been slightly improved; the provided example
342 init.d script should work better on it and the INSTALL file should
343 document the needed dependencies.
348 The deprecated ``QueryLocks`` LUXI request has been removed. Use
349 ``Query(what=QR_LOCK, ...)`` instead.
351 The LUXI requests :pyeval:`luxi.REQ_QUERY_JOBS`,
352 :pyeval:`luxi.REQ_QUERY_INSTANCES`, :pyeval:`luxi.REQ_QUERY_NODES`,
353 :pyeval:`luxi.REQ_QUERY_GROUPS`, :pyeval:`luxi.REQ_QUERY_EXPORTS` and
354 :pyeval:`luxi.REQ_QUERY_TAGS` are deprecated and will be removed in a
355 future version. :pyeval:`luxi.REQ_QUERY` should be used instead.
357 RAPI client: ``CertificateError`` now derives from
358 ``GanetiApiError``. This should make it more easy to handle Ganeti
361 Deprecation warnings due to PyCrypto/paramiko import in
362 ``tools/setup-ssh`` have been silenced, as usually they are safe; please
363 make sure to run an up-to-date paramiko version, if you use this tool.
365 The QA scripts now depend on Python 2.5 or above (the main code base
366 still works with Python 2.4).
368 The configuration file (``config.data``) is now written without
369 indentation for performance reasons; if you want to edit it, it can be
370 re-formatted via ``tools/fmtjson``.
372 A number of bugs has been fixed in the cluster merge tool.
374 ``x509`` certification verification (used in import-export) has been
375 changed to allow the same clock skew as permitted by the cluster
376 verification. This will remove some rare but hard to diagnose errors in
383 *(Released Thu, 19 Jul 2012)*
385 Very few changes from rc4 to the final release, only bugfixes:
387 - integrated fixes from release 2.5.2 (fix general boot flag for KVM
388 instance, fix CDROM booting for KVM instances)
389 - fixed node group modification of node parameters
390 - fixed issue in LUClusterVerifyGroup with multi-group clusters
391 - fixed generation of bash completion to ensure a stable ordering
398 *(Released Fri, 13 Jul 2012)*
400 Third release candidate for 2.6. The following changes were done from
403 - Fixed ``UpgradeConfig`` w.r.t. to disk parameters on disk objects.
404 - Fixed an inconsistency in the LUXI protocol with the provided
405 arguments (NOT backwards compatible)
406 - Fixed a bug with node groups ipolicy where ``min`` was greater than
407 the cluster ``std`` value
408 - Implemented a new ``gnt-node list-drbd`` call to list DRBD minors for
409 easier instance debugging on nodes (requires ``hconfd`` to work)
415 *(Released Tue, 03 Jul 2012)*
417 Second release candidate for 2.6. The following changes were done from
420 - Fixed ``gnt-cluster verify`` regarding ``master-ip-script`` on non
422 - Fixed a RAPI regression on missing beparams/memory
423 - Fixed redistribution of files on offline nodes
424 - Added possibility to run activate-disks even though secondaries are
425 offline. With this change it relaxes also the strictness on some other
426 commands which use activate disks internally:
427 * ``gnt-instance start|reboot|rename|backup|export``
428 - Made it possible to remove safely an instance if its secondaries are
430 - Made it possible to reinstall even though secondaries are offline
436 *(Released Mon, 25 Jun 2012)*
438 First release candidate for 2.6. The following changes were done from
441 - Fixed bugs with disk parameters and ``rbd`` templates as well as
443 - Made ``gnt-instance modify`` more consistent regarding new NIC/Disk
444 behaviour. It supports now the modify operation
445 - ``hcheck`` implemented to analyze cluster health and possibility of
446 improving health by rebalance
447 - ``hbal`` has been improved in dealing with split instances
453 *(Released Mon, 11 Jun 2012)*
455 Second beta release of 2.6. The following changes were done from beta2
458 - Fixed ``daemon-util`` with non-root user models
459 - Fixed creation of plain instances with ``--no-wait-for-sync``
460 - Fix wrong iv_names when running ``cfgupgrade``
461 - Export more information in RAPI group queries
462 - Fixed bug when changing instance network interfaces
463 - Extended burnin to do NIC changes
464 - query: Added ``<``, ``>``, ``<=``, ``>=`` comparison operators
465 - Changed default for DRBD barriers
466 - Fixed DRBD error reporting for syncer rate
467 - Verify the options on disk parameters
469 And of course various fixes to documentation and improved unittests and
476 *(Released Wed, 23 May 2012)*
478 First beta release of 2.6. The following changes were done from beta1 to
481 - integrated patch for distributions without ``start-stop-daemon``
482 - adapted example init.d script to work on Fedora
483 - fixed log handling in Haskell daemons
484 - adapted checks in the watcher for pycurl linked against libnss
485 - add partial support for ``xl`` instead of ``xm`` for Xen
486 - fixed a type issue in cluster verification
487 - fixed ssconf handling in the Haskell code (was breaking confd in IPv6
490 Plus integrated fixes from the 2.5 branch:
492 - fixed ``kvm-ifup`` to use ``/bin/bash``
493 - fixed parallel build failures
494 - KVM live migration when using a custom keymap
500 *(Released Tue, 24 Jul 2012)*
502 A small bugfix release, with no new features:
504 - fixed bash-isms in kvm-ifup, for compatibility with systems which use a
505 different default shell (e.g. Debian, Ubuntu)
506 - fixed KVM startup and live migration with a custom keymap (fixes Issue
507 243 and Debian bug #650664)
508 - fixed compatibility with KVM versions that don't support multiple boot
509 devices (fixes Issue 230 and Debian bug #624256)
511 Additionally, a few fixes were done to the build system (fixed parallel
512 build failures) and to the unittests (fixed race condition in test for
513 FileID functions, and the default enable/disable mode for QA test is now
520 *(Released Fri, 11 May 2012)*
522 A small bugfix release.
524 The main issues solved are on the topic of compatibility with newer LVM
527 - fixed parsing of ``lv_attr`` field
528 - adapted to new ``vgreduce --removemissing`` behaviour where sometimes
529 the ``--force`` flag is needed
531 Also on the topic of compatibility, ``tools/lvmstrap`` has been changed
532 to accept kernel 3.x too (was hardcoded to 2.6.*).
534 A regression present in 2.5.0 that broke handling (in the gnt-* scripts)
535 of hook results and that also made display of other errors suboptimal
536 was fixed; the code behaves now like 2.4 and earlier.
538 Another change in 2.5, the cleanup of the OS scripts environment, is too
539 aggressive: it removed even the ``PATH`` variable, which requires the OS
540 scripts to *always* need to export it. Since this is a bit too strict,
541 we now export a minimal PATH, the same that we export for hooks.
543 The fix for issue 201 (Preserve bridge MTU in KVM ifup script) was
544 integrated into this release.
546 Finally, a few other miscellaneous changes were done (no new features,
547 just small improvements):
549 - Fix ``gnt-group --help`` display
550 - Fix hardcoded Xen kernel path
551 - Fix grow-disk handling of invalid units
552 - Update synopsis for ``gnt-cluster repair-disk-sizes``
553 - Accept both PUT and POST in noded (makes future upgrade to 2.6 easier)
559 *(Released Thu, 12 Apr 2012)*
561 Incompatible/important changes and bugfixes
562 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
564 - The default of the ``/2/instances/[instance_name]/rename`` RAPI
565 resource's ``ip_check`` parameter changed from ``True`` to ``False``
566 to match the underlying LUXI interface.
567 - The ``/2/nodes/[node_name]/evacuate`` RAPI resource was changed to use
568 body parameters, see :doc:`RAPI documentation <rapi>`. The server does
569 not maintain backwards-compatibility as the underlying operation
570 changed in an incompatible way. The RAPI client can talk to old
571 servers, but it needs to be told so as the return value changed.
572 - When creating file-based instances via RAPI, the ``file_driver``
573 parameter no longer defaults to ``loop`` and must be specified.
574 - The deprecated ``bridge`` NIC parameter is no longer supported. Use
576 - Support for the undocumented and deprecated RAPI instance creation
577 request format version 0 has been dropped. Use version 1, supported
578 since Ganeti 2.1.3 and :doc:`documented <rapi>`, instead.
579 - Pyparsing 1.4.6 or above is required, see :doc:`installation
580 documentation <install>`.
581 - The "cluster-verify" hooks are now executed per group by the
582 ``OP_CLUSTER_VERIFY_GROUP`` opcode. This maintains the same behavior
583 if you just run ``gnt-cluster verify``, which generates one opcode per
585 - The environment as passed to the OS scripts is cleared, and thus no
586 environment variables defined in the node daemon's environment will be
587 inherited by the scripts.
