8 *(Released Mon, 24 Jun 2013)*
10 Incompatible/important changes
11 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
13 - Instance policy can contain multiple instance specs, as described in
14 the “Constrained instance sizes” section of :doc:`Partitioned Ganeti
15 <design-partitioned>`. As a consequence, it's not possible to partially change
16 or override instance specs. Bounding specs (min and max) can be specified as a
17 whole using the new option ``--ipolicy-bounds-specs``, while standard
18 specs use the new option ``--ipolicy-std-specs``.
19 - The output of the info command of gnt-cluster, gnt-group, gnt-node,
20 gnt-instance is a valid YAML object.
21 - hail now honors network restrictions when allocating nodes. This led to an
22 update of the IAllocator protocol. See the IAllocator documentation for
28 - The :doc:`Remote API <rapi>` daemon now supports a command line flag
29 to always require authentication, ``--require-authentication``. It can
30 be specified in ``$sysconfdir/default/ganeti``.
31 - A new cluster attribute 'enabled_disk_templates' is introduced. It will
32 be used to manage the disk templates to be used by instances in the cluster.
33 Initially, it will be set to a list that includes plain, drbd, if they were
34 enabled by specifying a volume group name, and file and sharedfile, if those
35 were enabled at configure time. Additionally, it will include all disk
36 templates that are currently used by instances. The order of disk templates
37 will be based on Ganeti's history of supporting them. In the future, the
38 first entry of the list will be used as a default disk template on instance
40 - ``cfgupgrade`` now supports a ``--downgrade`` option to bring the
41 configuration back to the previous stable version.
42 - Disk templates in group ipolicy can be restored to the default value.
43 - Initial support for diskless instances and virtual clusters in QA.
44 - More QA and unit tests for instance policies.
45 - Every opcode now contains a reason trail (visible through ``gnt-job info``)
46 describing why the opcode itself was executed.
47 - The monitoring daemon is now available. It allows users to query the cluster
48 for obtaining information about the status of the system. The daemon is only
49 responsible for providing the information over the network: the actual data
50 gathering is performed by data collectors (currently, only the DRBD status
51 collector is available).
52 - In order to help developers work on Ganeti, a new script
53 (``devel/build_chroot``) is provided, for building a chroot that contains all
54 the required development libraries and tools for compiling Ganeti on a Debian
56 - A new tool, ``harep``, for performing self-repair and recreation of instances
57 in Ganeti has been added.
58 - Split queries are enabled for tags, network, exports, cluster info, groups,
60 - New command ``show-ispecs-cmd`` for ``gnt-cluster`` and ``gnt-group``.
61 It prints the command line to set the current policies, to ease
66 The following new dependencies have been added:
69 - The ``curl`` library is not optional anymore for compiling the Haskell code.
70 - ``snap-server`` library (if monitoring is enabled).
73 - The minimum Python version needed to run Ganeti is now 2.6.
74 - ``yaml`` library (only for running the QA).
80 *(Released Fri, 24 May 2013)*
82 Incompatible/important changes
83 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
85 - Instance policies for disk size were documented to be on a per-disk
86 basis, but hail applied them to the sum of all disks. This has been
88 - ``hbal`` will now exit with status 0 if, during job execution over
89 LUXI, early exit has been requested and all jobs are successful;
90 before, exit status 1 was used, which cannot be differentiated from
92 - Compatibility with newer versions of rbd has been fixed
93 - ``gnt-instance batch-create`` has been changed to use the bulk create
94 opcode from Ganeti. This lead to incompatible changes in the format of
95 the JSON file. It's now not a custom dict anymore but a dict
96 compatible with the ``OpInstanceCreate`` opcode.
97 - Parent directories for file storage need to be listed in
98 ``$sysconfdir/ganeti/file-storage-paths`` now. ``cfgupgrade`` will
99 write the file automatically based on old configuration values, but it
100 can not distribute it across all nodes and the file contents should be
101 verified. Use ``gnt-cluster copyfile
102 $sysconfdir/ganeti/file-storage-paths`` once the cluster has been
103 upgraded. The reason for requiring this list of paths now is that
104 before it would have been possible to inject new paths via RPC,
105 allowing files to be created in arbitrary locations. The RPC protocol
106 is protected using SSL/X.509 certificates, but as a design principle
107 Ganeti does not permit arbitrary paths to be passed.
108 - The parsing of the variants file for OSes (see
109 :manpage:`ganeti-os-interface(7)`) has been slightly changed: now empty
110 lines and comment lines (starting with ``#``) are ignored for better
112 - The ``setup-ssh`` tool added in Ganeti 2.2 has been replaced and is no
113 longer available. ``gnt-node add`` now invokes a new tool on the
114 destination node, named ``prepare-node-join``, to configure the SSH
115 daemon. Paramiko is no longer necessary to configure nodes' SSH
116 daemons via ``gnt-node add``.
117 - Draining (``gnt-cluster queue drain``) and un-draining the job queue
118 (``gnt-cluster queue undrain``) now affects all nodes in a cluster and
119 the flag is not reset after a master failover.
120 - Python 2.4 has *not* been tested with this release. Using 2.6 or above
121 is recommended. 2.6 will be mandatory from the 2.8 series.
127 - New network management functionality to support automatic allocation
128 of IP addresses and managing of network parameters. See
129 :manpage:`gnt-network(8)` for more details.
130 - New external storage backend, to allow managing arbitrary storage
131 systems external to the cluster. See
132 :manpage:`ganeti-extstorage-interface(7)`.
133 - New ``exclusive-storage`` node parameter added, restricted to
134 nodegroup level. When it's set to true, physical disks are assigned in
135 an exclusive fashion to instances, as documented in :doc:`Partitioned
136 Ganeti <design-partitioned>`. Currently, only instances using the
137 ``plain`` disk template are supported.
138 - The KVM hypervisor has been updated with many new hypervisor
139 parameters, including a generic one for passing arbitrary command line
140 values. See a complete list in :manpage:`gnt-instance(8)`. It is now
141 compatible up to qemu 1.4.
142 - A new tool, called ``mon-collector``, is the stand-alone executor of
143 the data collectors for a monitoring system. As of this version, it
144 just includes the DRBD data collector, that can be executed by calling
145 ``mon-collector`` using the ``drbd`` parameter. See
146 :manpage:`mon-collector(7)`.
147 - A new user option, :pyeval:`rapi.RAPI_ACCESS_READ`, has been added
148 for RAPI users. It allows granting permissions to query for
149 information to a specific user without giving
150 :pyeval:`rapi.RAPI_ACCESS_WRITE` permissions.
151 - A new tool named ``node-cleanup`` has been added. It cleans remains of
152 a cluster from a machine by stopping all daemons, removing
153 certificates and ssconf files. Unless the ``--no-backup`` option is
154 given, copies of the certificates are made.
155 - Instance creations now support the use of opportunistic locking,
156 potentially speeding up the (parallel) creation of multiple instances.
157 This feature is currently only available via the :doc:`RAPI
158 <rapi>` interface and when an instance allocator is used. If the
159 ``opportunistic_locking`` parameter is set the opcode will try to
160 acquire as many locks as possible, but will not wait for any locks
161 held by other opcodes. If not enough resources can be found to
162 allocate the instance, the temporary error code
163 :pyeval:`errors.ECODE_TEMP_NORES` is returned. The operation can be
164 retried thereafter, with or without opportunistic locking.
165 - New experimental linux-ha resource scripts.
166 - Restricted-commands support: ganeti can now be asked (via command line
167 or rapi) to perform commands on a node. These are passed via ganeti
168 RPC rather than ssh. This functionality is restricted to commands
169 specified on the ``$sysconfdir/ganeti/restricted-commands`` for security
170 reasons. The file is not copied automatically.
176 - Diskless instances are now externally mirrored (Issue 237). This for
177 now has only been tested in conjunction with explicit target nodes for
179 - Queries not needing locks or RPC access to the node can now be
180 performed by the confd daemon, making them independent from jobs, and
181 thus faster to execute. This is selectable at configure time.
182 - The functionality for allocating multiple instances at once has been
183 overhauled and is now also available through :doc:`RAPI <rapi>`.
187 - ``devel/upload`` now works when ``/var/run`` on the target nodes is a
189 - Disks added through ``gnt-instance modify`` or created through
190 ``gnt-instance recreate-disks`` are wiped, if the
191 ``prealloc_wipe_disks`` flag is set.
192 - If wiping newly created disks fails, the disks are removed. Also,
193 partial failures in creating disks through ``gnt-instance modify``
194 triggers a cleanup of the partially-created disks.
195 - Removing the master IP address doesn't fail if the address has been
197 - Fix ownership of the OS log dir
198 - Workaround missing SO_PEERCRED constant (Issue 191)
204 *(Released Fri, 3 May 2013)*
206 This was the first release candidate of the 2.7 series. Since beta3:
208 - Fix kvm compatibility with qemu 1.4 (Issue 389)
209 - Documentation updates (admin guide, upgrade notes, install
210 instructions) (Issue 372)
211 - Fix gnt-group list nodes and instances count (Issue 436)
212 - Fix compilation without non-mandatory libraries (Issue 441)
213 - Fix xen-hvm hypervisor forcing nics to type 'ioemu' (Issue 247)
214 - Make confd logging more verbose at INFO level (Issue 435)
215 - Improve "networks" documentation in :manpage:`gnt-instance(8)`
216 - Fix failure path for instance storage type conversion (Issue 229)
217 - Update htools text backend documentation
218 - Improve the renew-crypto section of :manpage:`gnt-cluster(8)`
219 - Disable inter-cluster instance move for file-based instances, because
220 it is dependant on instance export, which is not supported for
221 file-based instances. (Issue 414)
222 - Fix gnt-job crashes on non-ascii characters (Issue 427)
223 - Fix volume group checks on non-vm-capable nodes (Issue 432)
229 *(Released Mon, 22 Apr 2013)*
231 This was the third beta release of the 2.7 series. Since beta2:
233 - Fix hail to verify disk instance policies on a per-disk basis (Issue 418).
234 - Fix data loss on wrong usage of ``gnt-instance move``
235 - Properly export errors in confd-based job queries
236 - Add ``users-setup`` tool
237 - Fix iallocator protocol to report 0 as a disk size for diskless
238 instances. This avoids hail breaking when a diskless instance is
240 - Fix job queue directory permission problem that made confd job queries
241 fail. This requires running an ``ensure-dirs --full-run`` on upgrade
242 for access to archived jobs (Issue 406).
243 - Limit the sizes of networks supported by ``gnt-network`` to something
244 between a ``/16`` and a ``/30`` to prevent memory bloat and crashes.
245 - Fix bugs in instance disk template conversion
246 - Fix GHC 7 compatibility
247 - Fix ``burnin`` install path (Issue 426).
248 - Allow very small disk grows (Issue 347).
249 - Fix a ``ganeti-noded`` memory bloat introduced in 2.5, by making sure
250 that noded doesn't import masterd code (Issue 419).
251 - Make sure the default metavg at cluster init is the same as the vg, if
252 unspecified (Issue 358).
253 - Fix cleanup of partially created disks (part of Issue 416)
259 *(Released Tue, 2 Apr 2013)*
261 This was the second beta release of the 2.7 series. Since beta1:
263 - Networks no longer have a "type" slot, since this information was
264 unused in Ganeti: instead of it tags should be used.
265 - The rapi client now has a ``target_node`` option to MigrateInstance.
266 - Fix early exit return code for hbal (Issue 386).
267 - Fix ``gnt-instance migrate/failover -n`` (Issue 396).
268 - Fix ``rbd showmapped`` output parsing (Issue 312).
269 - Networks are now referenced indexed by UUID, rather than name. This
270 will require running cfgupgrade, from 2.7.0beta1, if networks are in
272 - The OS environment now includes network information.
273 - Deleting of a network is now disallowed if any instance nic is using
274 it, to prevent dangling references.
