10 - DRBD 8.4 support. Depending on the installed DRBD version, Ganeti now uses
11 the correct command syntax. It is possible to use different DRBD versions
12 on different nodes as long as they are compatible to each other. This
13 enables rolling upgrades of DRBD with no downtime. As permanent operation
14 of different DRBD versions within a node group is discouraged,
15 ``gnt-cluster verify`` will emit a warning if it detects such a situation.
16 - hroller now also plans for capacity to move non-redundant instances off
17 any node to be rebooted; the old behavior of completely ignoring any
18 non-redundant instances can be restored by adding the --ignore-non-redundant
20 - The cluster option '--no-lvm-storage' was removed in favor of the new option
21 '--enabled-disk-templates'.
22 - On instance creation, disk templates no longer need to be specified
23 with '-t'. The default disk template will be taken from the list of
24 enabled disk templates.
25 - New "inst-status-xen" data collector for the monitoring daemon, providing
26 information about the state of the xen instances on the nodes.
27 - New "lv" data collector for the monitoring daemon, collecting data about the
28 logical volumes on the nodes, and pairing them with the name of the instances
30 - New "diskstats" data collector, collecting the data from /proc/diskstats and
31 presenting them over the monitoring daemon interface.
32 - The monitoring daemon is now running as root, in order to be able to collect
33 information only available to root (such as the state of Xen instances).
34 - The ConfD client is now IPv6 compatible.
40 *(Released Mon, 24 Jun 2013)*
42 Incompatible/important changes
43 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
45 - Instance policy can contain multiple instance specs, as described in
46 the “Constrained instance sizes” section of :doc:`Partitioned Ganeti
47 <design-partitioned>`. As a consequence, it's not possible to partially change
48 or override instance specs. Bounding specs (min and max) can be specified as a
49 whole using the new option ``--ipolicy-bounds-specs``, while standard
50 specs use the new option ``--ipolicy-std-specs``.
51 - The output of the info command of gnt-cluster, gnt-group, gnt-node,
52 gnt-instance is a valid YAML object.
53 - hail now honors network restrictions when allocating nodes. This led to an
54 update of the IAllocator protocol. See the IAllocator documentation for
60 - The :doc:`Remote API <rapi>` daemon now supports a command line flag
61 to always require authentication, ``--require-authentication``. It can
62 be specified in ``$sysconfdir/default/ganeti``.
63 - A new cluster attribute 'enabled_disk_templates' is introduced. It will
64 be used to manage the disk templates to be used by instances in the cluster.
65 Initially, it will be set to a list that includes plain, drbd, if they were
66 enabled by specifying a volume group name, and file and sharedfile, if those
67 were enabled at configure time. Additionally, it will include all disk
68 templates that are currently used by instances. The order of disk templates
69 will be based on Ganeti's history of supporting them. In the future, the
70 first entry of the list will be used as a default disk template on instance
72 - ``cfgupgrade`` now supports a ``--downgrade`` option to bring the
73 configuration back to the previous stable version.
74 - Disk templates in group ipolicy can be restored to the default value.
75 - Initial support for diskless instances and virtual clusters in QA.
76 - More QA and unit tests for instance policies.
77 - Every opcode now contains a reason trail (visible through ``gnt-job info``)
78 describing why the opcode itself was executed.
79 - The monitoring daemon is now available. It allows users to query the cluster
80 for obtaining information about the status of the system. The daemon is only
81 responsible for providing the information over the network: the actual data
82 gathering is performed by data collectors (currently, only the DRBD status
83 collector is available).
84 - In order to help developers work on Ganeti, a new script
85 (``devel/build_chroot``) is provided, for building a chroot that contains all
86 the required development libraries and tools for compiling Ganeti on a Debian
88 - A new tool, ``harep``, for performing self-repair and recreation of instances
89 in Ganeti has been added.
90 - Split queries are enabled for tags, network, exports, cluster info, groups,
92 - New command ``show-ispecs-cmd`` for ``gnt-cluster`` and ``gnt-group``.
93 It prints the command line to set the current policies, to ease
98 The following new dependencies have been added:
101 - The ``curl`` library is not optional anymore for compiling the Haskell code.