588 - The :doc:`iallocator <iallocator>` mode ``multi-evacuate`` has been
590 - :doc:`New iallocator modes <design-multi-reloc>` have been added to
591 support operations involving multiple node groups.
592 - Offline nodes are ignored when failing over an instance.
593 - Support for KVM version 1.0, which changed the version reporting format
595 - TCP/IP ports used by DRBD disks are returned to a pool upon instance
597 - ``Makefile`` is now compatible with Automake 1.11.2
598 - Includes all bugfixes made in the 2.4 series
603 - The ganeti-htools project has been merged into the ganeti-core source
604 tree and will be built as part of Ganeti (see :doc:`install-quick`).
605 - Implemented support for :doc:`shared storage <design-shared-storage>`.
606 - Add support for disks larger than 2 TB in ``lvmstrap`` by supporting
607 GPT-style partition tables (requires `parted
608 <http://www.gnu.org/s/parted/>`_).
609 - Added support for floppy drive and 2nd CD-ROM drive in KVM hypervisor.
610 - Allowed adding tags on instance creation.
611 - Export instance tags to hooks (``INSTANCE_TAGS``, see :doc:`hooks`)
612 - Allow instances to be started in a paused state, enabling the user to
613 see the complete console output on boot using the console.
614 - Added new hypervisor flag to control default reboot behaviour
615 (``reboot_behavior``).
616 - Added support for KVM keymaps (hypervisor parameter ``keymap``).
617 - Improved out-of-band management support:
619 - Added ``gnt-node health`` command reporting the health status of
621 - Added ``gnt-node power`` command to manage power status of nodes.
622 - Added command for emergency power-off (EPO), ``gnt-cluster epo``.
624 - Instance migration can fall back to failover if instance is not
626 - Filters can be used when listing nodes, instances, groups and locks;
627 see *ganeti(7)* manpage.
628 - Added post-execution status as variables to :doc:`hooks <hooks>`
630 - Instance tags are exported/imported together with the instance.
631 - When given an explicit job ID, ``gnt-job info`` will work for archived
633 - Jobs can define dependencies on other jobs (not yet supported via
634 RAPI or command line, but used by internal commands and usable via
637 - Lock monitor (``gnt-debug locks``) shows jobs waiting for
640 - Instance failover is now available as a RAPI resource
641 (``/2/instances/[instance_name]/failover``).
642 - ``gnt-instance info`` defaults to static information if primary node
644 - Opcodes have a new ``comment`` attribute.
645 - Added basic SPICE support to KVM hypervisor.
646 - ``tools/ganeti-listrunner`` allows passing of arguments to executable.
648 Node group improvements
649 ~~~~~~~~~~~~~~~~~~~~~~~
651 - ``gnt-cluster verify`` has been modified to check groups separately,
652 thereby improving performance.
653 - Node group support has been added to ``gnt-cluster verify-disks``,
654 which now operates per node group.
655 - Watcher has been changed to work better with node groups.
657 - One process and state file per node group.
658 - Slow watcher in one group doesn't block other group's watcher.
660 - Added new command, ``gnt-group evacuate``, to move all instances in a
661 node group to other groups.
662 - Added ``gnt-instance change-group`` to move an instance to another
664 - ``gnt-cluster command`` and ``gnt-cluster copyfile`` now support
665 per-group operations.
666 - Node groups can be tagged.
667 - Some operations switch from an exclusive to a shared lock as soon as
669 - Instance's primary and secondary nodes' groups are now available as
670 query fields (``pnode.group``, ``pnode.group.uuid``, ``snodes.group``
671 and ``snodes.group.uuid``).
676 - Numerous updates to documentation and manpages.
678 - :doc:`RAPI <rapi>` documentation now has detailed parameter
680 - Some opcode/job results are now also documented, see :doc:`RAPI
683 - A lockset's internal lock is now also visible in lock monitor.
684 - Log messages from job queue workers now contain information about the
685 opcode they're processing.
686 - ``gnt-instance console`` no longer requires the instance lock.
687 - A short delay when waiting for job changes reduces the number of LUXI
688 requests significantly.
689 - DRBD metadata volumes are overwritten with zeros during disk creation.
690 - Out-of-band commands no longer acquire the cluster lock in exclusive
692 - ``devel/upload`` now uses correct permissions for directories.
698 *(Released Fri, 23 Mar 2012)*
700 This was the sixth release candidate of the 2.5 series.
706 *(Released Mon, 9 Jan 2012)*
708 This was the fifth release candidate of the 2.5 series.
714 *(Released Thu, 27 Oct 2011)*
716 This was the fourth release candidate of the 2.5 series.
722 *(Released Wed, 26 Oct 2011)*
724 This was the third release candidate of the 2.5 series.
730 *(Released Tue, 18 Oct 2011)*
732 This was the second release candidate of the 2.5 series.
738 *(Released Tue, 4 Oct 2011)*
740 This was the first release candidate of the 2.5 series.
746 *(Released Wed, 31 Aug 2011)*
748 This was the third beta release of the 2.5 series.
754 *(Released Mon, 22 Aug 2011)*
756 This was the second beta release of the 2.5 series.
762 *(Released Fri, 12 Aug 2011)*
764 This was the first beta release of the 2.5 series.
770 *(Released Thu, 27 Oct 2011)*
772 - Fixed bug when parsing command line parameter values ending in
774 - Fixed assertion error after unclean master shutdown
775 - Disable HTTP client pool for RPC, significantly reducing memory usage
777 - Fixed queue archive creation with wrong permissions
783 *(Released Tue, 23 Aug 2011)*
787 - Fixed documentation for importing with ``--src-dir`` option
788 - Fixed a bug in ``ensure-dirs`` with queue/archive permissions
789 - Fixed a parsing issue with DRBD 8.3.11 in the Linux kernel
795 *(Released Fri, 5 Aug 2011)*
797 Many bug-fixes and a few small features:
799 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
800 issues when you tried to add an instance with two MAC addresses in one
802 - KVM: fixed per-instance stored UID value
803 - KVM: configure bridged NICs at migration start
804 - KVM: Fix a bug where instance will not start with never KVM versions
806 - Added OS search path to ``gnt-cluster info``
807 - Fixed an issue with ``file_storage_dir`` where you were forced to
808 provide an absolute path, but the documentation states it is a
809 relative path, the documentation was right
810 - Added a new parameter to instance stop/start called ``--no-remember``
811 that will make the state change to not be remembered
812 - Implemented ``no_remember`` at RAPI level
813 - Improved the documentation
814 - Node evacuation: don't call IAllocator if node is already empty
815 - Fixed bug in DRBD8 replace disks on current nodes
816 - Fixed bug in recreate-disks for DRBD instances
817 - Moved assertion checking locks in ``gnt-instance replace-disks``
818 causing it to abort with not owning the right locks for some situation
819 - Job queue: Fixed potential race condition when cancelling queued jobs
820 - Fixed off-by-one bug in job serial generation
821 - ``gnt-node volumes``: Fix instance names
822 - Fixed aliases in bash completion
823 - Fixed a bug in reopening log files after being sent a SIGHUP
824 - Added a flag to burnin to allow specifying VCPU count
825 - Bugfixes to non-root Ganeti configuration
831 *(Released Thu, 12 May 2011)*
833 Many bug-fixes and a few new small features:
835 - Fixed a bug related to log opening failures
836 - Fixed a bug in instance listing with orphan instances
837 - Fixed a bug which prevented resetting the cluster-level node parameter
838 ``oob_program`` to the default
839 - Many fixes related to the ``cluster-merge`` tool
840 - Fixed a race condition in the lock monitor, which caused failures
841 during (at least) creation of many instances in parallel
842 - Improved output for gnt-job info
843 - Removed the quiet flag on some ssh calls which prevented debugging
845 - Improved the N+1 failure messages in cluster verify by actually
846 showing the memory values (needed and available)
847 - Increased lock attempt timeouts so that when executing long operations
848 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
849 too early and thus prevent the use of the 'fair' mechanism
850 - Changed instance query data (``gnt-instance info``) to not acquire
851 locks unless needed, thus allowing its use on locked instance if only
852 static information is asked for
853 - Improved behaviour with filesystems that do not support rename on an
855 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
856 kept locks on all nodes during the wipe, which is unneeded
857 - Fixed ``gnt-watcher`` handling of errors during hooks execution
858 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
859 10GiB) which caused the wipe to fail right at the end in some cases
860 - Fixed master IP activation when doing master failover with no-voting
861 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
862 the master node itself
863 - Fixed potential data-loss in under disk full conditions, where Ganeti
864 wouldn't check correctly the return code and would consider
865 partially-written files 'correct'
866 - Fixed bug related to multiple VGs and DRBD disk replacing
867 - Added new disk parameter ``metavg`` that allows placement of the meta
868 device for DRBD in a different volume group
869 - Fixed error handling in the node daemon when the system libc doesn't
870 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
871 - Fixed lock release during replace-disks, which kept cluster-wide locks
872 when doing disk replaces with an iallocator script
873 - Added check for missing bridges in cluster verify
874 - Handle EPIPE errors while writing to the terminal better, so that
875 piping the output to e.g. ``less`` doesn't cause a backtrace
876 - Fixed rare case where a ^C during Luxi calls could have been
877 interpreted as server errors, instead of simply terminating
878 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
880 - Added a few more parameters to the KVM hypervisor, allowing a second
881 CDROM, custom disk type for CDROMs and a floppy image
882 - Removed redundant message in instance rename when the name is given
884 - Added option to ``gnt-instance recreate-disks`` to allow creating the
885 disks on new nodes, allowing recreation when the original instance
886 nodes are completely gone
887 - Added option when converting disk templates to DRBD to skip waiting
888 for the resync, in order to make the instance available sooner
889 - Added two new variables to the OS scripts environment (containing the
891 - Made the root_path and optional parameter for the xen-pvm hypervisor,
892 to allow use of ``pvgrub`` as bootloader
893 - Changed the instance memory modifications to only check out-of-memory
894 conditions on memory increases, and turned the secondary node warnings
895 into errors (they can still be overridden via ``--force``)
896 - Fixed the handling of a corner case when the Python installation gets
897 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
898 to execute a command that doesn't exist
899 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
900 node of the instance returned failures during instance shutdown; this
901 adds the option ``--ignore-consistency`` to gnt-instance move
903 And as usual, various improvements to the error messages, documentation
910 *(Released Wed, 09 Mar 2011)*
912 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
913 the RAPI users file if run twice (even with ``--dry-run``).