275 - External storage is now documented in man pages.
276 - The exclusive_storage flag can now only be set at nodegroup level.
277 - Hbal can now submit an explicit priority with its jobs.
278 - Many network related locking fixes.
279 - Bump up the required pylint version to 0.25.1.
280 - Fix the ``no_remember`` option in RAPI client.
281 - Many ipolicy related tests, qa, and fixes.
282 - Many documentation improvements and fixes.
283 - Fix building with ``--disable-file-storage``.
284 - Fix ``-q`` option in htools, which was broken if passed more than
286 - Some haskell/python interaction improvements and fixes.
287 - Fix iallocator in case of missing LVM storage.
288 - Fix confd config load in case of ``--no-lvm-storage``.
289 - The confd/query functionality is now mentioned in the security
296 *(Released Wed, 6 Feb 2013)*
298 This was the first beta release of the 2.7 series. All important changes
299 are listed in the latest 2.7 entry.
305 *(Released Fri, 21 Dec 2012)*
307 Important behaviour change: hbal won't rebalance anymore instances which
308 have the ``auto_balance`` attribute set to false. This was the intention
309 all along, but until now it only skipped those from the N+1 memory
310 reservation (DRBD-specific).
312 A significant number of bug fixes in this release:
314 - Fixed disk adoption interaction with ipolicy checks.
315 - Fixed networking issues when instances are started, stopped or
316 migrated, by forcing the tap device's MAC prefix to "fe" (issue 217).
317 - Fixed the warning in cluster verify for shared storage instances not
319 - Fixed removal of storage directory on shared file storage (issue 262).
320 - Fixed validation of LVM volume group name in OpClusterSetParams
321 (``gnt-cluster modify``) (issue 285).
322 - Fixed runtime memory increases (``gnt-instance modify -m``).
323 - Fixed live migration under Xen's ``xl`` mode.
324 - Fixed ``gnt-instance console`` with ``xl``.
325 - Fixed building with newer Haskell compiler/libraries.
326 - Fixed PID file writing in Haskell daemons (confd); this prevents
327 restart issues if confd was launched manually (outside of
328 ``daemon-util``) while another copy of it was running
329 - Fixed a type error when doing live migrations with KVM (issue 297) and
330 the error messages for failing migrations have been improved.
331 - Fixed opcode validation for the out-of-band commands (``gnt-node
333 - Fixed a type error when unsetting OS hypervisor parameters (issue
334 311); now it's possible to unset all OS-specific hypervisor
336 - Fixed the ``dry-run`` mode for many operations: verification of
337 results was over-zealous but didn't take into account the ``dry-run``
338 operation, resulting in "wrong" failures.
339 - Fixed bash completion in ``gnt-job list`` when the job queue has
340 hundreds of entries; especially with older ``bash`` versions, this
341 results in significant CPU usage.
343 And lastly, a few other improvements have been made:
345 - Added option to force master-failover without voting (issue 282).
346 - Clarified error message on lock conflict (issue 287).
347 - Logging of newly submitted jobs has been improved (issue 290).
348 - Hostname checks have been made uniform between instance rename and
350 - The ``--submit`` option is now supported by ``gnt-debug delay``.
351 - Shutting down the master daemon by sending SIGTERM now stops it from
352 processing jobs waiting for locks; instead, those jobs will be started
353 once again after the master daemon is started the next time (issue
355 - Support for Xen's ``xl`` program has been improved (besides the fixes
357 - Reduced logging noise in the Haskell confd daemon (only show one log
358 entry for each config reload, instead of two).
359 - Several man page updates and typo fixes.
365 *(Released Fri, 12 Oct 2012)*
367 A small bugfix release. Among the bugs fixed:
369 - Fixed double use of ``PRIORITY_OPT`` in ``gnt-node migrate``, that
370 made the command unusable.
371 - Commands that issue many jobs don't fail anymore just because some jobs
372 take so long that other jobs are archived.
373 - Failures during ``gnt-instance reinstall`` are reflected by the exit
375 - Issue 190 fixed. Check for DRBD in cluster verify is enabled only when
377 - When ``always_failover`` is set, ``--allow-failover`` is not required
378 in migrate commands anymore.
379 - ``bash_completion`` works even if extglob is disabled.
380 - Fixed bug with locks that made failover for RDB-based instances fail.
381 - Fixed bug in non-mirrored instance allocation that made Ganeti choose
382 a random node instead of one based on the allocator metric.
383 - Support for newer versions of pylint and pep8.
384 - Hail doesn't fail anymore when trying to add an instance of type
385 ``file``, ``sharedfile`` or ``rbd``.
386 - Added new Makefile target to rebuild the whole distribution, so that
387 all files are included.
393 *(Released Fri, 27 Jul 2012)*
396 .. attention:: The ``LUXI`` protocol has been made more consistent
397 regarding its handling of command arguments. This, however, leads to
398 incompatibility issues with previous versions. Please ensure that you
399 restart Ganeti daemons soon after the upgrade, otherwise most
400 ``LUXI`` calls (job submission, setting/resetting the drain flag,
401 pausing/resuming the watcher, cancelling and archiving jobs, querying
402 the cluster configuration) will fail.
411 The current ``admin_up`` field, which used to denote whether an instance
412 should be running or not, has been removed. Instead, ``admin_state`` is
413 introduced, with 3 possible values -- ``up``, ``down`` and ``offline``.
415 The rational behind this is that an instance being “down” can have
418 - it could be down during a reboot
419 - it could be temporarily be down for a reinstall
420 - or it could be down because it is deprecated and kept just for its
423 The previous Boolean state was making it difficult to do capacity
424 calculations: should Ganeti reserve memory for a down instance? Now, the
425 tri-state field makes it clear:
427 - in ``up`` and ``down`` state, all resources are reserved for the
428 instance, and it can be at any time brought up if it is down
429 - in ``offline`` state, only disk space is reserved for it, but not
432 The field can have an extra use: since the transition between ``up`` and
433 ``down`` and vice-versus is done via ``gnt-instance start/stop``, but
434 transition between ``offline`` and ``down`` is done via ``gnt-instance
435 modify``, it is possible to given different rights to users. For
436 example, owners of an instance could be allowed to start/stop it, but
437 not transition it out of the offline state.
439 Instance policies and specs
440 +++++++++++++++++++++++++++
442 In previous Ganeti versions, an instance creation request was not
443 limited on the minimum size and on the maximum size just by the cluster
444 resources. As such, any policy could be implemented only in third-party
445 clients (RAPI clients, or shell wrappers over ``gnt-*``
446 tools). Furthermore, calculating cluster capacity via ``hspace`` again
447 required external input with regards to instance sizes.
449 In order to improve these workflows and to allow for example better
450 per-node group differentiation, we introduced instance specs, which
453 - minimum instance disk size, disk count, memory size, cpu count
454 - maximum values for the above metrics
455 - and “standard” values (used in ``hspace`` to calculate the standard
458 The minimum/maximum values can be also customised at node-group level,
459 for example allowing more powerful hardware to support bigger instance
462 Beside the instance specs, there are a few other settings belonging to
463 the instance policy framework. It is possible now to customise, per
464 cluster and node-group:
466 - the list of allowed disk templates
467 - the maximum ratio of VCPUs per PCPUs (to control CPU oversubscription)
468 - the maximum ratio of instance to spindles (see below for more
469 information) for local storage
471 All these together should allow all tools that talk to Ganeti to know
472 what are the ranges of allowed values for instances and the
473 over-subscription that is allowed.
475 For the VCPU/PCPU ratio, we already have the VCPU configuration from the
476 instance configuration, and the physical CPU configuration from the
477 node. For the spindle ratios however, we didn't track before these
478 values, so new parameters have been added:
480 - a new node parameter ``spindle_count``, defaults to 1, customisable at
481 node group or node level
482 - at new backend parameter (for instances), ``spindle_use`` defaults to 1
484 Note that spindles in this context doesn't need to mean actual
485 mechanical hard-drives; it's just a relative number for both the node
486 I/O capacity and instance I/O consumption.
488 Instance migration behaviour
489 ++++++++++++++++++++++++++++
491 While live-migration is in general desirable over failover, it is
492 possible that for some workloads it is actually worse, due to the
493 variable time of the “suspend” phase during live migration.
495 To allow the tools to work consistently over such instances (without
496 having to hard-code instance names), a new backend parameter
497 ``always_failover`` has been added to control the migration/failover
498 behaviour. When set to True, all migration requests for an instance will
499 instead fall-back to failover.
501 Instance memory ballooning
502 ++++++++++++++++++++++++++
504 Initial support for memory ballooning has been added. The memory for an
505 instance is no longer fixed (backend parameter ``memory``), but instead
506 can vary between minimum and maximum values (backend parameters
507 ``minmem`` and ``maxmem``). Currently we only change an instance's
510 - live migrating or failing over and instance and the target node
511 doesn't have enough memory
512 - user requests changing the memory via ``gnt-instance modify
518 In order to control the use of specific CPUs by instance, support for
519 controlling CPU pinning has been added for the Xen, HVM and LXC
520 hypervisors. This is controlled by a new hypervisor parameter
521 ``cpu_mask``; details about possible values for this are in the
522 :manpage:`gnt-instance(8)`. Note that use of the most specific (precise
523 VCPU-to-CPU mapping) form will work well only when all nodes in your
524 cluster have the same amount of CPUs.
529 Another area in which Ganeti was not customisable were the parameters
530 used for storage configuration, e.g. how many stripes to use for LVM,
531 DRBD resync configuration, etc.
533 To improve this area, we've added disks parameters, which are
534 customisable at cluster and node group level, and which allow to
535 specify various parameters for disks (DRBD has the most parameters
536 currently), for example:
538 - DRBD resync algorithm and parameters (e.g. speed)
539 - the default VG for meta-data volumes for DRBD
540 - number of stripes for LVM (plain disk template)
543 These parameters can be modified via ``gnt-cluster modify -D …`` and
544 ``gnt-group modify -D …``, and are used at either instance creation (in
545 case of LVM stripes, for example) or at disk “activation” time
548 Rados block device support
549 ++++++++++++++++++++++++++
551 A Rados (http://ceph.com/wiki/Rbd) storage backend has been added,
552 denoted by the ``rbd`` disk template type. This is considered
553 experimental, feedback is welcome. For details on configuring it, see
554 the :doc:`install` document and the :manpage:`gnt-cluster(8)` man page.
559 The existing master IP functionality works well only in simple setups (a
560 single network shared by all nodes); however, if nodes belong to
561 different networks, then the ``/32`` setup and lack of routing
562 information is not enough.
564 To allow the master IP to function well in more complex cases, the
565 system was reworked as follows:
567 - a master IP netmask setting has been added
568 - the master IP activation/turn-down code was moved from the node daemon
570 - whether to run the Ganeti-supplied master IP script or a user-supplied
571 on is a ``gnt-cluster init`` setting
573 Details about the location of the standard and custom setup scripts are
574 in the man page :manpage:`gnt-cluster(8)`; for information about the
575 setup script protocol, look at the Ganeti-supplied script.
580 The `SPICE <http://www.linux-kvm.org/page/SPICE>`_ support has been
583 It is now possible to use TLS-protected connections, and when renewing
584 or changing the cluster certificates (via ``gnt-cluster renew-crypto``,
585 it is now possible to specify spice or spice CA certificates. Also, it
586 is possible to configure a password for SPICE sessions via the
587 hypervisor parameter ``spice_password_file``.
589 There are also new parameters to control the compression and streaming
590 options (e.g. ``spice_image_compression``, ``spice_streaming_video``,
591 etc.). For details, see the man page :manpage:`gnt-instance(8)` and look
592 for the spice parameters.
594 Lastly, it is now possible to see the SPICE connection information via
595 ``gnt-instance console``.