102 - ``snap-server`` library (if monitoring is enabled).
105 - The minimum Python version needed to run Ganeti is now 2.6.
106 - ``yaml`` library (only for running the QA).
112 *(Released Thu, 04 Jul 2013)*
114 Incompatible/important changes
115 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
117 - Instance policies for disk size were documented to be on a per-disk
118 basis, but hail applied them to the sum of all disks. This has been
120 - ``hbal`` will now exit with status 0 if, during job execution over
121 LUXI, early exit has been requested and all jobs are successful;
122 before, exit status 1 was used, which cannot be differentiated from
124 - Compatibility with newer versions of rbd has been fixed
125 - ``gnt-instance batch-create`` has been changed to use the bulk create
126 opcode from Ganeti. This lead to incompatible changes in the format of
127 the JSON file. It's now not a custom dict anymore but a dict
128 compatible with the ``OpInstanceCreate`` opcode.
129 - Parent directories for file storage need to be listed in
130 ``$sysconfdir/ganeti/file-storage-paths`` now. ``cfgupgrade`` will
131 write the file automatically based on old configuration values, but it
132 can not distribute it across all nodes and the file contents should be
133 verified. Use ``gnt-cluster copyfile
134 $sysconfdir/ganeti/file-storage-paths`` once the cluster has been
135 upgraded. The reason for requiring this list of paths now is that
136 before it would have been possible to inject new paths via RPC,
137 allowing files to be created in arbitrary locations. The RPC protocol
138 is protected using SSL/X.509 certificates, but as a design principle
139 Ganeti does not permit arbitrary paths to be passed.
140 - The parsing of the variants file for OSes (see
141 :manpage:`ganeti-os-interface(7)`) has been slightly changed: now empty
142 lines and comment lines (starting with ``#``) are ignored for better
144 - The ``setup-ssh`` tool added in Ganeti 2.2 has been replaced and is no
145 longer available. ``gnt-node add`` now invokes a new tool on the
146 destination node, named ``prepare-node-join``, to configure the SSH
147 daemon. Paramiko is no longer necessary to configure nodes' SSH
148 daemons via ``gnt-node add``.
149 - Draining (``gnt-cluster queue drain``) and un-draining the job queue
150 (``gnt-cluster queue undrain``) now affects all nodes in a cluster and
151 the flag is not reset after a master failover.
152 - Python 2.4 has *not* been tested with this release. Using 2.6 or above
153 is recommended. 2.6 will be mandatory from the 2.8 series.
159 - New network management functionality to support automatic allocation
160 of IP addresses and managing of network parameters. See
161 :manpage:`gnt-network(8)` for more details.
162 - New external storage backend, to allow managing arbitrary storage
163 systems external to the cluster. See
164 :manpage:`ganeti-extstorage-interface(7)`.
165 - New ``exclusive-storage`` node parameter added, restricted to
166 nodegroup level. When it's set to true, physical disks are assigned in
167 an exclusive fashion to instances, as documented in :doc:`Partitioned
168 Ganeti <design-partitioned>`. Currently, only instances using the
169 ``plain`` disk template are supported.
170 - The KVM hypervisor has been updated with many new hypervisor
171 parameters, including a generic one for passing arbitrary command line
172 values. See a complete list in :manpage:`gnt-instance(8)`. It is now
173 compatible up to qemu 1.4.
174 - A new tool, called ``mon-collector``, is the stand-alone executor of
175 the data collectors for a monitoring system. As of this version, it
176 just includes the DRBD data collector, that can be executed by calling
177 ``mon-collector`` using the ``drbd`` parameter. See
178 :manpage:`mon-collector(7)`.
179 - A new user option, :pyeval:`rapi.RAPI_ACCESS_READ`, has been added
180 for RAPI users. It allows granting permissions to query for
181 information to a specific user without giving
182 :pyeval:`rapi.RAPI_ACCESS_WRITE` permissions.
183 - A new tool named ``node-cleanup`` has been added. It cleans remains of
184 a cluster from a machine by stopping all daemons, removing
185 certificates and ssconf files. Unless the ``--no-backup`` option is
186 given, copies of the certificates are made.