915 The release fixes that bug (nothing else changed).
921 *(Released Mon, 07 Mar 2011)*
923 Final 2.4.0 release. Just a few small fixes:
925 - Fixed RAPI node evacuate
926 - Fixed the kvm-ifup script
927 - Fixed internal error handling for special job cases
928 - Updated man page to specify the escaping feature for options
934 *(Released Mon, 28 Feb 2011)*
936 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
937 that this feature wiped the disks of the wrong instance, leading to loss
942 - Fixed title of query field containing instance name
943 - Expanded the glossary in the documentation
944 - Fixed one unittest (internal issue)
950 *(Released Mon, 21 Feb 2011)*
952 A number of bug fixes plus just a couple functionality changes.
954 On the user-visible side, the ``gnt-* list`` command output has changed
955 with respect to "special" field states. The current rc1 style of display
956 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
957 the default output mode special fields states are displayed as follows:
959 - Offline resource: ``*``
960 - Unavailable/not applicable: ``-``
961 - Data missing (RPC failure): ``?``
962 - Unknown field: ``??``
964 Another user-visible change is the addition of ``--force-join`` to
969 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
970 - Fixed regression in RAPI/instance reinstall where all parameters were
971 required (instead of optional)
972 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
973 - Fixed iallocator usage (offline nodes were not considered offline)
974 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
975 - Fixed hypervisor and OS parameter validation with respect to
977 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
979 - Fixed ``tools/listrunner`` with respect to agent-based usage
985 *(Released Fri, 4 Feb 2011)*
987 Many changes and fixes since the beta1 release. While there were some
988 internal changes, the code has been mostly stabilised for the RC
991 Note: the dumb allocator was removed in this release, as it was not kept
992 up-to-date with the IAllocator protocol changes. It is recommended to
993 use the ``hail`` command from the ganeti-htools package.
995 Note: the 2.4 and up versions of Ganeti are not compatible with the
996 0.2.x branch of ganeti-htools. You need to upgrade to
997 ganeti-htools-0.3.0 (or later).
999 Regressions fixed from 2.3
1000 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1002 - Fixed the ``gnt-cluster verify-disks`` command
1003 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
1005 - Fixed disk adoption breakage
1006 - Fixed wrong headers in instance listing for field aliases
1011 - Fixed corner case in KVM handling of NICs
1012 - Fixed many cases of wrong handling of non-vm_capable nodes
1013 - Fixed a bug where a missing instance symlink was not possible to
1014 recreate with any ``gnt-*`` command (now ``gnt-instance
1015 activate-disks`` does it)
1016 - Fixed the volume group name as reported by ``gnt-cluster
1018 - Increased timeouts for the import-export code, hopefully leading to
1019 fewer aborts due network or instance timeouts
1020 - Fixed bug in ``gnt-node list-storage``
1021 - Fixed bug where not all daemons were started on cluster
1022 initialisation, but only at the first watcher run
1023 - Fixed many bugs in the OOB implementation
1024 - Fixed watcher behaviour in presence of instances with offline
1026 - Fixed instance list output for instances running on the wrong node
1027 - a few fixes to the cluster-merge tool, but it still cannot merge
1028 multi-node groups (currently it is not recommended to use this tool)
1034 - Improved network configuration for the KVM hypervisor
1035 - Added e1000 as a supported NIC for Xen-HVM
1036 - Improved the lvmstrap tool to also be able to use partitions, as
1037 opposed to full disks
1038 - Improved speed of disk wiping (the cluster parameter
1039 ``prealloc_wipe_disks``, so that it has a low impact on the total time
1040 of instance creations
1041 - Added documentation for the OS parameters
1042 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
1043 hypervisor is not responding
1044 - Added display of blacklisted and hidden OS information in
1045 ``gnt-cluster info``
1046 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
1047 NIC and node parameters, which might create problems with currently
1048 invalid (but undetected) configuration files, but prevents validation
1049 failures when unrelated parameters are modified
1050 - Changed cluster initialisation to wait for the master daemon to become
1052 - Expanded the RAPI interface:
1054 - Added config redistribution resource
1055 - Added activation/deactivation of instance disks
1056 - Added export of console information
1058 - Implemented log file reopening on SIGHUP, which allows using
1059 logrotate(8) for the Ganeti log files
1060 - Added a basic OOB helper script as an example
1066 *(Released Fri, 14 Jan 2011)*
1071 - Fixed timezone issues when formatting timestamps
1072 - Added support for node groups, available via ``gnt-group`` and other
1074 - Added out-of-band framework and management, see :doc:`design
1075 document <design-oob>`
1076 - Removed support for roman numbers from ``gnt-node list`` and
1077 ``gnt-instance list``.
1078 - Allowed modification of master network interface via ``gnt-cluster
1079 modify --master-netdev``
1080 - Accept offline secondaries while shutting down instance disks
1081 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
1082 - Added support for multiple LVM volume groups
1083 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
1085 - Added commands to list available fields:
1087 - ``gnt-node list-fields``
1088 - ``gnt-group list-fields``
1089 - ``gnt-instance list-fields``
1091 - Updated documentation and man pages
1096 - Moved ``rapi_users`` file into separate directory, now named
1097 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
1099 - Added new tool for running commands on many machines,
1100 ``tools/ganeti-listrunner``
1101 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
1102 improving the ``gnt-instance console`` output
1103 - Allowed customisation of disk index separator at ``configure`` time
1104 - Export node group allocation policy to :doc:`iallocator <iallocator>`
1105 - Added support for non-partitioned md disks in ``lvmstrap``
1106 - Added script to gracefully power off KVM instances
1107 - Split ``utils`` module into smaller parts
1108 - Changed query operations to return more detailed information, e.g.