600 A new tool (``tools/ovfconverter``) has been added that supports
601 conversion between Ganeti and the `Open Virtualization Format
602 <http://en.wikipedia.org/wiki/Open_Virtualization_Format>`_ (both to and
605 This relies on the ``qemu-img`` tool to convert the disk formats, so the
606 actual compatibility with other virtualization solutions depends on it.
611 The configuration query daemon (``ganeti-confd``) is now optional, and
612 has been rewritten in Haskell; whether to use the daemon at all, use the
613 Python (default) or the Haskell version is selectable at configure time
614 via the ``--enable-confd`` parameter, which can take one of the
615 ``haskell``, ``python`` or ``no`` values. If not used, disabling the
616 daemon will result in a smaller footprint; for larger systems, we
617 welcome feedback on the Haskell version which might become the default
620 If you want to use ``gnt-node list-drbd`` you need to have the Haskell
621 daemon running. The Python version doesn't implement the new call.
624 User interface changes
625 ~~~~~~~~~~~~~~~~~~~~~~
627 We have replaced the ``--disks`` option of ``gnt-instance
628 replace-disks`` with a more flexible ``--disk`` option, which allows
629 adding and removing disks at arbitrary indices (Issue 188). Furthermore,
630 disk size and mode can be changed upon recreation (via ``gnt-instance
631 recreate-disks``, which accepts the same ``--disk`` option).
633 As many people are used to a ``show`` command, we have added that as an
634 alias to ``info`` on all ``gnt-*`` commands.
636 The ``gnt-instance grow-disk`` command has a new mode in which it can
637 accept the target size of the disk, instead of the delta; this can be
638 more safe since two runs in absolute mode will be idempotent, and
639 sometimes it's also easier to specify the desired size directly.
641 Also the handling of instances with regard to offline secondaries has
642 been improved. Instance operations should not fail because one of it's
643 secondary nodes is offline, even though it's safe to proceed.
645 A new command ``list-drbd`` has been added to the ``gnt-node`` script to
646 support debugging of DRBD issues on nodes. It provides a mapping of DRBD
647 minors to instance name.
652 RAPI coverage has improved, with (for example) new resources for
653 recreate-disks, node power-cycle, etc.
658 There is partial support for ``xl`` in the Xen hypervisor; feedback is
661 Python 2.7 is better supported, and after Ganeti 2.6 we will investigate
662 whether to still support Python 2.4 or move to Python 2.6 as minimum
665 Support for Fedora has been slightly improved; the provided example
666 init.d script should work better on it and the INSTALL file should
667 document the needed dependencies.
672 The deprecated ``QueryLocks`` LUXI request has been removed. Use
673 ``Query(what=QR_LOCK, ...)`` instead.
675 The LUXI requests :pyeval:`luxi.REQ_QUERY_JOBS`,
676 :pyeval:`luxi.REQ_QUERY_INSTANCES`, :pyeval:`luxi.REQ_QUERY_NODES`,
677 :pyeval:`luxi.REQ_QUERY_GROUPS`, :pyeval:`luxi.REQ_QUERY_EXPORTS` and
678 :pyeval:`luxi.REQ_QUERY_TAGS` are deprecated and will be removed in a
679 future version. :pyeval:`luxi.REQ_QUERY` should be used instead.
681 RAPI client: ``CertificateError`` now derives from
682 ``GanetiApiError``. This should make it more easy to handle Ganeti
685 Deprecation warnings due to PyCrypto/paramiko import in
686 ``tools/setup-ssh`` have been silenced, as usually they are safe; please
687 make sure to run an up-to-date paramiko version, if you use this tool.
689 The QA scripts now depend on Python 2.5 or above (the main code base
690 still works with Python 2.4).
692 The configuration file (``config.data``) is now written without
693 indentation for performance reasons; if you want to edit it, it can be
694 re-formatted via ``tools/fmtjson``.
696 A number of bugs has been fixed in the cluster merge tool.
698 ``x509`` certification verification (used in import-export) has been
699 changed to allow the same clock skew as permitted by the cluster
700 verification. This will remove some rare but hard to diagnose errors in
707 *(Released Thu, 19 Jul 2012)*
709 Very few changes from rc4 to the final release, only bugfixes:
711 - integrated fixes from release 2.5.2 (fix general boot flag for KVM
712 instance, fix CDROM booting for KVM instances)
713 - fixed node group modification of node parameters
714 - fixed issue in LUClusterVerifyGroup with multi-group clusters
715 - fixed generation of bash completion to ensure a stable ordering
722 *(Released Fri, 13 Jul 2012)*
724 Third release candidate for 2.6. The following changes were done from
727 - Fixed ``UpgradeConfig`` w.r.t. to disk parameters on disk objects.
728 - Fixed an inconsistency in the LUXI protocol with the provided
729 arguments (NOT backwards compatible)
730 - Fixed a bug with node groups ipolicy where ``min`` was greater than
731 the cluster ``std`` value
732 - Implemented a new ``gnt-node list-drbd`` call to list DRBD minors for
733 easier instance debugging on nodes (requires ``hconfd`` to work)
739 *(Released Tue, 03 Jul 2012)*
741 Second release candidate for 2.6. The following changes were done from
744 - Fixed ``gnt-cluster verify`` regarding ``master-ip-script`` on non
746 - Fixed a RAPI regression on missing beparams/memory
747 - Fixed redistribution of files on offline nodes
748 - Added possibility to run activate-disks even though secondaries are
749 offline. With this change it relaxes also the strictness on some other
750 commands which use activate disks internally:
751 * ``gnt-instance start|reboot|rename|backup|export``
752 - Made it possible to remove safely an instance if its secondaries are
754 - Made it possible to reinstall even though secondaries are offline
760 *(Released Mon, 25 Jun 2012)*
762 First release candidate for 2.6. The following changes were done from
765 - Fixed bugs with disk parameters and ``rbd`` templates as well as
767 - Made ``gnt-instance modify`` more consistent regarding new NIC/Disk
768 behaviour. It supports now the modify operation
769 - ``hcheck`` implemented to analyze cluster health and possibility of
770 improving health by rebalance
771 - ``hbal`` has been improved in dealing with split instances
777 *(Released Mon, 11 Jun 2012)*
779 Second beta release of 2.6. The following changes were done from beta2
782 - Fixed ``daemon-util`` with non-root user models
783 - Fixed creation of plain instances with ``--no-wait-for-sync``
784 - Fix wrong iv_names when running ``cfgupgrade``
785 - Export more information in RAPI group queries
786 - Fixed bug when changing instance network interfaces
787 - Extended burnin to do NIC changes
788 - query: Added ``<``, ``>``, ``<=``, ``>=`` comparison operators
789 - Changed default for DRBD barriers
790 - Fixed DRBD error reporting for syncer rate
791 - Verify the options on disk parameters
793 And of course various fixes to documentation and improved unittests and
800 *(Released Wed, 23 May 2012)*
802 First beta release of 2.6. The following changes were done from beta1 to
805 - integrated patch for distributions without ``start-stop-daemon``
806 - adapted example init.d script to work on Fedora
807 - fixed log handling in Haskell daemons
808 - adapted checks in the watcher for pycurl linked against libnss
809 - add partial support for ``xl`` instead of ``xm`` for Xen
810 - fixed a type issue in cluster verification
811 - fixed ssconf handling in the Haskell code (was breaking confd in IPv6
814 Plus integrated fixes from the 2.5 branch:
816 - fixed ``kvm-ifup`` to use ``/bin/bash``
817 - fixed parallel build failures
818 - KVM live migration when using a custom keymap
824 *(Released Tue, 24 Jul 2012)*
826 A small bugfix release, with no new features:
828 - fixed bash-isms in kvm-ifup, for compatibility with systems which use a
829 different default shell (e.g. Debian, Ubuntu)
830 - fixed KVM startup and live migration with a custom keymap (fixes Issue
831 243 and Debian bug #650664)
832 - fixed compatibility with KVM versions that don't support multiple boot
833 devices (fixes Issue 230 and Debian bug #624256)
835 Additionally, a few fixes were done to the build system (fixed parallel
836 build failures) and to the unittests (fixed race condition in test for
837 FileID functions, and the default enable/disable mode for QA test is now
844 *(Released Fri, 11 May 2012)*
846 A small bugfix release.
848 The main issues solved are on the topic of compatibility with newer LVM
851 - fixed parsing of ``lv_attr`` field
852 - adapted to new ``vgreduce --removemissing`` behaviour where sometimes
853 the ``--force`` flag is needed
855 Also on the topic of compatibility, ``tools/lvmstrap`` has been changed
856 to accept kernel 3.x too (was hardcoded to 2.6.*).
858 A regression present in 2.5.0 that broke handling (in the gnt-* scripts)
859 of hook results and that also made display of other errors suboptimal
860 was fixed; the code behaves now like 2.4 and earlier.
862 Another change in 2.5, the cleanup of the OS scripts environment, is too
863 aggressive: it removed even the ``PATH`` variable, which requires the OS
864 scripts to *always* need to export it. Since this is a bit too strict,
865 we now export a minimal PATH, the same that we export for hooks.
867 The fix for issue 201 (Preserve bridge MTU in KVM ifup script) was
868 integrated into this release.
870 Finally, a few other miscellaneous changes were done (no new features,
871 just small improvements):
873 - Fix ``gnt-group --help`` display
874 - Fix hardcoded Xen kernel path
875 - Fix grow-disk handling of invalid units
876 - Update synopsis for ``gnt-cluster repair-disk-sizes``
877 - Accept both PUT and POST in noded (makes future upgrade to 2.6 easier)
883 *(Released Thu, 12 Apr 2012)*
885 Incompatible/important changes and bugfixes
886 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
888 - The default of the ``/2/instances/[instance_name]/rename`` RAPI
889 resource's ``ip_check`` parameter changed from ``True`` to ``False``
890 to match the underlying LUXI interface.
891 - The ``/2/nodes/[node_name]/evacuate`` RAPI resource was changed to use
892 body parameters, see :doc:`RAPI documentation <rapi>`. The server does
893 not maintain backwards-compatibility as the underlying operation
894 changed in an incompatible way. The RAPI client can talk to old
895 servers, but it needs to be told so as the return value changed.
896 - When creating file-based instances via RAPI, the ``file_driver``
897 parameter no longer defaults to ``loop`` and must be specified.
898 - The deprecated ``bridge`` NIC parameter is no longer supported. Use
900 - Support for the undocumented and deprecated RAPI instance creation
901 request format version 0 has been dropped. Use version 1, supported
902 since Ganeti 2.1.3 and :doc:`documented <rapi>`, instead.
903 - Pyparsing 1.4.6 or above is required, see :doc:`installation
904 documentation <install>`.
905 - The "cluster-verify" hooks are now executed per group by the
906 ``OP_CLUSTER_VERIFY_GROUP`` opcode. This maintains the same behavior
907 if you just run ``gnt-cluster verify``, which generates one opcode per
909 - The environment as passed to the OS scripts is cleared, and thus no
910 environment variables defined in the node daemon's environment will be
911 inherited by the scripts.
912 - The :doc:`iallocator <iallocator>` mode ``multi-evacuate`` has been
914 - :doc:`New iallocator modes <design-multi-reloc>` have been added to
915 support operations involving multiple node groups.
916 - Offline nodes are ignored when failing over an instance.
917 - Support for KVM version 1.0, which changed the version reporting format
919 - TCP/IP ports used by DRBD disks are returned to a pool upon instance
921 - ``Makefile`` is now compatible with Automake 1.11.2
922 - Includes all bugfixes made in the 2.4 series
927 - The ganeti-htools project has been merged into the ganeti-core source
928 tree and will be built as part of Ganeti (see :doc:`install-quick`).
929 - Implemented support for :doc:`shared storage <design-shared-storage>`.
930 - Add support for disks larger than 2 TB in ``lvmstrap`` by supporting
931 GPT-style partition tables (requires `parted
932 <http://www.gnu.org/s/parted/>`_).