187 - Instance creations now support the use of opportunistic locking,
188 potentially speeding up the (parallel) creation of multiple instances.
189 This feature is currently only available via the :doc:`RAPI
190 <rapi>` interface and when an instance allocator is used. If the
191 ``opportunistic_locking`` parameter is set the opcode will try to
192 acquire as many locks as possible, but will not wait for any locks
193 held by other opcodes. If not enough resources can be found to
194 allocate the instance, the temporary error code
195 :pyeval:`errors.ECODE_TEMP_NORES` is returned. The operation can be
196 retried thereafter, with or without opportunistic locking.
197 - New experimental linux-ha resource scripts.
198 - Restricted-commands support: ganeti can now be asked (via command line
199 or rapi) to perform commands on a node. These are passed via ganeti
200 RPC rather than ssh. This functionality is restricted to commands
201 specified on the ``$sysconfdir/ganeti/restricted-commands`` for security
202 reasons. The file is not copied automatically.
208 - Diskless instances are now externally mirrored (Issue 237). This for
209 now has only been tested in conjunction with explicit target nodes for
211 - Queries not needing locks or RPC access to the node can now be
212 performed by the confd daemon, making them independent from jobs, and
213 thus faster to execute. This is selectable at configure time.
214 - The functionality for allocating multiple instances at once has been
215 overhauled and is now also available through :doc:`RAPI <rapi>`.
217 There are no significant changes from version 2.7.0~rc3.
223 *(Released Tue, 25 Jun 2013)*
225 - Fix permissions on the confd query socket (Issue 477)
226 - Fix permissions on the job archive dir (Issue 498)
227 - Fix handling of an internal exception in replace-disks (Issue 472)
228 - Fix gnt-node info handling of shortened names (Issue 497)
229 - Fix gnt-instance grow-disk when wiping is enabled
230 - Documentation improvements, and support for newer pandoc
231 - Fix hspace honoring ipolicy for disks (Issue 484)
232 - Improve handling of the ``kvm_extra`` HV parameter
238 *(Released Fri, 24 May 2013)*
240 - ``devel/upload`` now works when ``/var/run`` on the target nodes is a
242 - Disks added through ``gnt-instance modify`` or created through
243 ``gnt-instance recreate-disks`` are wiped, if the
244 ``prealloc_wipe_disks`` flag is set.
245 - If wiping newly created disks fails, the disks are removed. Also,
246 partial failures in creating disks through ``gnt-instance modify``
247 triggers a cleanup of the partially-created disks.
248 - Removing the master IP address doesn't fail if the address has been
250 - Fix ownership of the OS log dir
251 - Workaround missing SO_PEERCRED constant (Issue 191)
257 *(Released Fri, 3 May 2013)*
259 This was the first release candidate of the 2.7 series. Since beta3:
261 - Fix kvm compatibility with qemu 1.4 (Issue 389)
262 - Documentation updates (admin guide, upgrade notes, install
263 instructions) (Issue 372)
264 - Fix gnt-group list nodes and instances count (Issue 436)
265 - Fix compilation without non-mandatory libraries (Issue 441)
266 - Fix xen-hvm hypervisor forcing nics to type 'ioemu' (Issue 247)
267 - Make confd logging more verbose at INFO level (Issue 435)
268 - Improve "networks" documentation in :manpage:`gnt-instance(8)`
269 - Fix failure path for instance storage type conversion (Issue 229)
270 - Update htools text backend documentation
271 - Improve the renew-crypto section of :manpage:`gnt-cluster(8)`
272 - Disable inter-cluster instance move for file-based instances, because
273 it is dependant on instance export, which is not supported for
274 file-based instances. (Issue 414)
275 - Fix gnt-job crashes on non-ascii characters (Issue 427)
276 - Fix volume group checks on non-vm-capable nodes (Issue 432)
282 *(Released Mon, 22 Apr 2013)*
284 This was the third beta release of the 2.7 series. Since beta2:
286 - Fix hail to verify disk instance policies on a per-disk basis (Issue 418).