1109 whether an information is unavailable due to an offline node. To use
1110 this new functionality, the LUXI call ``Query`` must be used. Field
1111 information is now stored by the master daemon and can be retrieved
1112 using ``QueryFields``. Instances, nodes and groups can also be queried
1113 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
1114 exposed via RAPI). The following commands make use of this
1115 infrastructure change:
1117 - ``gnt-group list``
1118 - ``gnt-group list-fields``
1120 - ``gnt-node list-fields``
1121 - ``gnt-instance list``
1122 - ``gnt-instance list-fields``
1123 - ``gnt-debug locks``
1128 - New RAPI resources (see :doc:`rapi`):
1132 - ``/2/groups/[group_name]``
1133 - ``/2/groups/[group_name]/assign-nodes``
1134 - ``/2/groups/[group_name]/modify``
1135 - ``/2/groups/[group_name]/rename``
1136 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
1140 - Implemented ``no_install`` for instance creation
1141 - Implemented OS parameters for instance reinstallation, allowing
1142 use of special settings on reinstallation (e.g. for preserving data)
1147 - Added IPv6 support in import/export
1148 - Pause DRBD synchronization while wiping disks on instance creation
1149 - Updated unittests and QA scripts
1150 - Improved network parameters passed to KVM
1151 - Converted man pages from docbook to reStructuredText
1157 *(Released Mon, 20 Dec 2010)*
1159 Released version 2.3.1~rc1 without any changes.
1165 *(Released Wed, 1 Dec 2010)*
1167 - impexpd: Disable OpenSSL compression in socat if possible (backport
1168 from master, commit e90739d625b, see :doc:`installation guide
1169 <install-quick>` for details)
1170 - Changed unittest coverage report to exclude test scripts
1171 - Added script to check version format
1177 *(Released Wed, 1 Dec 2010)*
1179 Released version 2.3.0~rc1 without any changes.
1185 *(Released Fri, 19 Nov 2010)*
1187 A number of bugfixes and documentation updates:
1189 - Update ganeti-os-interface documentation
1190 - Fixed a bug related to duplicate MACs or similar items which should be
1192 - Fix breakage in OS state modify
1193 - Reinstall instance: disallow offline secondaries (fixes bug related to
1194 OS changing but reinstall failing)
1195 - plus all the other fixes between 2.2.1 and 2.2.2
1201 *(Released Tue, 2 Nov 2010)*
1203 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
1204 - Fixed master failover race with watcher
1205 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
1207 - Accept previously stopped instance for export with instance removal
1208 - Simplify and extend the environment variables for instance OS scripts
1209 - Added new node flags, ``master_capable`` and ``vm_capable``
1210 - Added optional instance disk wiping prior during allocation. This is a
1211 cluster-wide option and can be set/modified using
1212 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
1213 - Added IPv6 support, see :doc:`design document <design-2.3>` and
1214 :doc:`install-quick`
1215 - Added a new watcher option (``--ignore-pause``)
1216 - Added option to ignore offline node on instance start/stop
1217 (``--ignore-offline``)
1218 - Allow overriding OS parameters with ``gnt-instance reinstall``
1219 - Added ability to change node's secondary IP address using ``gnt-node
1221 - Implemented privilege separation for all daemons except
1222 ``ganeti-noded``, see ``configure`` options
1223 - Complain if an instance's disk is marked faulty in ``gnt-cluster
1225 - Implemented job priorities (see ``ganeti(7)`` manpage)
1226 - Ignore failures while shutting down instances during failover from
1228 - Exit daemon's bootstrap process only once daemon is ready
1229 - Export more information via ``LUInstanceQuery``/remote API
1230 - Improved documentation, QA and unittests
1231 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
1232 restart if the file was created or changed
1233 - Added LUXI protocol version sent with each request and response,
1234 allowing detection of server/client mismatches
1235 - Moved the Python scripts among gnt-* and ganeti-* into modules
1236 - Moved all code related to setting up SSH to an external script,
1238 - Infrastructure changes for node group support in future versions
1244 *(Released Fri, 19 Nov 2010)*
1246 A few small bugs fixed, and some improvements to the build system:
1248 - Fix documentation regarding conversion to drbd
1249 - Fix validation of parameters in cluster modify (``gnt-cluster modify
1251 - Fix error handling in node modify with multiple changes
1252 - Allow remote imports without checked names
1258 *(Released Tue, 19 Oct 2010)*
1260 - Disable SSL session ID cache in RPC client
1266 *(Released Thu, 14 Oct 2010)*
1268 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
1269 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
1271 - Fix problems with interaction between Curl and Python's HTTP server,
1272 resulting in increased speed in many RPC calls
1273 - Improve our release script to prevent breakage with older aclocal and
1280 *(Released Thu, 7 Oct 2010)*
1282 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
1283 value retrieved from master
1284 - Added support for blacklisted or hidden OS definitions
1285 - Added simple lock monitor (accessible via (``gnt-debug locks``)
1286 - Added support for -mem-path in KVM hypervisor abstraction layer
1287 - Allow overriding instance parameters in tool for inter-cluster
1288 instance moves (``tools/move-instance``)
1289 - Improved opcode summaries (e.g. in ``gnt-job list``)
1290 - Improve consistency of OS listing by sorting it
1291 - Documentation updates
1297 *(Released Fri, 8 Oct 2010)*
1299 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
1305 *(Released Mon, 4 Oct 2010)*
1307 - Fixed regression in ``gnt-instance rename``
1313 *(Released Wed, 22 Sep 2010)*
1315 - Fixed OS_VARIANT variable for OS scripts
1316 - Fixed cluster tag operations via RAPI
1317 - Made ``setup-ssh`` exit with non-zero code if an error occurred
1318 - Disabled RAPI CA checks in watcher
1324 *(Released Mon, 23 Aug 2010)*
1326 - Support DRBD versions of the format "a.b.c.d"
1328 - Re-introduce support for usage from multiple threads in RAPI client
1329 - Instance renames and modify via RAPI
1330 - Work around race condition between processing and archival in job
1332 - Mark opcodes following failed one as failed, too
1333 - Job field ``lock_status`` was removed due to difficulties making it
1334 work with the changed job queue in Ganeti 2.2; a better way to monitor
1335 locks is expected for a later 2.2.x release
1336 - Fixed dry-run behaviour with many commands
1337 - Support ``ssh-agent`` again when adding nodes
1338 - Many additional bugfixes
1344 *(Released Fri, 30 Jul 2010)*
1346 Important change: the internal RPC mechanism between Ganeti nodes has
1347 changed from using a home-grown http library (based on the Python base
1348 libraries) to use the PycURL library. This requires that PycURL is
1349 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
1350 against GnuTLS by default. cURL's support for GnuTLS had known issues
1351 before cURL 7.21.0 and we recommend using the latest cURL release or
1352 linking against OpenSSL. Most other distributions already link PycURL
1353 and cURL against OpenSSL. The command::
1355 python -c 'import pycurl; print pycurl.version'
1357 can be used to determine the libraries PycURL and cURL are linked
1360 Other significant changes:
1362 - Rewrote much of the internals of the job queue, in order to achieve
1363 better parallelism; this decouples job query operations from the job
1364 processing, and it should allow much nicer behaviour of the master
1365 daemon under load, and it also has uncovered some long-standing bugs
1366 related to the job serialisation (now fixed)
1367 - Added a default iallocator setting to the cluster parameters,
1368 eliminating the need to always pass nodes or an iallocator for
1369 operations that require selection of new node(s)
1370 - Added experimental support for the LXC virtualization method
1371 - Added support for OS parameters, which allows the installation of
1372 instances to pass parameter to OS scripts in order to customise the
1374 - Added a hypervisor parameter controlling the migration type (live or
1375 non-live), since hypervisors have various levels of reliability; this
1376 has renamed the 'live' parameter to 'mode'
1377 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
1378 logical volumes, meaning that cluster verify will ignore them and not
1379 flag their presence as errors
1380 - The watcher will now reset the error count for failed instances after
1381 8 hours, thus allowing self-healing if the problem that caused the
1382 instances to be down/fail to start has cleared in the meantime
1383 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
1384 check for, and warn, if the drbd module parameter ``usermode_helper``
1385 is not consistent with the cluster-wide setting; this is needed to
1386 make diagnose easier of failed drbd creations
1387 - Started adding base IPv6 support, but this is not yet
1388 enabled/available for use
1389 - Rename operations (cluster, instance) will now return the new name,
1390 which is especially useful if a short name was passed in
1391 - Added support for instance migration in RAPI
1392 - Added a tool to pre-configure nodes for the SSH setup, before joining
1393 them to the cluster; this will allow in the future a simplified model
1394 for node joining (but not yet fully enabled in 2.2); this needs the
1395 paramiko python library
1396 - Fixed handling of name-resolving errors
1397 - Fixed consistency of job results on the error path
1398 - Fixed master-failover race condition when executed multiple times in
1400 - Fixed many bugs related to the job queue (mostly introduced during the
1401 2.2 development cycle, so not all are impacting 2.1)
1402 - Fixed instance migration with missing disk symlinks
1403 - Fixed handling of unknown jobs in ``gnt-job archive``
1404 - And many other small fixes/improvements
1408 - Enhanced both the unittest and the QA coverage
1409 - Switched the opcode validation to a generic model, and extended the
1410 validation to all opcode parameters
1411 - Changed more parts of the code that write shell scripts to use the
1413 - Switched the master daemon to use the asyncore library for the Luxi
1420 *(Released Thu, 17 Jun 2010)*
1422 - Added tool (``move-instance``) and infrastructure to move instances
1423 between separate clusters (see :doc:`separate documentation
1424 <move-instance>` and :doc:`design document <design-2.2>`)
1425 - Added per-request RPC timeout
1426 - RAPI now requires a Content-Type header for requests with a body (e.g.