933 - Added support for floppy drive and 2nd CD-ROM drive in KVM hypervisor.
934 - Allowed adding tags on instance creation.
935 - Export instance tags to hooks (``INSTANCE_TAGS``, see :doc:`hooks`)
936 - Allow instances to be started in a paused state, enabling the user to
937 see the complete console output on boot using the console.
938 - Added new hypervisor flag to control default reboot behaviour
939 (``reboot_behavior``).
940 - Added support for KVM keymaps (hypervisor parameter ``keymap``).
941 - Improved out-of-band management support:
943 - Added ``gnt-node health`` command reporting the health status of
945 - Added ``gnt-node power`` command to manage power status of nodes.
946 - Added command for emergency power-off (EPO), ``gnt-cluster epo``.
948 - Instance migration can fall back to failover if instance is not
950 - Filters can be used when listing nodes, instances, groups and locks;
951 see :manpage:`ganeti(7)` manpage.
952 - Added post-execution status as variables to :doc:`hooks <hooks>`
954 - Instance tags are exported/imported together with the instance.
955 - When given an explicit job ID, ``gnt-job info`` will work for archived
957 - Jobs can define dependencies on other jobs (not yet supported via
958 RAPI or command line, but used by internal commands and usable via
961 - Lock monitor (``gnt-debug locks``) shows jobs waiting for
964 - Instance failover is now available as a RAPI resource
965 (``/2/instances/[instance_name]/failover``).
966 - ``gnt-instance info`` defaults to static information if primary node
968 - Opcodes have a new ``comment`` attribute.
969 - Added basic SPICE support to KVM hypervisor.
970 - ``tools/ganeti-listrunner`` allows passing of arguments to executable.
972 Node group improvements
973 ~~~~~~~~~~~~~~~~~~~~~~~
975 - ``gnt-cluster verify`` has been modified to check groups separately,
976 thereby improving performance.
977 - Node group support has been added to ``gnt-cluster verify-disks``,
978 which now operates per node group.
979 - Watcher has been changed to work better with node groups.
981 - One process and state file per node group.
982 - Slow watcher in one group doesn't block other group's watcher.
984 - Added new command, ``gnt-group evacuate``, to move all instances in a
985 node group to other groups.
986 - Added ``gnt-instance change-group`` to move an instance to another
988 - ``gnt-cluster command`` and ``gnt-cluster copyfile`` now support
989 per-group operations.
990 - Node groups can be tagged.
991 - Some operations switch from an exclusive to a shared lock as soon as
993 - Instance's primary and secondary nodes' groups are now available as
994 query fields (``pnode.group``, ``pnode.group.uuid``, ``snodes.group``
995 and ``snodes.group.uuid``).
1000 - Numerous updates to documentation and manpages.
1002 - :doc:`RAPI <rapi>` documentation now has detailed parameter
1004 - Some opcode/job results are now also documented, see :doc:`RAPI
1007 - A lockset's internal lock is now also visible in lock monitor.
1008 - Log messages from job queue workers now contain information about the
1009 opcode they're processing.
1010 - ``gnt-instance console`` no longer requires the instance lock.
1011 - A short delay when waiting for job changes reduces the number of LUXI
1012 requests significantly.
1013 - DRBD metadata volumes are overwritten with zeros during disk creation.
1014 - Out-of-band commands no longer acquire the cluster lock in exclusive
1016 - ``devel/upload`` now uses correct permissions for directories.
1022 *(Released Fri, 23 Mar 2012)*
1024 This was the sixth release candidate of the 2.5 series.
1030 *(Released Mon, 9 Jan 2012)*
1032 This was the fifth release candidate of the 2.5 series.
1038 *(Released Thu, 27 Oct 2011)*
1040 This was the fourth release candidate of the 2.5 series.
1046 *(Released Wed, 26 Oct 2011)*
1048 This was the third release candidate of the 2.5 series.
1054 *(Released Tue, 18 Oct 2011)*
1056 This was the second release candidate of the 2.5 series.
1062 *(Released Tue, 4 Oct 2011)*
1064 This was the first release candidate of the 2.5 series.
1070 *(Released Wed, 31 Aug 2011)*
1072 This was the third beta release of the 2.5 series.
1078 *(Released Mon, 22 Aug 2011)*
1080 This was the second beta release of the 2.5 series.
1086 *(Released Fri, 12 Aug 2011)*
1088 This was the first beta release of the 2.5 series.
1094 *(Released Thu, 27 Oct 2011)*
1096 - Fixed bug when parsing command line parameter values ending in
1098 - Fixed assertion error after unclean master shutdown
1099 - Disable HTTP client pool for RPC, significantly reducing memory usage
1101 - Fixed queue archive creation with wrong permissions
1107 *(Released Tue, 23 Aug 2011)*
1111 - Fixed documentation for importing with ``--src-dir`` option
1112 - Fixed a bug in ``ensure-dirs`` with queue/archive permissions
1113 - Fixed a parsing issue with DRBD 8.3.11 in the Linux kernel
1119 *(Released Fri, 5 Aug 2011)*
1121 Many bug-fixes and a few small features:
1123 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
1124 issues when you tried to add an instance with two MAC addresses in one
1126 - KVM: fixed per-instance stored UID value
1127 - KVM: configure bridged NICs at migration start
1128 - KVM: Fix a bug where instance will not start with never KVM versions
1130 - Added OS search path to ``gnt-cluster info``
1131 - Fixed an issue with ``file_storage_dir`` where you were forced to
1132 provide an absolute path, but the documentation states it is a
1133 relative path, the documentation was right
1134 - Added a new parameter to instance stop/start called ``--no-remember``
1135 that will make the state change to not be remembered
1136 - Implemented ``no_remember`` at RAPI level
1137 - Improved the documentation
1138 - Node evacuation: don't call IAllocator if node is already empty
1139 - Fixed bug in DRBD8 replace disks on current nodes
1140 - Fixed bug in recreate-disks for DRBD instances
1141 - Moved assertion checking locks in ``gnt-instance replace-disks``
1142 causing it to abort with not owning the right locks for some situation
1143 - Job queue: Fixed potential race condition when cancelling queued jobs
1144 - Fixed off-by-one bug in job serial generation
1145 - ``gnt-node volumes``: Fix instance names
1146 - Fixed aliases in bash completion
1147 - Fixed a bug in reopening log files after being sent a SIGHUP
1148 - Added a flag to burnin to allow specifying VCPU count
1149 - Bugfixes to non-root Ganeti configuration
1155 *(Released Thu, 12 May 2011)*
1157 Many bug-fixes and a few new small features:
1159 - Fixed a bug related to log opening failures
1160 - Fixed a bug in instance listing with orphan instances
1161 - Fixed a bug which prevented resetting the cluster-level node parameter
1162 ``oob_program`` to the default
1163 - Many fixes related to the ``cluster-merge`` tool
1164 - Fixed a race condition in the lock monitor, which caused failures
1165 during (at least) creation of many instances in parallel
1166 - Improved output for gnt-job info
1167 - Removed the quiet flag on some ssh calls which prevented debugging
1169 - Improved the N+1 failure messages in cluster verify by actually
1170 showing the memory values (needed and available)
1171 - Increased lock attempt timeouts so that when executing long operations
1172 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
1173 too early and thus prevent the use of the 'fair' mechanism
1174 - Changed instance query data (``gnt-instance info``) to not acquire
1175 locks unless needed, thus allowing its use on locked instance if only
1176 static information is asked for
1177 - Improved behaviour with filesystems that do not support rename on an
1179 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
1180 kept locks on all nodes during the wipe, which is unneeded
1181 - Fixed ``gnt-watcher`` handling of errors during hooks execution
1182 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
1183 10GiB) which caused the wipe to fail right at the end in some cases
1184 - Fixed master IP activation when doing master failover with no-voting
1185 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
1186 the master node itself
1187 - Fixed potential data-loss in under disk full conditions, where Ganeti
1188 wouldn't check correctly the return code and would consider
1189 partially-written files 'correct'
1190 - Fixed bug related to multiple VGs and DRBD disk replacing
1191 - Added new disk parameter ``metavg`` that allows placement of the meta
1192 device for DRBD in a different volume group
1193 - Fixed error handling in the node daemon when the system libc doesn't
1194 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
1195 - Fixed lock release during replace-disks, which kept cluster-wide locks
1196 when doing disk replaces with an iallocator script
1197 - Added check for missing bridges in cluster verify
1198 - Handle EPIPE errors while writing to the terminal better, so that
1199 piping the output to e.g. ``less`` doesn't cause a backtrace
1200 - Fixed rare case where a ^C during Luxi calls could have been
1201 interpreted as server errors, instead of simply terminating
1202 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
1204 - Added a few more parameters to the KVM hypervisor, allowing a second
1205 CDROM, custom disk type for CDROMs and a floppy image
1206 - Removed redundant message in instance rename when the name is given
1208 - Added option to ``gnt-instance recreate-disks`` to allow creating the
1209 disks on new nodes, allowing recreation when the original instance
1210 nodes are completely gone
1211 - Added option when converting disk templates to DRBD to skip waiting
1212 for the resync, in order to make the instance available sooner
1213 - Added two new variables to the OS scripts environment (containing the
1215 - Made the root_path and optional parameter for the xen-pvm hypervisor,
1216 to allow use of ``pvgrub`` as bootloader
1217 - Changed the instance memory modifications to only check out-of-memory
1218 conditions on memory increases, and turned the secondary node warnings
1219 into errors (they can still be overridden via ``--force``)
1220 - Fixed the handling of a corner case when the Python installation gets
1221 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
1222 to execute a command that doesn't exist
1223 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
1224 node of the instance returned failures during instance shutdown; this
1225 adds the option ``--ignore-consistency`` to gnt-instance move
1227 And as usual, various improvements to the error messages, documentation
1234 *(Released Wed, 09 Mar 2011)*
1236 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
1237 the RAPI users file if run twice (even with ``--dry-run``).
1239 The release fixes that bug (nothing else changed).
1245 *(Released Mon, 07 Mar 2011)*
1247 Final 2.4.0 release. Just a few small fixes:
1249 - Fixed RAPI node evacuate
1250 - Fixed the kvm-ifup script
1251 - Fixed internal error handling for special job cases
1252 - Updated man page to specify the escaping feature for options
1258 *(Released Mon, 28 Feb 2011)*
1260 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
1261 that this feature wiped the disks of the wrong instance, leading to loss
1266 - Fixed title of query field containing instance name
1267 - Expanded the glossary in the documentation
1268 - Fixed one unittest (internal issue)
1274 *(Released Mon, 21 Feb 2011)*
1276 A number of bug fixes plus just a couple functionality changes.
1278 On the user-visible side, the ``gnt-* list`` command output has changed
1279 with respect to "special" field states. The current rc1 style of display
1280 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
1281 the default output mode special fields states are displayed as follows:
1283 - Offline resource: ``*``
1284 - Unavailable/not applicable: ``-``
1285 - Data missing (RPC failure): ``?``
1286 - Unknown field: ``??``
1288 Another user-visible change is the addition of ``--force-join`` to
1293 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
1294 - Fixed regression in RAPI/instance reinstall where all parameters were
1295 required (instead of optional)
1296 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
1297 - Fixed iallocator usage (offline nodes were not considered offline)
1298 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
1299 - Fixed hypervisor and OS parameter validation with respect to
1300 non-vm_capable nodes
1301 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
1303 - Fixed ``tools/listrunner`` with respect to agent-based usage
1309 *(Released Fri, 4 Feb 2011)*
1311 Many changes and fixes since the beta1 release. While there were some
1312 internal changes, the code has been mostly stabilised for the RC
1315 Note: the dumb allocator was removed in this release, as it was not kept
1316 up-to-date with the IAllocator protocol changes. It is recommended to
1317 use the ``hail`` command from the ganeti-htools package.
1319 Note: the 2.4 and up versions of Ganeti are not compatible with the
1320 0.2.x branch of ganeti-htools. You need to upgrade to
1321 ganeti-htools-0.3.0 (or later).
1323 Regressions fixed from 2.3
1324 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1326 - Fixed the ``gnt-cluster verify-disks`` command
1327 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
1329 - Fixed disk adoption breakage
1330 - Fixed wrong headers in instance listing for field aliases
1335 - Fixed corner case in KVM handling of NICs
1336 - Fixed many cases of wrong handling of non-vm_capable nodes
1337 - Fixed a bug where a missing instance symlink was not possible to
1338 recreate with any ``gnt-*`` command (now ``gnt-instance
1339 activate-disks`` does it)
1340 - Fixed the volume group name as reported by ``gnt-cluster
1342 - Increased timeouts for the import-export code, hopefully leading to
1343 fewer aborts due network or instance timeouts
1344 - Fixed bug in ``gnt-node list-storage``
1345 - Fixed bug where not all daemons were started on cluster
1346 initialisation, but only at the first watcher run
1347 - Fixed many bugs in the OOB implementation
1348 - Fixed watcher behaviour in presence of instances with offline
1350 - Fixed instance list output for instances running on the wrong node
1351 - a few fixes to the cluster-merge tool, but it still cannot merge
1352 multi-node groups (currently it is not recommended to use this tool)
1358 - Improved network configuration for the KVM hypervisor
1359 - Added e1000 as a supported NIC for Xen-HVM
1360 - Improved the lvmstrap tool to also be able to use partitions, as
1361 opposed to full disks
1362 - Improved speed of disk wiping (the cluster parameter
1363 ``prealloc_wipe_disks``, so that it has a low impact on the total time
1364 of instance creations
1365 - Added documentation for the OS parameters
1366 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
1367 hypervisor is not responding
1368 - Added display of blacklisted and hidden OS information in
1369 ``gnt-cluster info``
1370 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
1371 NIC and node parameters, which might create problems with currently
1372 invalid (but undetected) configuration files, but prevents validation
1373 failures when unrelated parameters are modified
1374 - Changed cluster initialisation to wait for the master daemon to become
1376 - Expanded the RAPI interface:
1378 - Added config redistribution resource
1379 - Added activation/deactivation of instance disks
1380 - Added export of console information
1382 - Implemented log file reopening on SIGHUP, which allows using
1383 logrotate(8) for the Ganeti log files
1384 - Added a basic OOB helper script as an example
1390 *(Released Fri, 14 Jan 2011)*
1395 - Fixed timezone issues when formatting timestamps
1396 - Added support for node groups, available via ``gnt-group`` and other
1398 - Added out-of-band framework and management, see :doc:`design
1399 document <design-oob>`
1400 - Removed support for roman numbers from ``gnt-node list`` and
1401 ``gnt-instance list``.
1402 - Allowed modification of master network interface via ``gnt-cluster
1403 modify --master-netdev``
1404 - Accept offline secondaries while shutting down instance disks
1405 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
1406 - Added support for multiple LVM volume groups
1407 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
1409 - Added commands to list available fields:
1411 - ``gnt-node list-fields``
1412 - ``gnt-group list-fields``
1413 - ``gnt-instance list-fields``
1415 - Updated documentation and man pages
1420 - Moved ``rapi_users`` file into separate directory, now named
1421 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
1423 - Added new tool for running commands on many machines,
1424 ``tools/ganeti-listrunner``
1425 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
1426 improving the ``gnt-instance console`` output
1427 - Allowed customisation of disk index separator at ``configure`` time
1428 - Export node group allocation policy to :doc:`iallocator <iallocator>`
1429 - Added support for non-partitioned md disks in ``lvmstrap``
1430 - Added script to gracefully power off KVM instances
1431 - Split ``utils`` module into smaller parts
1432 - Changed query operations to return more detailed information, e.g.
1433 whether an information is unavailable due to an offline node. To use
1434 this new functionality, the LUXI call ``Query`` must be used. Field
1435 information is now stored by the master daemon and can be retrieved
1436 using ``QueryFields``. Instances, nodes and groups can also be queried
1437 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
1438 exposed via RAPI). The following commands make use of this
1439 infrastructure change:
1441 - ``gnt-group list``
1442 - ``gnt-group list-fields``
1444 - ``gnt-node list-fields``
1445 - ``gnt-instance list``
1446 - ``gnt-instance list-fields``
1447 - ``gnt-debug locks``
1452 - New RAPI resources (see :doc:`rapi`):
1456 - ``/2/groups/[group_name]``
1457 - ``/2/groups/[group_name]/assign-nodes``
1458 - ``/2/groups/[group_name]/modify``
1459 - ``/2/groups/[group_name]/rename``
1460 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
1464 - Implemented ``no_install`` for instance creation
1465 - Implemented OS parameters for instance reinstallation, allowing
1466 use of special settings on reinstallation (e.g. for preserving data)
1471 - Added IPv6 support in import/export
1472 - Pause DRBD synchronization while wiping disks on instance creation
1473 - Updated unittests and QA scripts
1474 - Improved network parameters passed to KVM
1475 - Converted man pages from docbook to reStructuredText
1481 *(Released Mon, 20 Dec 2010)*
1483 Released version 2.3.1~rc1 without any changes.
1489 *(Released Wed, 1 Dec 2010)*
1491 - impexpd: Disable OpenSSL compression in socat if possible (backport
1492 from master, commit e90739d625b, see :doc:`installation guide
1493 <install-quick>` for details)
1494 - Changed unittest coverage report to exclude test scripts
1495 - Added script to check version format
1501 *(Released Wed, 1 Dec 2010)*
1503 Released version 2.3.0~rc1 without any changes.
1509 *(Released Fri, 19 Nov 2010)*
1511 A number of bugfixes and documentation updates:
1513 - Update ganeti-os-interface documentation
1514 - Fixed a bug related to duplicate MACs or similar items which should be
1516 - Fix breakage in OS state modify
1517 - Reinstall instance: disallow offline secondaries (fixes bug related to
1518 OS changing but reinstall failing)
1519 - plus all the other fixes between 2.2.1 and 2.2.2
1525 *(Released Tue, 2 Nov 2010)*
1527 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
1528 - Fixed master failover race with watcher
1529 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
1531 - Accept previously stopped instance for export with instance removal
1532 - Simplify and extend the environment variables for instance OS scripts
1533 - Added new node flags, ``master_capable`` and ``vm_capable``
1534 - Added optional instance disk wiping prior during allocation. This is a
1535 cluster-wide option and can be set/modified using
1536 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
1537 - Added IPv6 support, see :doc:`design document <design-2.3>` and
1538 :doc:`install-quick`
1539 - Added a new watcher option (``--ignore-pause``)
1540 - Added option to ignore offline node on instance start/stop
1541 (``--ignore-offline``)
1542 - Allow overriding OS parameters with ``gnt-instance reinstall``
1543 - Added ability to change node's secondary IP address using ``gnt-node
1545 - Implemented privilege separation for all daemons except
1546 ``ganeti-noded``, see ``configure`` options
1547 - Complain if an instance's disk is marked faulty in ``gnt-cluster
1549 - Implemented job priorities (see ``ganeti(7)`` manpage)
1550 - Ignore failures while shutting down instances during failover from
1552 - Exit daemon's bootstrap process only once daemon is ready
1553 - Export more information via ``LUInstanceQuery``/remote API
1554 - Improved documentation, QA and unittests
1555 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
1556 restart if the file was created or changed
1557 - Added LUXI protocol version sent with each request and response,
1558 allowing detection of server/client mismatches
1559 - Moved the Python scripts among gnt-* and ganeti-* into modules
1560 - Moved all code related to setting up SSH to an external script,
1562 - Infrastructure changes for node group support in future versions
1568 *(Released Fri, 19 Nov 2010)*
1570 A few small bugs fixed, and some improvements to the build system:
1572 - Fix documentation regarding conversion to drbd
1573 - Fix validation of parameters in cluster modify (``gnt-cluster modify
1575 - Fix error handling in node modify with multiple changes
1576 - Allow remote imports without checked names
1582 *(Released Tue, 19 Oct 2010)*
1584 - Disable SSL session ID cache in RPC client
1590 *(Released Thu, 14 Oct 2010)*
1592 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
1593 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
1595 - Fix problems with interaction between Curl and Python's HTTP server,
1596 resulting in increased speed in many RPC calls
1597 - Improve our release script to prevent breakage with older aclocal and
1604 *(Released Thu, 7 Oct 2010)*
1606 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
1607 value retrieved from master
1608 - Added support for blacklisted or hidden OS definitions
1609 - Added simple lock monitor (accessible via (``gnt-debug locks``)
1610 - Added support for -mem-path in KVM hypervisor abstraction layer
1611 - Allow overriding instance parameters in tool for inter-cluster
1612 instance moves (``tools/move-instance``)
1613 - Improved opcode summaries (e.g. in ``gnt-job list``)
1614 - Improve consistency of OS listing by sorting it
1615 - Documentation updates
1621 *(Released Fri, 8 Oct 2010)*
1623 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
1629 *(Released Mon, 4 Oct 2010)*
1631 - Fixed regression in ``gnt-instance rename``
1637 *(Released Wed, 22 Sep 2010)*
1639 - Fixed OS_VARIANT variable for OS scripts
1640 - Fixed cluster tag operations via RAPI
1641 - Made ``setup-ssh`` exit with non-zero code if an error occurred
1642 - Disabled RAPI CA checks in watcher
1648 *(Released Mon, 23 Aug 2010)*
1650 - Support DRBD versions of the format "a.b.c.d"
1652 - Re-introduce support for usage from multiple threads in RAPI client
1653 - Instance renames and modify via RAPI
1654 - Work around race condition between processing and archival in job
1656 - Mark opcodes following failed one as failed, too
1657 - Job field ``lock_status`` was removed due to difficulties making it
1658 work with the changed job queue in Ganeti 2.2; a better way to monitor
1659 locks is expected for a later 2.2.x release
1660 - Fixed dry-run behaviour with many commands
1661 - Support ``ssh-agent`` again when adding nodes
1662 - Many additional bugfixes
1668 *(Released Fri, 30 Jul 2010)*
1670 Important change: the internal RPC mechanism between Ganeti nodes has
1671 changed from using a home-grown http library (based on the Python base
1672 libraries) to use the PycURL library. This requires that PycURL is
1673 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
1674 against GnuTLS by default. cURL's support for GnuTLS had known issues
1675 before cURL 7.21.0 and we recommend using the latest cURL release or
1676 linking against OpenSSL. Most other distributions already link PycURL
1677 and cURL against OpenSSL. The command::
1679 python -c 'import pycurl; print pycurl.version'
1681 can be used to determine the libraries PycURL and cURL are linked
1684 Other significant changes:
1686 - Rewrote much of the internals of the job queue, in order to achieve
1687 better parallelism; this decouples job query operations from the job
1688 processing, and it should allow much nicer behaviour of the master
1689 daemon under load, and it also has uncovered some long-standing bugs
1690 related to the job serialisation (now fixed)
1691 - Added a default iallocator setting to the cluster parameters,
1692 eliminating the need to always pass nodes or an iallocator for
1693 operations that require selection of new node(s)
1694 - Added experimental support for the LXC virtualization method
1695 - Added support for OS parameters, which allows the installation of
1696 instances to pass parameter to OS scripts in order to customise the
1698 - Added a hypervisor parameter controlling the migration type (live or
1699 non-live), since hypervisors have various levels of reliability; this
1700 has renamed the 'live' parameter to 'mode'
1701 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
1702 logical volumes, meaning that cluster verify will ignore them and not
1703 flag their presence as errors
1704 - The watcher will now reset the error count for failed instances after
1705 8 hours, thus allowing self-healing if the problem that caused the
1706 instances to be down/fail to start has cleared in the meantime
1707 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
1708 check for, and warn, if the drbd module parameter ``usermode_helper``
1709 is not consistent with the cluster-wide setting; this is needed to
1710 make diagnose easier of failed drbd creations
1711 - Started adding base IPv6 support, but this is not yet
1712 enabled/available for use
1713 - Rename operations (cluster, instance) will now return the new name,
1714 which is especially useful if a short name was passed in
1715 - Added support for instance migration in RAPI
1716 - Added a tool to pre-configure nodes for the SSH setup, before joining
1717 them to the cluster; this will allow in the future a simplified model
1718 for node joining (but not yet fully enabled in 2.2); this needs the
1719 paramiko python library
1720 - Fixed handling of name-resolving errors
1721 - Fixed consistency of job results on the error path
1722 - Fixed master-failover race condition when executed multiple times in
1724 - Fixed many bugs related to the job queue (mostly introduced during the
1725 2.2 development cycle, so not all are impacting 2.1)
1726 - Fixed instance migration with missing disk symlinks
1727 - Fixed handling of unknown jobs in ``gnt-job archive``
1728 - And many other small fixes/improvements
1732 - Enhanced both the unittest and the QA coverage
1733 - Switched the opcode validation to a generic model, and extended the
1734 validation to all opcode parameters
1735 - Changed more parts of the code that write shell scripts to use the
1737 - Switched the master daemon to use the asyncore library for the Luxi
1744 *(Released Thu, 17 Jun 2010)*
1746 - Added tool (``move-instance``) and infrastructure to move instances
1747 between separate clusters (see :doc:`separate documentation
1748 <move-instance>` and :doc:`design document <design-2.2>`)
1749 - Added per-request RPC timeout
1750 - RAPI now requires a Content-Type header for requests with a body (e.g.