1427 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
1428 :rfc:`2616` (HTTP/1.1), section 7.2.1)
1429 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
1431 - Implemented initial support for running Ganeti daemons as separate
1432 users, see configure-time flags ``--with-user-prefix`` and
1433 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
1435 - Instances can be removed after export (``gnt-backup export
1436 --remove-instance``)
1437 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
1438 key (instead of 1024 bit)
1439 - Added new cluster configuration file for cluster domain secret
1440 - Import/export now use SSL instead of SSH
1441 - Added support for showing estimated time when exporting an instance,
1442 see the ``ganeti-os-interface(7)`` manpage and look for
1449 *(Released Tue, 16 Nov 2010)*
1451 Some more bugfixes. Unless critical bugs occur, this will be the last
1454 - Fix case of MAC special-values
1455 - Fix mac checker regex
1456 - backend: Fix typo causing "out of range" error
1457 - Add missing --units in gnt-instance list man page
1463 *(Released Tue, 24 Aug 2010)*
1466 - Don't ignore secondary node silently on non-mirrored disk templates
1468 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
1469 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
1470 - Properly document the usb_mouse parameter
1471 - Fix path in ganeti-rapi(8) (issue 116)
1472 - Adjust error message when the ganeti user's .ssh directory is
1474 - Add same-node-check when changing the disk template to drbd
1480 *(Released Fri, 16 Jul 2010)*
1483 - Add an option to only select some reboot types during qa/burnin.
1484 (on some hypervisors consequent reboots are not supported)
1485 - Fix infrequent race condition in master failover. Sometimes the old
1486 master ip address would be still detected as up for a short time
1487 after it was removed, causing failover to fail.
1488 - Decrease mlockall warnings when the ctypes module is missing. On
1489 Python 2.4 we support running even if no ctypes module is installed,
1490 but we were too verbose about this issue.
1491 - Fix building on old distributions, on which man doesn't have a
1493 - Fix RAPI not to ignore the MAC address on instance creation
1494 - Implement the old instance creation format in the RAPI client.
1500 *(Released Thu, 01 Jul 2010)*
1502 A small bugfix release:
1503 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
1504 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
1505 a non-existing option
1506 - Fix instance create: the --force-variant option was ignored
1507 - Improve pylint 0.21 compatibility and warnings with Python 2.6
1508 - Fix modify node storage with non-FQDN arguments
1509 - Fix RAPI client to authenticate under Python 2.6 when used
1510 for more than 5 requests needing authentication
1511 - Fix gnt-instance modify -t (storage) giving a wrong error message
1512 when converting a non-shutdown drbd instance to plain
1518 *(Released Fri, 18 Jun 2010)*
1520 A small bugfix release:
1522 - Fix live migration of KVM instances started with older Ganeti
1523 versions which had fewer hypervisor parameters
1524 - Fix gnt-instance grow-disk on down instances
1525 - Fix an error-reporting bug during instance migration
1526 - Better checking of the ``--net`` and ``--disk`` values, to avoid
1527 silently ignoring broken ones
1528 - Fix an RPC error reporting bug affecting, for example, RAPI client
1530 - Fix bug triggered by different API version os-es on different nodes
1531 - Fix a bug in instance startup with custom hvparams: OS level
1532 parameters would fail to be applied.
1533 - Fix the RAPI client under Python 2.6 (but more work is needed to
1534 make it work completely well with OpenSSL)
1535 - Fix handling of errors when resolving names from DNS
1541 *(Released Thu, 3 Jun 2010)*
1543 A medium sized development cycle. Some new features, and some
1544 fixes/small improvements/cleanups.
1546 Significant features
1547 ~~~~~~~~~~~~~~~~~~~~
1549 The node deamon now tries to mlock itself into memory, unless the
1550 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
1551 its logs, and falls back to console logging. This allows emergency
1552 features such as ``gnt-node powercycle`` to work even in the event of a
1553 broken node disk (tested offlining the disk hosting the node's
1554 filesystem and dropping its memory caches; don't try this at home)
1556 KVM: add vhost-net acceleration support. It can be tested with a new
1557 enough version of the kernel and of qemu-kvm.
1559 KVM: Add instance chrooting feature. If you use privilege dropping for
1560 your VMs you can also now force them to chroot to an empty directory,
1561 before starting the emulated guest.
1563 KVM: Add maximum migration bandwith and maximum downtime tweaking
1564 support (requires a new-enough version of qemu-kvm).
1566 Cluster verify will now warn if the master node doesn't have the master
1567 ip configured on it.
1569 Add a new (incompatible) instance creation request format to RAPI which
1570 supports all parameters (previously only a subset was supported, and it
1571 wasn't possible to extend the old format to accomodate all the new
1572 features. The old format is still supported, and a client can check for
1573 this feature, before using it, by checking for its presence in the
1574 ``features`` RAPI resource.
1576 Now with ancient latin support. Try it passing the ``--roman`` option to
1577 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
1578 (requires the python-roman module to be installed, in order to work).
1583 As usual many internal code refactorings, documentation updates, and
1586 - Lots of improvements and cleanups to the experimental Remote API
1587 (RAPI) client library.
1588 - A new unit test suite for the core daemon libraries.
1589 - A fix to creating missing directories makes sure the umask is not
1590 applied anymore. This enforces the same directory permissions
1592 - Better handling terminating daemons with ctrl+c (used when running
1593 them in debugging mode).
1594 - Fix a race condition in live migrating a KVM instance, when stat()
1595 on the old proc status file returned EINVAL, which is an unexpected
1597 - Fixed manpage checking with newer man and utf-8 charachters. But now
1598 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
1604 *(Released Fri, 7 May 2010)*
1606 Fix a bug which prevented untagged KVM instances from starting.
1612 *(Released Fri, 7 May 2010)*
1614 Another release with a long development cycle, during which many
1615 different features were added.
1617 Significant features
1618 ~~~~~~~~~~~~~~~~~~~~
1620 The KVM hypervisor now can run the individual instances as non-root, to
1621 reduce the impact of a VM being hijacked due to bugs in the
1622 hypervisor. It is possible to run all instances as a single (non-root)
1623 user, to manually specify a user for each instance, or to dynamically
1624 allocate a user out of a cluster-wide pool to each instance, with the
1625 guarantee that no two instances will run under the same user ID on any
1628 An experimental RAPI client library, that can be used standalone
1629 (without the other Ganeti libraries), is provided in the source tree as
1630 ``lib/rapi/client.py``. Note this client might change its interface in
1631 the future, as we iterate on its capabilities.
1633 A new command, ``gnt-cluster renew-crypto`` has been added to easily
1634 replace the cluster's certificates and crypto keys. This might help in
1635 case they have been compromised, or have simply expired.
1637 A new disk option for instance creation has been added that allows one
1638 to "adopt" currently existing logical volumes, with data
1639 preservation. This should allow easier migration to Ganeti from
1640 unmanaged (or managed via other software) instances.
1642 Another disk improvement is the possibility to convert between redundant
1643 (DRBD) and plain (LVM) disk configuration for an instance. This should
1644 allow better scalability (starting with one node and growing the
1645 cluster, or shrinking a two-node cluster to one node).
1647 A new feature that could help with automated node failovers has been
1648 implemented: if a node sees itself as offline (by querying the master
1649 candidates), it will try to shutdown (hard) all instances and any active
1650 DRBD devices. This reduces the risk of duplicate instances if an
1651 external script automatically failovers the instances on such nodes. To
1652 enable this, the cluster parameter ``maintain_node_health`` should be
1653 enabled; in the future this option (per the name) will enable other
1654 automatic maintenance features.
1656 Instance export/import now will reuse the original instance
1657 specifications for all parameters; that means exporting an instance,
1658 deleting it and the importing it back should give an almost identical
1659 instance. Note that the default import behaviour has changed from
1660 before, where it created only one NIC; now it recreates the original
1663 Cluster verify has added a few new checks: SSL certificates validity,
1664 /etc/hosts consistency across the cluster, etc.