1751 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
1752 :rfc:`2616` (HTTP/1.1), section 7.2.1)
1753 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
1755 - Implemented initial support for running Ganeti daemons as separate
1756 users, see configure-time flags ``--with-user-prefix`` and
1757 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
1759 - Instances can be removed after export (``gnt-backup export
1760 --remove-instance``)
1761 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
1762 key (instead of 1024 bit)
1763 - Added new cluster configuration file for cluster domain secret
1764 - Import/export now use SSL instead of SSH
1765 - Added support for showing estimated time when exporting an instance,
1766 see the ``ganeti-os-interface(7)`` manpage and look for
1773 *(Released Tue, 16 Nov 2010)*
1775 Some more bugfixes. Unless critical bugs occur, this will be the last
1778 - Fix case of MAC special-values
1779 - Fix mac checker regex
1780 - backend: Fix typo causing "out of range" error
1781 - Add missing --units in gnt-instance list man page
1787 *(Released Tue, 24 Aug 2010)*
1790 - Don't ignore secondary node silently on non-mirrored disk templates
1792 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
1793 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
1794 - Properly document the usb_mouse parameter
1795 - Fix path in ganeti-rapi(8) (issue 116)
1796 - Adjust error message when the ganeti user's .ssh directory is
1798 - Add same-node-check when changing the disk template to drbd
1804 *(Released Fri, 16 Jul 2010)*
1807 - Add an option to only select some reboot types during qa/burnin.
1808 (on some hypervisors consequent reboots are not supported)
1809 - Fix infrequent race condition in master failover. Sometimes the old
1810 master ip address would be still detected as up for a short time
1811 after it was removed, causing failover to fail.
1812 - Decrease mlockall warnings when the ctypes module is missing. On
1813 Python 2.4 we support running even if no ctypes module is installed,
1814 but we were too verbose about this issue.
1815 - Fix building on old distributions, on which man doesn't have a
1817 - Fix RAPI not to ignore the MAC address on instance creation
1818 - Implement the old instance creation format in the RAPI client.
1824 *(Released Thu, 01 Jul 2010)*
1826 A small bugfix release:
1827 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
1828 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
1829 a non-existing option
1830 - Fix instance create: the --force-variant option was ignored
1831 - Improve pylint 0.21 compatibility and warnings with Python 2.6
1832 - Fix modify node storage with non-FQDN arguments
1833 - Fix RAPI client to authenticate under Python 2.6 when used
1834 for more than 5 requests needing authentication
1835 - Fix gnt-instance modify -t (storage) giving a wrong error message
1836 when converting a non-shutdown drbd instance to plain
1842 *(Released Fri, 18 Jun 2010)*
1844 A small bugfix release:
1846 - Fix live migration of KVM instances started with older Ganeti
1847 versions which had fewer hypervisor parameters
1848 - Fix gnt-instance grow-disk on down instances
1849 - Fix an error-reporting bug during instance migration
1850 - Better checking of the ``--net`` and ``--disk`` values, to avoid
1851 silently ignoring broken ones
1852 - Fix an RPC error reporting bug affecting, for example, RAPI client
1854 - Fix bug triggered by different API version os-es on different nodes
1855 - Fix a bug in instance startup with custom hvparams: OS level
1856 parameters would fail to be applied.
1857 - Fix the RAPI client under Python 2.6 (but more work is needed to
1858 make it work completely well with OpenSSL)
1859 - Fix handling of errors when resolving names from DNS
1865 *(Released Thu, 3 Jun 2010)*
1867 A medium sized development cycle. Some new features, and some
1868 fixes/small improvements/cleanups.
1870 Significant features
1871 ~~~~~~~~~~~~~~~~~~~~
1873 The node deamon now tries to mlock itself into memory, unless the
1874 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
1875 its logs, and falls back to console logging. This allows emergency
1876 features such as ``gnt-node powercycle`` to work even in the event of a
1877 broken node disk (tested offlining the disk hosting the node's
1878 filesystem and dropping its memory caches; don't try this at home)
1880 KVM: add vhost-net acceleration support. It can be tested with a new
1881 enough version of the kernel and of qemu-kvm.
1883 KVM: Add instance chrooting feature. If you use privilege dropping for
1884 your VMs you can also now force them to chroot to an empty directory,
1885 before starting the emulated guest.
1887 KVM: Add maximum migration bandwith and maximum downtime tweaking
1888 support (requires a new-enough version of qemu-kvm).
1890 Cluster verify will now warn if the master node doesn't have the master
1891 ip configured on it.
1893 Add a new (incompatible) instance creation request format to RAPI which
1894 supports all parameters (previously only a subset was supported, and it
1895 wasn't possible to extend the old format to accomodate all the new
1896 features. The old format is still supported, and a client can check for
1897 this feature, before using it, by checking for its presence in the
1898 ``features`` RAPI resource.
1900 Now with ancient latin support. Try it passing the ``--roman`` option to
1901 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
1902 (requires the python-roman module to be installed, in order to work).
1907 As usual many internal code refactorings, documentation updates, and
1910 - Lots of improvements and cleanups to the experimental Remote API
1911 (RAPI) client library.
1912 - A new unit test suite for the core daemon libraries.
1913 - A fix to creating missing directories makes sure the umask is not
1914 applied anymore. This enforces the same directory permissions
1916 - Better handling terminating daemons with ctrl+c (used when running
1917 them in debugging mode).
1918 - Fix a race condition in live migrating a KVM instance, when stat()
1919 on the old proc status file returned EINVAL, which is an unexpected
1921 - Fixed manpage checking with newer man and utf-8 charachters. But now
1922 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
1928 *(Released Fri, 7 May 2010)*
1930 Fix a bug which prevented untagged KVM instances from starting.
1936 *(Released Fri, 7 May 2010)*
1938 Another release with a long development cycle, during which many
1939 different features were added.
1941 Significant features
1942 ~~~~~~~~~~~~~~~~~~~~
1944 The KVM hypervisor now can run the individual instances as non-root, to
1945 reduce the impact of a VM being hijacked due to bugs in the
1946 hypervisor. It is possible to run all instances as a single (non-root)
1947 user, to manually specify a user for each instance, or to dynamically
1948 allocate a user out of a cluster-wide pool to each instance, with the
1949 guarantee that no two instances will run under the same user ID on any
1952 An experimental RAPI client library, that can be used standalone
1953 (without the other Ganeti libraries), is provided in the source tree as
1954 ``lib/rapi/client.py``. Note this client might change its interface in
1955 the future, as we iterate on its capabilities.
1957 A new command, ``gnt-cluster renew-crypto`` has been added to easily
1958 replace the cluster's certificates and crypto keys. This might help in
1959 case they have been compromised, or have simply expired.
1961 A new disk option for instance creation has been added that allows one
1962 to "adopt" currently existing logical volumes, with data
1963 preservation. This should allow easier migration to Ganeti from
1964 unmanaged (or managed via other software) instances.
1966 Another disk improvement is the possibility to convert between redundant
1967 (DRBD) and plain (LVM) disk configuration for an instance. This should
1968 allow better scalability (starting with one node and growing the
1969 cluster, or shrinking a two-node cluster to one node).
1971 A new feature that could help with automated node failovers has been
1972 implemented: if a node sees itself as offline (by querying the master
1973 candidates), it will try to shutdown (hard) all instances and any active
1974 DRBD devices. This reduces the risk of duplicate instances if an
1975 external script automatically failovers the instances on such nodes. To
1976 enable this, the cluster parameter ``maintain_node_health`` should be
1977 enabled; in the future this option (per the name) will enable other
1978 automatic maintenance features.
1980 Instance export/import now will reuse the original instance
1981 specifications for all parameters; that means exporting an instance,
1982 deleting it and the importing it back should give an almost identical
1983 instance. Note that the default import behaviour has changed from
1984 before, where it created only one NIC; now it recreates the original
1987 Cluster verify has added a few new checks: SSL certificates validity,
1988 /etc/hosts consistency across the cluster, etc.
1993 As usual, many internal changes were done, documentation fixes,
1996 - Fixed cluster initialization with disabled cluster storage (regression
1997 introduced in 2.1.1)
1998 - File-based storage supports growing the disks
1999 - Fixed behaviour of node role changes
2000 - Fixed cluster verify for some corner cases, plus a general rewrite of
2001 cluster verify to allow future extension with more checks
2002 - Fixed log spamming by watcher and node daemon (regression introduced
2004 - Fixed possible validation issues when changing the list of enabled
2006 - Fixed cleanup of /etc/hosts during node removal
2007 - Fixed RAPI response for invalid methods
2008 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
2009 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
2010 from ide disks, etc.)
2011 - Allow OS changes without re-installation (to record a changed OS
2012 outside of Ganeti, or to allow OS renames)
2013 - Allow instance creation without OS installation (useful for example if
2014 the OS will be installed manually, or restored from a backup not in
2016 - Implemented option to make cluster ``copyfile`` use the replication
2018 - Added list of enabled hypervisors to ssconf (possibly useful for
2020 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
2022 - A partial form of node re-IP is possible via node readd, which now
2023 allows changed node primary IP
2024 - Command line utilities now show an informational message if the job is
2026 - The logs of the master daemon now show the PID/UID/GID of the
2033 *(Released Fri, 12 Mar 2010)*
2035 During the 2.1.0 long release candidate cycle, a lot of improvements and
2036 changes have accumulated with were released later as 2.1.1.
2041 The node evacuate command (``gnt-node evacuate``) was significantly
2042 rewritten, and as such the IAllocator protocol was changed - a new
2043 request type has been added. This unfortunate change during a stable
2044 series is designed to improve performance of node evacuations; on
2045 clusters with more than about five nodes and which are well-balanced,
2046 evacuation should proceed in parallel for all instances of the node
2047 being evacuated. As such, any existing IAllocator scripts need to be
2048 updated, otherwise the above command will fail due to the unknown
2049 request. The provided "dumb" allocator has not been updated; but the
2050 ganeti-htools package supports the new protocol since version 0.2.4.
2052 Another important change is increased validation of node and instance
2053 names. This might create problems in special cases, if invalid host
2054 names are being used.