1669 As usual, many internal changes were done, documentation fixes,
1672 - Fixed cluster initialization with disabled cluster storage (regression
1673 introduced in 2.1.1)
1674 - File-based storage supports growing the disks
1675 - Fixed behaviour of node role changes
1676 - Fixed cluster verify for some corner cases, plus a general rewrite of
1677 cluster verify to allow future extension with more checks
1678 - Fixed log spamming by watcher and node daemon (regression introduced
1680 - Fixed possible validation issues when changing the list of enabled
1682 - Fixed cleanup of /etc/hosts during node removal
1683 - Fixed RAPI response for invalid methods
1684 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
1685 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
1686 from ide disks, etc.)
1687 - Allow OS changes without re-installation (to record a changed OS
1688 outside of Ganeti, or to allow OS renames)
1689 - Allow instance creation without OS installation (useful for example if
1690 the OS will be installed manually, or restored from a backup not in
1692 - Implemented option to make cluster ``copyfile`` use the replication
1694 - Added list of enabled hypervisors to ssconf (possibly useful for
1696 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
1698 - A partial form of node re-IP is possible via node readd, which now
1699 allows changed node primary IP
1700 - Command line utilities now show an informational message if the job is
1702 - The logs of the master daemon now show the PID/UID/GID of the
1709 *(Released Fri, 12 Mar 2010)*
1711 During the 2.1.0 long release candidate cycle, a lot of improvements and
1712 changes have accumulated with were released later as 2.1.1.
1717 The node evacuate command (``gnt-node evacuate``) was significantly
1718 rewritten, and as such the IAllocator protocol was changed - a new
1719 request type has been added. This unfortunate change during a stable
1720 series is designed to improve performance of node evacuations; on
1721 clusters with more than about five nodes and which are well-balanced,
1722 evacuation should proceed in parallel for all instances of the node
1723 being evacuated. As such, any existing IAllocator scripts need to be
1724 updated, otherwise the above command will fail due to the unknown
1725 request. The provided "dumb" allocator has not been updated; but the
1726 ganeti-htools package supports the new protocol since version 0.2.4.
1728 Another important change is increased validation of node and instance
1729 names. This might create problems in special cases, if invalid host
1730 names are being used.
1732 Also, a new layer of hypervisor parameters has been added, that sits at
1733 OS level between the cluster defaults and the instance ones. This allows
1734 customisation of virtualization parameters depending on the installed
1735 OS. For example instances with OS 'X' may have a different KVM kernel
1736 (or any other parameter) than the cluster defaults. This is intended to
1737 help managing a multiple OSes on the same cluster, without manual
1738 modification of each instance's parameters.
1740 A tool for merging clusters, ``cluster-merge``, has been added in the
1741 tools sub-directory.
1746 - Improved the int/float conversions that should make the code more
1747 robust in face of errors from the node daemons
1748 - Fixed the remove node code in case of internal configuration errors
1749 - Fixed the node daemon behaviour in face of inconsistent queue
1750 directory (e.g. read-only file-system where we can't open the files
1752 - Fixed the behaviour of gnt-node modify for master candidate demotion;
1753 now it either aborts cleanly or, if given the new "auto_promote"
1754 parameter, will automatically promote other nodes as needed
1755 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
1756 completely prevent Ganeti from working
1757 - Fixed bug for instance export when not all disks were successfully
1759 - Fixed behaviour of node add when the new node is slow in starting up
1761 - Fixed handling of signals in the LUXI client, which should improve
1762 behaviour of command-line scripts
1763 - Added checks for invalid node/instance names in the configuration (now
1764 flagged during cluster verify)
1765 - Fixed watcher behaviour for disk activation errors
1766 - Fixed two potentially endless loops in http library, which led to the
1767 RAPI daemon hanging and consuming 100% CPU in some cases
1768 - Fixed bug in RAPI daemon related to hashed passwords
1769 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
1771 - Enhanced compatibility with non-Debian OSes, but not using absolute
1772 path in some commands and allowing customisation of the ssh
1773 configuration directory
1774 - Fixed possible future issue with new Python versions by abiding to the
1775 proper use of ``__slots__`` attribute on classes
1776 - Added checks that should prevent directory traversal attacks
1777 - Many documentation fixes based on feedback from users
1782 - Added an "early_release" more for instance replace disks and node
1783 evacuate, where we release locks earlier and thus allow higher
1784 parallelism within the cluster
1785 - Added watcher hooks, intended to allow the watcher to restart other
1786 daemons (e.g. from the ganeti-nbma project), but they can be used of
1787 course for any other purpose
1788 - Added a compile-time disable for DRBD barriers, to increase
1789 performance if the administrator trusts the power supply or the
1790 storage system to not lose writes
1791 - Added the option of using syslog for logging instead of, or in
1792 addition to, Ganeti's own log files
1793 - Removed boot restriction for paravirtual NICs for KVM, recent versions
1794 can indeed boot from a paravirtual NIC
1795 - Added a generic debug level for many operations; while this is not
1796 used widely yet, it allows one to pass the debug value all the way to
1798 - Enhanced the hooks environment for instance moves (failovers,
1799 migrations) where the primary/secondary nodes changed during the
1800 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
1801 - Enhanced data validations for many user-supplied values; one important
1802 item is the restrictions imposed on instance and node names, which
1803 might reject some (invalid) host names
1804 - Add a configure-time option to disable file-based storage, if it's not
1805 needed; this allows greater security separation between the master
1806 node and the other nodes from the point of view of the inter-node RPC
1808 - Added user notification in interactive tools if job is waiting in the
1809 job queue or trying to acquire locks
1810 - Added log messages when a job is waiting for locks
1811 - Added filtering by node tags in instance operations which admit
1812 multiple instances (start, stop, reboot, reinstall)
1813 - Added a new tool for cluster mergers, ``cluster-merge``
1814 - Parameters from command line which are of the form ``a=b,c=d`` can now
1815 use backslash escapes to pass in values which contain commas,
1816 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
1818 - For KVM, the instance name is the first parameter passed to KVM, so
1819 that it's more visible in the process list
1825 *(Released Tue, 2 Mar 2010)*
1827 Ganeti 2.1 brings many improvements with it. Major changes:
1829 - Added infrastructure to ease automated disk repairs
1830 - Added new daemon to export configuration data in a cheaper way than
1831 using the remote API
1832 - Instance NICs can now be routed instead of being associated with a
1834 - Improved job locking logic to reduce impact of jobs acquiring multiple
1835 locks waiting for other long-running jobs
1837 In-depth implementation details can be found in the Ganeti 2.1 design
1843 - Added chroot hypervisor
1844 - Added more options to xen-hvm hypervisor (``kernel_path`` and
1846 - Added more options to xen-pvm hypervisor (``use_bootloader``,
1847 ``bootloader_path`` and ``bootloader_args``)
1848 - Added the ``use_localtime`` option for the xen-hvm and kvm
1849 hypervisors, and the default value for this has changed to false (in
1850 2.0 xen-hvm always enabled it)
1851 - Added luxi call to submit multiple jobs in one go
1852 - Added cluster initialization option to not modify ``/etc/hosts``
1854 - Added network interface parameters
1855 - Added dry run mode to some LUs
1856 - Added RAPI resources:
1858 - ``/2/instances/[instance_name]/info``
1859 - ``/2/instances/[instance_name]/replace-disks``
1860 - ``/2/nodes/[node_name]/evacuate``
1861 - ``/2/nodes/[node_name]/migrate``
1862 - ``/2/nodes/[node_name]/role``
1863 - ``/2/nodes/[node_name]/storage``
1864 - ``/2/nodes/[node_name]/storage/modify``
1865 - ``/2/nodes/[node_name]/storage/repair``
1867 - Added OpCodes to evacuate or migrate all instances on a node
1868 - Added new command to list storage elements on nodes (``gnt-node
1869 list-storage``) and modify them (``gnt-node modify-storage``)
1870 - Added new ssconf files with master candidate IP address
1871 (``ssconf_master_candidates_ips``), node primary IP address
1872 (``ssconf_node_primary_ips``) and node secondary IP address
1873 (``ssconf_node_secondary_ips``)
1874 - Added ``ganeti-confd`` and a client library to query the Ganeti
1875 configuration via UDP
1876 - Added ability to run hooks after cluster initialization and before
1878 - Added automatic mode for disk replace (``gnt-instance replace-disks
1880 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
1881 after catastrophic data-loss
1882 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
1884 - Added ``gnt-instance move`` command to move instances
1885 - Added ``gnt-cluster watcher`` command to control watcher
1886 - Added ``gnt-node powercycle`` command to powercycle nodes
1887 - Added new job status field ``lock_status``
1888 - Added parseable error codes to cluster verification (``gnt-cluster
1889 verify --error-codes``) and made output less verbose (use
1890 ``--verbose`` to restore previous behaviour)
1891 - Added UUIDs to the main config entities (cluster, nodes, instances)
1892 - Added support for OS variants
1893 - Added support for hashed passwords in the Ganeti remote API users file
1895 - Added option to specify maximum timeout on instance shutdown
1896 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
1897 - Added new helper script to start and stop Ganeti daemons
1898 (``daemon-util``), with the intent to reduce the work necessary to
1899 adjust Ganeti for non-Debian distributions and to start/stop daemons
1901 - Added more unittests
1902 - Fixed critical bug in ganeti-masterd startup
1903 - Removed the configure-time ``kvm-migration-port`` parameter, this is
1904 now customisable at the cluster level for both the KVM and Xen
1905 hypervisors using the new ``migration_port`` parameter
1906 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
1907 reinstalling an instance