2056 Also, a new layer of hypervisor parameters has been added, that sits at
2057 OS level between the cluster defaults and the instance ones. This allows
2058 customisation of virtualization parameters depending on the installed
2059 OS. For example instances with OS 'X' may have a different KVM kernel
2060 (or any other parameter) than the cluster defaults. This is intended to
2061 help managing a multiple OSes on the same cluster, without manual
2062 modification of each instance's parameters.
2064 A tool for merging clusters, ``cluster-merge``, has been added in the
2065 tools sub-directory.
2070 - Improved the int/float conversions that should make the code more
2071 robust in face of errors from the node daemons
2072 - Fixed the remove node code in case of internal configuration errors
2073 - Fixed the node daemon behaviour in face of inconsistent queue
2074 directory (e.g. read-only file-system where we can't open the files
2076 - Fixed the behaviour of gnt-node modify for master candidate demotion;
2077 now it either aborts cleanly or, if given the new "auto_promote"
2078 parameter, will automatically promote other nodes as needed
2079 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
2080 completely prevent Ganeti from working
2081 - Fixed bug for instance export when not all disks were successfully
2083 - Fixed behaviour of node add when the new node is slow in starting up
2085 - Fixed handling of signals in the LUXI client, which should improve
2086 behaviour of command-line scripts
2087 - Added checks for invalid node/instance names in the configuration (now
2088 flagged during cluster verify)
2089 - Fixed watcher behaviour for disk activation errors
2090 - Fixed two potentially endless loops in http library, which led to the
2091 RAPI daemon hanging and consuming 100% CPU in some cases
2092 - Fixed bug in RAPI daemon related to hashed passwords
2093 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
2095 - Enhanced compatibility with non-Debian OSes, but not using absolute
2096 path in some commands and allowing customisation of the ssh
2097 configuration directory
2098 - Fixed possible future issue with new Python versions by abiding to the
2099 proper use of ``__slots__`` attribute on classes
2100 - Added checks that should prevent directory traversal attacks
2101 - Many documentation fixes based on feedback from users
2106 - Added an "early_release" more for instance replace disks and node
2107 evacuate, where we release locks earlier and thus allow higher
2108 parallelism within the cluster
2109 - Added watcher hooks, intended to allow the watcher to restart other
2110 daemons (e.g. from the ganeti-nbma project), but they can be used of
2111 course for any other purpose
2112 - Added a compile-time disable for DRBD barriers, to increase
2113 performance if the administrator trusts the power supply or the
2114 storage system to not lose writes
2115 - Added the option of using syslog for logging instead of, or in
2116 addition to, Ganeti's own log files
2117 - Removed boot restriction for paravirtual NICs for KVM, recent versions
2118 can indeed boot from a paravirtual NIC
2119 - Added a generic debug level for many operations; while this is not
2120 used widely yet, it allows one to pass the debug value all the way to
2122 - Enhanced the hooks environment for instance moves (failovers,
2123 migrations) where the primary/secondary nodes changed during the
2124 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
2125 - Enhanced data validations for many user-supplied values; one important
2126 item is the restrictions imposed on instance and node names, which
2127 might reject some (invalid) host names
2128 - Add a configure-time option to disable file-based storage, if it's not
2129 needed; this allows greater security separation between the master
2130 node and the other nodes from the point of view of the inter-node RPC
2132 - Added user notification in interactive tools if job is waiting in the
2133 job queue or trying to acquire locks
2134 - Added log messages when a job is waiting for locks
2135 - Added filtering by node tags in instance operations which admit
2136 multiple instances (start, stop, reboot, reinstall)
2137 - Added a new tool for cluster mergers, ``cluster-merge``
2138 - Parameters from command line which are of the form ``a=b,c=d`` can now
2139 use backslash escapes to pass in values which contain commas,
2140 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
2142 - For KVM, the instance name is the first parameter passed to KVM, so
2143 that it's more visible in the process list
2149 *(Released Tue, 2 Mar 2010)*
2151 Ganeti 2.1 brings many improvements with it. Major changes:
2153 - Added infrastructure to ease automated disk repairs
2154 - Added new daemon to export configuration data in a cheaper way than
2155 using the remote API
2156 - Instance NICs can now be routed instead of being associated with a
2158 - Improved job locking logic to reduce impact of jobs acquiring multiple
2159 locks waiting for other long-running jobs
2161 In-depth implementation details can be found in the Ganeti 2.1 design
2167 - Added chroot hypervisor
2168 - Added more options to xen-hvm hypervisor (``kernel_path`` and
2170 - Added more options to xen-pvm hypervisor (``use_bootloader``,
2171 ``bootloader_path`` and ``bootloader_args``)
2172 - Added the ``use_localtime`` option for the xen-hvm and kvm
2173 hypervisors, and the default value for this has changed to false (in
2174 2.0 xen-hvm always enabled it)
2175 - Added luxi call to submit multiple jobs in one go
2176 - Added cluster initialization option to not modify ``/etc/hosts``
2178 - Added network interface parameters
2179 - Added dry run mode to some LUs
2180 - Added RAPI resources:
2182 - ``/2/instances/[instance_name]/info``
2183 - ``/2/instances/[instance_name]/replace-disks``
2184 - ``/2/nodes/[node_name]/evacuate``
2185 - ``/2/nodes/[node_name]/migrate``
2186 - ``/2/nodes/[node_name]/role``
2187 - ``/2/nodes/[node_name]/storage``
2188 - ``/2/nodes/[node_name]/storage/modify``
2189 - ``/2/nodes/[node_name]/storage/repair``
2191 - Added OpCodes to evacuate or migrate all instances on a node
2192 - Added new command to list storage elements on nodes (``gnt-node
2193 list-storage``) and modify them (``gnt-node modify-storage``)
2194 - Added new ssconf files with master candidate IP address
2195 (``ssconf_master_candidates_ips``), node primary IP address
2196 (``ssconf_node_primary_ips``) and node secondary IP address
2197 (``ssconf_node_secondary_ips``)
2198 - Added ``ganeti-confd`` and a client library to query the Ganeti
2199 configuration via UDP
2200 - Added ability to run hooks after cluster initialization and before
2202 - Added automatic mode for disk replace (``gnt-instance replace-disks
2204 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
2205 after catastrophic data-loss
2206 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
2208 - Added ``gnt-instance move`` command to move instances
2209 - Added ``gnt-cluster watcher`` command to control watcher
2210 - Added ``gnt-node powercycle`` command to powercycle nodes
2211 - Added new job status field ``lock_status``
2212 - Added parseable error codes to cluster verification (``gnt-cluster
2213 verify --error-codes``) and made output less verbose (use
2214 ``--verbose`` to restore previous behaviour)
2215 - Added UUIDs to the main config entities (cluster, nodes, instances)
2216 - Added support for OS variants
2217 - Added support for hashed passwords in the Ganeti remote API users file
2219 - Added option to specify maximum timeout on instance shutdown
2220 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
2221 - Added new helper script to start and stop Ganeti daemons
2222 (``daemon-util``), with the intent to reduce the work necessary to
2223 adjust Ganeti for non-Debian distributions and to start/stop daemons
2225 - Added more unittests
2226 - Fixed critical bug in ganeti-masterd startup
2227 - Removed the configure-time ``kvm-migration-port`` parameter, this is
2228 now customisable at the cluster level for both the KVM and Xen
2229 hypervisors using the new ``migration_port`` parameter
2230 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
2231 reinstalling an instance
2232 - Allowed ``@`` in tag names
2233 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
2234 - Many documentation updates
2235 - Distribute hypervisor files on ``gnt-cluster redist-conf``
2236 - ``gnt-instance reinstall`` can now reinstall multiple instances
2237 - Updated many command line parameters
2238 - Introduced new OS API version 15
2239 - No longer support a default hypervisor
2240 - Treat virtual LVs as inexistent
2241 - Improved job locking logic to reduce lock contention
2242 - Match instance and node names case insensitively
2243 - Reimplemented bash completion script to be more complete
2250 *(Released Thu, 4 Feb 2010)*
2252 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
2253 - Fix a string formatting bug
2254 - Improve safety of the code in some error paths
2255 - Improve data validation in the master of values returned from nodes
2261 *(Released Thu, 17 Dec 2009)*
2263 - Fix security issue due to missing validation of iallocator names; this
2264 allows local and remote execution of arbitrary executables
2265 - Fix failure of gnt-node list during instance removal
2266 - Ship the RAPI documentation in the archive
2272 *(Released Wed, 30 Sep 2009)*
2274 - Fixed many wrong messages
2275 - Fixed a few bugs related to the locking library
2276 - Fixed MAC checking at instance creation time
2277 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
2278 - Fixed a few issues related to signal handling in both daemons and
2280 - Fixed the example startup script provided
2281 - Fixed insserv dependencies in the example startup script (patch from
2283 - Fixed handling of drained nodes in the iallocator framework
2284 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
2286 - Fixed error related to invalid job IDs in job polling
2287 - Fixed job/opcode persistence on unclean master shutdown
2288 - Fixed handling of partial job processing after unclean master
2290 - Fixed error reporting from LUs, previously all errors were converted
2291 into execution errors
2292 - Fixed error reporting from burnin
2293 - Decreased significantly the memory usage of the job queue
2294 - Optimised slightly multi-job submission
2295 - Optimised slightly opcode loading
2296 - Backported the multi-job submit framework from the development
2297 branch; multi-instance start and stop should be faster
2298 - Added script to clean archived jobs after 21 days; this will reduce
2299 the size of the queue directory
2300 - Added some extra checks in disk size tracking
2301 - Added an example ethers hook script
2302 - Added a cluster parameter that prevents Ganeti from modifying of
2304 - Added more node information to RAPI responses
2305 - Added a ``gnt-job watch`` command that allows following the ouput of a
2307 - Added a bind-address option to ganeti-rapi
2308 - Added more checks to the configuration verify
2309 - Enhanced the burnin script such that some operations can be retried
2311 - Converted instance reinstall to multi-instance model
2317 *(Released Fri, 7 Aug 2009)*
2319 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
2320 to allow using the pre-2.0.2 behaviour in activation, if any existing
2321 instances have mismatched disk sizes in the configuration
2322 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
2323 any configuration mismatches for disk sizes
2324 - Added ``gnt-master cluste-failover --no-voting`` to allow master
2325 failover to work on two-node clusters
2326 - Fixed the ``--net`` option of ``gnt-backup import``, which was
2328 - Fixed detection of OS script errors in ``gnt-backup export``
2329 - Fixed exit code of ``gnt-backup export``
2335 *(Released Fri, 17 Jul 2009)*
2337 - Added experimental support for stripped logical volumes; this should
2338 enhance performance but comes with a higher complexity in the block
2339 device handling; stripping is only enabled when passing
2340 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
2341 affected even in the non-stripped mode
2342 - Improved resiliency against transient failures at the end of DRBD
2343 resyncs, and in general of DRBD resync checks
2344 - Fixed a couple of issues with exports and snapshot errors
2345 - Fixed a couple of issues in instance listing
2346 - Added display of the disk size in ``gnt-instance info``
2347 - Fixed checking for valid OSes in instance creation
2348 - Fixed handling of the "vcpus" parameter in instance listing and in
2349 general of invalid parameters
2350 - Fixed http server library, and thus RAPI, to handle invalid
2351 username/password combinations correctly; this means that now they
2352 report unauthorized for queries too, not only for modifications,
2353 allowing earlier detect of configuration problems
2354 - Added a new "role" node list field, equivalent to the master/master
2355 candidate/drained/offline flags combinations
2356 - Fixed cluster modify and changes of candidate pool size
2357 - Fixed cluster verify error messages for wrong files on regular nodes
2358 - Fixed a couple of issues with node demotion from master candidate role
2359 - Fixed node readd issues
2360 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
2361 - Added a new ``--no-voting`` option for masterfailover to fix failover
2362 on two-nodes clusters when the former master node is unreachable
2363 - Added instance reinstall over RAPI
2369 *(Released Tue, 16 Jun 2009)*
2371 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
2372 allow re-adding the start in single-user option (regression from 1.2)