1908 - Allowed ``@`` in tag names
1909 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
1910 - Many documentation updates
1911 - Distribute hypervisor files on ``gnt-cluster redist-conf``
1912 - ``gnt-instance reinstall`` can now reinstall multiple instances
1913 - Updated many command line parameters
1914 - Introduced new OS API version 15
1915 - No longer support a default hypervisor
1916 - Treat virtual LVs as inexistent
1917 - Improved job locking logic to reduce lock contention
1918 - Match instance and node names case insensitively
1919 - Reimplemented bash completion script to be more complete
1926 *(Released Thu, 4 Feb 2010)*
1928 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
1929 - Fix a string formatting bug
1930 - Improve safety of the code in some error paths
1931 - Improve data validation in the master of values returned from nodes
1937 *(Released Thu, 17 Dec 2009)*
1939 - Fix security issue due to missing validation of iallocator names; this
1940 allows local and remote execution of arbitrary executables
1941 - Fix failure of gnt-node list during instance removal
1942 - Ship the RAPI documentation in the archive
1948 *(Released Wed, 30 Sep 2009)*
1950 - Fixed many wrong messages
1951 - Fixed a few bugs related to the locking library
1952 - Fixed MAC checking at instance creation time
1953 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
1954 - Fixed a few issues related to signal handling in both daemons and
1956 - Fixed the example startup script provided
1957 - Fixed insserv dependencies in the example startup script (patch from
1959 - Fixed handling of drained nodes in the iallocator framework
1960 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
1962 - Fixed error related to invalid job IDs in job polling
1963 - Fixed job/opcode persistence on unclean master shutdown
1964 - Fixed handling of partial job processing after unclean master
1966 - Fixed error reporting from LUs, previously all errors were converted
1967 into execution errors
1968 - Fixed error reporting from burnin
1969 - Decreased significantly the memory usage of the job queue
1970 - Optimised slightly multi-job submission
1971 - Optimised slightly opcode loading
1972 - Backported the multi-job submit framework from the development
1973 branch; multi-instance start and stop should be faster
1974 - Added script to clean archived jobs after 21 days; this will reduce
1975 the size of the queue directory
1976 - Added some extra checks in disk size tracking
1977 - Added an example ethers hook script
1978 - Added a cluster parameter that prevents Ganeti from modifying of
1980 - Added more node information to RAPI responses
1981 - Added a ``gnt-job watch`` command that allows following the ouput of a
1983 - Added a bind-address option to ganeti-rapi
1984 - Added more checks to the configuration verify
1985 - Enhanced the burnin script such that some operations can be retried
1987 - Converted instance reinstall to multi-instance model
1993 *(Released Fri, 7 Aug 2009)*
1995 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
1996 to allow using the pre-2.0.2 behaviour in activation, if any existing
1997 instances have mismatched disk sizes in the configuration
1998 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
1999 any configuration mismatches for disk sizes
2000 - Added ``gnt-master cluste-failover --no-voting`` to allow master
2001 failover to work on two-node clusters
2002 - Fixed the ``--net`` option of ``gnt-backup import``, which was
2004 - Fixed detection of OS script errors in ``gnt-backup export``
2005 - Fixed exit code of ``gnt-backup export``
2011 *(Released Fri, 17 Jul 2009)*
2013 - Added experimental support for stripped logical volumes; this should
2014 enhance performance but comes with a higher complexity in the block
2015 device handling; stripping is only enabled when passing
2016 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
2017 affected even in the non-stripped mode
2018 - Improved resiliency against transient failures at the end of DRBD
2019 resyncs, and in general of DRBD resync checks
2020 - Fixed a couple of issues with exports and snapshot errors
2021 - Fixed a couple of issues in instance listing
2022 - Added display of the disk size in ``gnt-instance info``
2023 - Fixed checking for valid OSes in instance creation
2024 - Fixed handling of the "vcpus" parameter in instance listing and in
2025 general of invalid parameters
2026 - Fixed http server library, and thus RAPI, to handle invalid
2027 username/password combinations correctly; this means that now they
2028 report unauthorized for queries too, not only for modifications,
2029 allowing earlier detect of configuration problems
2030 - Added a new "role" node list field, equivalent to the master/master
2031 candidate/drained/offline flags combinations
2032 - Fixed cluster modify and changes of candidate pool size
2033 - Fixed cluster verify error messages for wrong files on regular nodes
2034 - Fixed a couple of issues with node demotion from master candidate role
2035 - Fixed node readd issues
2036 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
2037 - Added a new ``--no-voting`` option for masterfailover to fix failover
2038 on two-nodes clusters when the former master node is unreachable
2039 - Added instance reinstall over RAPI
2045 *(Released Tue, 16 Jun 2009)*
2047 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
2048 allow re-adding the start in single-user option (regression from 1.2)
2049 - the watcher writes the instance status to a file, to allow monitoring
2050 to report the instance status (from the master) based on cached
2051 results of the watcher's queries; while this can get stale if the
2052 watcher is being locked due to other work on the cluster, this is
2053 still an improvement
2054 - the watcher now also restarts the node daemon and the rapi daemon if
2056 - fixed the watcher to handle full and drained queue cases
2057 - hooks export more instance data in the environment, which helps if
2058 hook scripts need to take action based on the instance's properties
2059 (no longer need to query back into ganeti)
2060 - instance failovers when the instance is stopped do not check for free
2061 RAM, so that failing over a stopped instance is possible in low memory
2063 - rapi uses queries for tags instead of jobs (for less job traffic), and
2064 for cluster tags it won't talk to masterd at all but read them from
2066 - a couple of error handling fixes in RAPI
2067 - drbd handling: improved the error handling of inconsistent disks after
2068 resync to reduce the frequency of "there are some degraded disks for
2069 this instance" messages
2070 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
2071 error handling path called a wrong function name)
2077 *(Released Wed, 27 May 2009)*
2079 - no changes from rc5
2085 *(Released Wed, 20 May 2009)*
2087 - fix a couple of bugs (validation, argument checks)
2088 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
2089 - some small improvements to RAPI and IAllocator
2090 - make watcher automatically start the master daemon if down
2096 *(Released Mon, 27 Apr 2009)*
2098 - change the OS list to not require locks; this helps with big clusters
2099 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
2100 volume group is broken
2101 - ``gnt-instance info``, without any arguments, doesn't run for all
2102 instances anymore; either pass ``--all`` or pass the desired
2103 instances; this helps against mistakes on big clusters where listing
2104 the information for all instances takes a long time
2105 - miscellaneous doc and man pages fixes
2111 *(Released Wed, 8 Apr 2009)*
2113 - Change the internal locking model of some ``gnt-node`` commands, in
2114 order to reduce contention (and blocking of master daemon) when
2115 batching many creation/reinstall jobs
2116 - Fixes to Xen soft reboot
2117 - No longer build documentation at build time, instead distribute it in
2118 the archive, in order to reduce the need for the whole docbook/rst
2125 *(Released Fri, 27 Mar 2009)*
2127 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
2128 - Fix watcher startup sequence, improves the behaviour of busy clusters
2129 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
2130 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
2131 bind address and other places
2132 - Some documentation fixes and updates
2138 *(Released Mon, 2 Mar 2009)*
2140 - More documentation updates, now all docs should be more-or-less
2142 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
2144 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
2146 - Fix an issue related to $libdir/run/ganeti and cluster creation
2152 *(Released Thu, 19 Feb 2009)*
2154 - Xen PVM and KVM have switched the default value for the instance root
2155 disk to the first partition on the first drive, instead of the whole
2156 drive; this means that the OS installation scripts must be changed
2158 - Man pages have been updated
2159 - RAPI has been switched by default to HTTPS, and the exported functions
2160 should all work correctly
2161 - RAPI v1 has been removed
2162 - Many improvements to the KVM hypervisor
2163 - Block device errors are now better reported
2164 - Many other bugfixes and small improvements
2170 *(Released Mon, 26 Jan 2009)*
2172 - Version 2 is a general rewrite of the code and therefore the
2173 differences are too many to list, see the design document for 2.0 in
2174 the ``doc/`` subdirectory for more details
2175 - In this beta version there is not yet a migration path from 1.2 (there
2176 will be one in the final 2.0 release)
2177 - A few significant changes are:
2179 - all commands are executed by a daemon (``ganeti-masterd``) and the
2180 various ``gnt-*`` commands are just front-ends to it
2181 - all the commands are entered into, and executed from a job queue,
2182 see the ``gnt-job(8)`` manpage
2183 - the RAPI daemon supports read-write operations, secured by basic
2184 HTTP authentication on top of HTTPS
2185 - DRBD version 0.7 support has been removed, DRBD 8 is the only
2186 supported version (when migrating from Ganeti 1.2 to 2.0, you need
2187 to migrate to DRBD 8 first while still running Ganeti 1.2)
2188 - DRBD devices are using statically allocated minor numbers, which
2189 will be assigned to existing instances during the migration process
2190 - there is support for both Xen PVM and Xen HVM instances running on
2192 - KVM virtualization is supported too
2193 - file-based storage has been implemented, which means that it is
2194 possible to run the cluster without LVM and DRBD storage, for
2195 example using a shared filesystem exported from shared storage (and
2196 still have live migration)
2202 *(Released Tue, 13 Jan 2009)*
2204 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
2205 - Reuse the old instance mac address by default on instance import, if
2206 the instance name is the same.