2373 - the watcher writes the instance status to a file, to allow monitoring
2374 to report the instance status (from the master) based on cached
2375 results of the watcher's queries; while this can get stale if the
2376 watcher is being locked due to other work on the cluster, this is
2377 still an improvement
2378 - the watcher now also restarts the node daemon and the rapi daemon if
2380 - fixed the watcher to handle full and drained queue cases
2381 - hooks export more instance data in the environment, which helps if
2382 hook scripts need to take action based on the instance's properties
2383 (no longer need to query back into ganeti)
2384 - instance failovers when the instance is stopped do not check for free
2385 RAM, so that failing over a stopped instance is possible in low memory
2387 - rapi uses queries for tags instead of jobs (for less job traffic), and
2388 for cluster tags it won't talk to masterd at all but read them from
2390 - a couple of error handling fixes in RAPI
2391 - drbd handling: improved the error handling of inconsistent disks after
2392 resync to reduce the frequency of "there are some degraded disks for
2393 this instance" messages
2394 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
2395 error handling path called a wrong function name)
2401 *(Released Wed, 27 May 2009)*
2403 - no changes from rc5
2409 *(Released Wed, 20 May 2009)*
2411 - fix a couple of bugs (validation, argument checks)
2412 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
2413 - some small improvements to RAPI and IAllocator
2414 - make watcher automatically start the master daemon if down
2420 *(Released Mon, 27 Apr 2009)*
2422 - change the OS list to not require locks; this helps with big clusters
2423 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
2424 volume group is broken
2425 - ``gnt-instance info``, without any arguments, doesn't run for all
2426 instances anymore; either pass ``--all`` or pass the desired
2427 instances; this helps against mistakes on big clusters where listing
2428 the information for all instances takes a long time
2429 - miscellaneous doc and man pages fixes
2435 *(Released Wed, 8 Apr 2009)*
2437 - Change the internal locking model of some ``gnt-node`` commands, in
2438 order to reduce contention (and blocking of master daemon) when
2439 batching many creation/reinstall jobs
2440 - Fixes to Xen soft reboot
2441 - No longer build documentation at build time, instead distribute it in
2442 the archive, in order to reduce the need for the whole docbook/rst
2449 *(Released Fri, 27 Mar 2009)*
2451 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
2452 - Fix watcher startup sequence, improves the behaviour of busy clusters
2453 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
2454 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
2455 bind address and other places
2456 - Some documentation fixes and updates
2462 *(Released Mon, 2 Mar 2009)*
2464 - More documentation updates, now all docs should be more-or-less
2466 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
2468 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
2470 - Fix an issue related to $libdir/run/ganeti and cluster creation
2476 *(Released Thu, 19 Feb 2009)*
2478 - Xen PVM and KVM have switched the default value for the instance root
2479 disk to the first partition on the first drive, instead of the whole
2480 drive; this means that the OS installation scripts must be changed
2482 - Man pages have been updated
2483 - RAPI has been switched by default to HTTPS, and the exported functions
2484 should all work correctly
2485 - RAPI v1 has been removed
2486 - Many improvements to the KVM hypervisor
2487 - Block device errors are now better reported
2488 - Many other bugfixes and small improvements
2494 *(Released Mon, 26 Jan 2009)*
2496 - Version 2 is a general rewrite of the code and therefore the
2497 differences are too many to list, see the design document for 2.0 in
2498 the ``doc/`` subdirectory for more details
2499 - In this beta version there is not yet a migration path from 1.2 (there
2500 will be one in the final 2.0 release)
2501 - A few significant changes are:
2503 - all commands are executed by a daemon (``ganeti-masterd``) and the
2504 various ``gnt-*`` commands are just front-ends to it
2505 - all the commands are entered into, and executed from a job queue,
2506 see the ``gnt-job(8)`` manpage
2507 - the RAPI daemon supports read-write operations, secured by basic
2508 HTTP authentication on top of HTTPS
2509 - DRBD version 0.7 support has been removed, DRBD 8 is the only
2510 supported version (when migrating from Ganeti 1.2 to 2.0, you need
2511 to migrate to DRBD 8 first while still running Ganeti 1.2)
2512 - DRBD devices are using statically allocated minor numbers, which
2513 will be assigned to existing instances during the migration process
2514 - there is support for both Xen PVM and Xen HVM instances running on
2516 - KVM virtualization is supported too
2517 - file-based storage has been implemented, which means that it is
2518 possible to run the cluster without LVM and DRBD storage, for
2519 example using a shared filesystem exported from shared storage (and
2520 still have live migration)
2526 *(Released Tue, 13 Jan 2009)*
2528 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
2529 - Reuse the old instance mac address by default on instance import, if
2530 the instance name is the same.
2531 - Handle situations in which the node info rpc returns incomplete
2533 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
2534 - Improved version of batcher:
2536 - state file support
2537 - instance mac address support
2538 - support for HVM clusters/instances
2540 - Add an option to show the number of cpu sockets and nodes in
2542 - Support OSes that handle more than one version of the OS api (but do
2543 not change the current API in any other way)
2544 - Fix ``gnt-node migrate``
2545 - ``gnt-debug`` man page
2546 - Fixes various more typos and small issues
2547 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
2553 *(Released Wed, 24 Sep 2008)*
2555 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
2556 type of disk exported to fully virtualized instances.
2557 - provide access to the serial console of HVM instances
2558 - instance auto_balance flag, set by default. If turned off it will
2559 avoid warnings on cluster verify if there is not enough memory to fail
2560 over an instance. in the future it will prevent automatically failing
2561 it over when we will support that.
2562 - batcher tool for instance creation, see ``tools/README.batcher``
2563 - ``gnt-instance reinstall --select-os`` to interactively select a new
2564 operating system when reinstalling an instance.
2565 - when changing the memory amount on instance modify a check has been
2566 added that the instance will be able to start. also warnings are
2567 emitted if the instance will not be able to fail over, if auto_balance
2569 - documentation fixes
2570 - sync fields between ``gnt-instance list/modify/add/import``
2571 - fix a race condition in drbd when the sync speed was set after giving
2572 the device a remote peer.
2578 *(Released Tue, 22 Jul 2008)*
2580 - note: the allowed size and number of tags per object were reduced
2581 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
2582 - fixed twisted 8.x compatibility
2583 - fixed ``gnt-instance replace-disks`` with iallocator
2584 - add TCP keepalives on twisted connections to detect restarted nodes
2585 - disk increase support, see ``gnt-instance grow-disk``
2586 - implement bulk node/instance query for RAPI
2587 - add tags in node/instance listing (optional)
2588 - experimental migration (and live migration) support, read the man page
2589 for ``gnt-instance migrate``
2590 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
2591 has some small improvements in handling its state file
2597 *(Released Fri, 13 Jun 2008)*
2599 - Experimental readonly, REST-based remote API implementation;
2600 automatically started on master node, TCP port 5080, if enabled by
2601 ``--enable-rapi`` parameter to configure script.
2602 - Instance allocator support. Add and import instance accept a
2603 ``--iallocator`` parameter, and call that instance allocator to decide
2604 which node to use for the instance. The iallocator document describes
2605 what's expected from an allocator script.
2606 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
2607 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
2608 node is lost there is still enough memory to fail over the instances
2610 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
2611 ``gnt-cluster verify``, to check for site-specific compliance. All the
2612 hooks will run, and their output, if any, will be displayed. Any
2613 failing hook will make the verification return an error value.
2614 - ``gnt-cluster verify`` now checks that its peers are reachable on the
2615 primary and secondary interfaces
2616 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
2617 that is still declared as part of the cluster and has failed.
2618 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
2619 specifying output fields, that just adds the chosen fields to the
2621 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
2622 export from the filesystem.
2623 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
2624 have been added. Using them you can enable/disable acpi and pae
2625 support, and specify a path for a cd image to be exported to the
2626 instance. These parameters as the name suggest only work on HVM
2628 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
2629 PAE support will be set to the previously hardcoded values, but the
2630 (previously hardcoded) path to the CDROM ISO image will be unset and
2631 if required, needs to be set manually with ``gnt-instance modify``
2633 - The address to which an instance's VNC console is bound is now
2634 selectable per-instance, rather than being cluster wide. Of course
2635 this only applies to instances controlled via VNC, so currently just
2636 applies to HVM clusters.
2642 *(Released Mon, 18 Feb 2008)*
2644 - more tweaks to the disk activation code (especially helpful for DRBD)
2645 - change the default ``gnt-instance list`` output format, now there is
2646 one combined status field (see the manpage for the exact values this
2648 - some more fixes for the mac export to hooks change
2649 - make Ganeti not break with DRBD 8.2.x (which changed the version
2650 format in ``/proc/drbd``) (issue 24)
2651 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
2652 template, allowing migration from DRBD0.7+MD to DRBD8
2658 *(Released Wed, 30 Jan 2008)*
2660 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
2662 - add command aliases infrastructure and a few aliases
2663 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
2664 man pages and the ``--help`` option of ``gnt-node
2665 list``/``gnt-instance list``
2666 - fix ``gnt-backup list`` with down nodes (issue 21)
2667 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
2668 - fix the dist archive and add a check for including svn/git files in
2670 - some developer-related changes: improve the burnin and the QA suite,
2671 add an upload script for testing during development
2677 *(Released Wed, 16 Jan 2008)*
2679 - experimental HVM support, read the install document, section
2680 "Initializing the cluster"
2681 - allow for the PVM hypervisor per-instance kernel and initrd paths
2682 - add a new command ``gnt-cluster verify-disks`` which uses a new
2683 algorithm to improve the reconnection of the DRBD pairs if the device
2684 on the secondary node has gone away
2685 - make logical volume code auto-activate LVs at disk activation time
2686 - slightly improve the speed of activating disks
2687 - allow specification of the MAC address at instance creation time, and
2688 changing it later via ``gnt-instance modify``
2689 - fix handling of external commands that generate lots of output on
2691 - update documentation with regard to minimum version of DRBD8 supported
2697 *(Released Tue, 4 Dec 2007)*
2699 - Log the ``xm create`` output to the node daemon log on failure (to
2700 help diagnosing the error)
2701 - In debug mode, log all external commands output if failed to the logs
2702 - Change parsing of lvm commands to ignore stderr
2708 *(Released Wed, 28 Nov 2007)*
2710 - Another round of updates to the DRBD 8 code to deal with more failures
2711 in the replace secondary node operation
2712 - Some more logging of failures in disk operations (lvm, drbd)
2713 - A few documentation updates
2720 *(Released Tue, 13 Nov 2007)*
2722 - Change configuration file format from Python's Pickle to JSON.
2723 Upgrading is possible using the cfgupgrade utility.
2724 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
2725 faster replace disks and is more stable (DRBD 8 has many improvements
2726 compared to DRBD 0.7)
2727 - Added command line tags support (see man pages for ``gnt-instance``,
2728 ``gnt-node``, ``gnt-cluster``)
2729 - Added instance rename support
2730 - Added multi-instance startup/shutdown
2731 - Added cluster rename support
2732 - Added ``gnt-node evacuate`` to simplify some node operations
2733 - Added instance reboot operation that can speedup reboot as compared to
2735 - Soften the requirement that hostnames are in FQDN format
2736 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
2738 - Removed dependency on debian's patched fping that uses the
2739 non-standard ``-S`` option
2740 - Now the OS definitions are searched for in multiple, configurable
2741 paths (easier for distros to package)
2742 - Some changes to the hooks infrastructure (especially the new
2743 post-configuration update hook)
2744 - Other small bugfixes
2746 .. vim: set textwidth=72 syntax=rst :