2207 - Handle situations in which the node info rpc returns incomplete
2209 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
2210 - Improved version of batcher:
2212 - state file support
2213 - instance mac address support
2214 - support for HVM clusters/instances
2216 - Add an option to show the number of cpu sockets and nodes in
2218 - Support OSes that handle more than one version of the OS api (but do
2219 not change the current API in any other way)
2220 - Fix ``gnt-node migrate``
2221 - ``gnt-debug`` man page
2222 - Fixes various more typos and small issues
2223 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
2229 *(Released Wed, 24 Sep 2008)*
2231 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
2232 type of disk exported to fully virtualized instances.
2233 - provide access to the serial console of HVM instances
2234 - instance auto_balance flag, set by default. If turned off it will
2235 avoid warnings on cluster verify if there is not enough memory to fail
2236 over an instance. in the future it will prevent automatically failing
2237 it over when we will support that.
2238 - batcher tool for instance creation, see ``tools/README.batcher``
2239 - ``gnt-instance reinstall --select-os`` to interactively select a new
2240 operating system when reinstalling an instance.
2241 - when changing the memory amount on instance modify a check has been
2242 added that the instance will be able to start. also warnings are
2243 emitted if the instance will not be able to fail over, if auto_balance
2245 - documentation fixes
2246 - sync fields between ``gnt-instance list/modify/add/import``
2247 - fix a race condition in drbd when the sync speed was set after giving
2248 the device a remote peer.
2254 *(Released Tue, 22 Jul 2008)*
2256 - note: the allowed size and number of tags per object were reduced
2257 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
2258 - fixed twisted 8.x compatibility
2259 - fixed ``gnt-instance replace-disks`` with iallocator
2260 - add TCP keepalives on twisted connections to detect restarted nodes
2261 - disk increase support, see ``gnt-instance grow-disk``
2262 - implement bulk node/instance query for RAPI
2263 - add tags in node/instance listing (optional)
2264 - experimental migration (and live migration) support, read the man page
2265 for ``gnt-instance migrate``
2266 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
2267 has some small improvements in handling its state file
2273 *(Released Fri, 13 Jun 2008)*
2275 - Experimental readonly, REST-based remote API implementation;
2276 automatically started on master node, TCP port 5080, if enabled by
2277 ``--enable-rapi`` parameter to configure script.
2278 - Instance allocator support. Add and import instance accept a
2279 ``--iallocator`` parameter, and call that instance allocator to decide
2280 which node to use for the instance. The iallocator document describes
2281 what's expected from an allocator script.
2282 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
2283 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
2284 node is lost there is still enough memory to fail over the instances
2286 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
2287 ``gnt-cluster verify``, to check for site-specific compliance. All the
2288 hooks will run, and their output, if any, will be displayed. Any
2289 failing hook will make the verification return an error value.
2290 - ``gnt-cluster verify`` now checks that its peers are reachable on the
2291 primary and secondary interfaces
2292 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
2293 that is still declared as part of the cluster and has failed.
2294 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
2295 specifying output fields, that just adds the chosen fields to the
2297 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
2298 export from the filesystem.
2299 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
2300 have been added. Using them you can enable/disable acpi and pae
2301 support, and specify a path for a cd image to be exported to the
2302 instance. These parameters as the name suggest only work on HVM
2304 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
2305 PAE support will be set to the previously hardcoded values, but the
2306 (previously hardcoded) path to the CDROM ISO image will be unset and
2307 if required, needs to be set manually with ``gnt-instance modify``
2309 - The address to which an instance's VNC console is bound is now
2310 selectable per-instance, rather than being cluster wide. Of course
2311 this only applies to instances controlled via VNC, so currently just
2312 applies to HVM clusters.
2318 *(Released Mon, 18 Feb 2008)*
2320 - more tweaks to the disk activation code (especially helpful for DRBD)
2321 - change the default ``gnt-instance list`` output format, now there is
2322 one combined status field (see the manpage for the exact values this
2324 - some more fixes for the mac export to hooks change
2325 - make Ganeti not break with DRBD 8.2.x (which changed the version
2326 format in ``/proc/drbd``) (issue 24)
2327 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
2328 template, allowing migration from DRBD0.7+MD to DRBD8
2334 *(Released Wed, 30 Jan 2008)*
2336 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
2338 - add command aliases infrastructure and a few aliases
2339 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
2340 man pages and the ``--help`` option of ``gnt-node
2341 list``/``gnt-instance list``
2342 - fix ``gnt-backup list`` with down nodes (issue 21)
2343 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
2344 - fix the dist archive and add a check for including svn/git files in
2346 - some developer-related changes: improve the burnin and the QA suite,
2347 add an upload script for testing during development
2353 *(Released Wed, 16 Jan 2008)*
2355 - experimental HVM support, read the install document, section
2356 "Initializing the cluster"
2357 - allow for the PVM hypervisor per-instance kernel and initrd paths
2358 - add a new command ``gnt-cluster verify-disks`` which uses a new
2359 algorithm to improve the reconnection of the DRBD pairs if the device
2360 on the secondary node has gone away
2361 - make logical volume code auto-activate LVs at disk activation time
2362 - slightly improve the speed of activating disks
2363 - allow specification of the MAC address at instance creation time, and
2364 changing it later via ``gnt-instance modify``
2365 - fix handling of external commands that generate lots of output on
2367 - update documentation with regard to minimum version of DRBD8 supported
2373 *(Released Tue, 4 Dec 2007)*
2375 - Log the ``xm create`` output to the node daemon log on failure (to
2376 help diagnosing the error)
2377 - In debug mode, log all external commands output if failed to the logs
2378 - Change parsing of lvm commands to ignore stderr
2384 *(Released Wed, 28 Nov 2007)*
2386 - Another round of updates to the DRBD 8 code to deal with more failures
2387 in the replace secondary node operation
2388 - Some more logging of failures in disk operations (lvm, drbd)
2389 - A few documentation updates
2396 *(Released Tue, 13 Nov 2007)*
2398 - Change configuration file format from Python's Pickle to JSON.
2399 Upgrading is possible using the cfgupgrade utility.
2400 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
2401 faster replace disks and is more stable (DRBD 8 has many improvements
2402 compared to DRBD 0.7)
2403 - Added command line tags support (see man pages for ``gnt-instance``,
2404 ``gnt-node``, ``gnt-cluster``)
2405 - Added instance rename support
2406 - Added multi-instance startup/shutdown
2407 - Added cluster rename support
2408 - Added ``gnt-node evacuate`` to simplify some node operations
2409 - Added instance reboot operation that can speedup reboot as compared to
2411 - Soften the requirement that hostnames are in FQDN format
2412 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
2414 - Removed dependency on debian's patched fping that uses the
2415 non-standard ``-S`` option
2416 - Now the OS definitions are searched for in multiple, configurable
2417 paths (easier for distros to package)
2418 - Some changes to the hooks infrastructure (especially the new
2419 post-configuration update hook)
2420 - Other small bugfixes
2422 .. vim: set textwidth=72 syntax=rst :