287 - Fix data loss on wrong usage of ``gnt-instance move``
288 - Properly export errors in confd-based job queries
289 - Add ``users-setup`` tool
290 - Fix iallocator protocol to report 0 as a disk size for diskless
291 instances. This avoids hail breaking when a diskless instance is
293 - Fix job queue directory permission problem that made confd job queries
294 fail. This requires running an ``ensure-dirs --full-run`` on upgrade
295 for access to archived jobs (Issue 406).
296 - Limit the sizes of networks supported by ``gnt-network`` to something
297 between a ``/16`` and a ``/30`` to prevent memory bloat and crashes.
298 - Fix bugs in instance disk template conversion
299 - Fix GHC 7 compatibility
300 - Fix ``burnin`` install path (Issue 426).
301 - Allow very small disk grows (Issue 347).
302 - Fix a ``ganeti-noded`` memory bloat introduced in 2.5, by making sure
303 that noded doesn't import masterd code (Issue 419).
304 - Make sure the default metavg at cluster init is the same as the vg, if
305 unspecified (Issue 358).
306 - Fix cleanup of partially created disks (part of Issue 416)
312 *(Released Tue, 2 Apr 2013)*
314 This was the second beta release of the 2.7 series. Since beta1:
316 - Networks no longer have a "type" slot, since this information was
317 unused in Ganeti: instead of it tags should be used.
318 - The rapi client now has a ``target_node`` option to MigrateInstance.
319 - Fix early exit return code for hbal (Issue 386).
320 - Fix ``gnt-instance migrate/failover -n`` (Issue 396).
321 - Fix ``rbd showmapped`` output parsing (Issue 312).
322 - Networks are now referenced indexed by UUID, rather than name. This
323 will require running cfgupgrade, from 2.7.0beta1, if networks are in
325 - The OS environment now includes network information.
326 - Deleting of a network is now disallowed if any instance nic is using
327 it, to prevent dangling references.
328 - External storage is now documented in man pages.
329 - The exclusive_storage flag can now only be set at nodegroup level.
330 - Hbal can now submit an explicit priority with its jobs.
331 - Many network related locking fixes.
332 - Bump up the required pylint version to 0.25.1.
333 - Fix the ``no_remember`` option in RAPI client.
334 - Many ipolicy related tests, qa, and fixes.
335 - Many documentation improvements and fixes.
336 - Fix building with ``--disable-file-storage``.
337 - Fix ``-q`` option in htools, which was broken if passed more than
339 - Some haskell/python interaction improvements and fixes.
340 - Fix iallocator in case of missing LVM storage.
341 - Fix confd config load in case of ``--no-lvm-storage``.
342 - The confd/query functionality is now mentioned in the security
349 *(Released Wed, 6 Feb 2013)*
351 This was the first beta release of the 2.7 series. All important changes
352 are listed in the latest 2.7 entry.
358 *(Released Fri, 21 Dec 2012)*
360 Important behaviour change: hbal won't rebalance anymore instances which
361 have the ``auto_balance`` attribute set to false. This was the intention
362 all along, but until now it only skipped those from the N+1 memory
363 reservation (DRBD-specific).
365 A significant number of bug fixes in this release:
367 - Fixed disk adoption interaction with ipolicy checks.
368 - Fixed networking issues when instances are started, stopped or
369 migrated, by forcing the tap device's MAC prefix to "fe" (issue 217).
370 - Fixed the warning in cluster verify for shared storage instances not
372 - Fixed removal of storage directory on shared file storage (issue 262).
373 - Fixed validation of LVM volume group name in OpClusterSetParams
374 (``gnt-cluster modify``) (issue 285).
375 - Fixed runtime memory increases (``gnt-instance modify -m``).
376 - Fixed live migration under Xen's ``xl`` mode.
377 - Fixed ``gnt-instance console`` with ``xl``.
378 - Fixed building with newer Haskell compiler/libraries.
379 - Fixed PID file writing in Haskell daemons (confd); this prevents
380 restart issues if confd was launched manually (outside of
381 ``daemon-util``) while another copy of it was running
382 - Fixed a type error when doing live migrations with KVM (issue 297) and
383 the error messages for failing migrations have been improved.
384 - Fixed opcode validation for the out-of-band commands (``gnt-node
386 - Fixed a type error when unsetting OS hypervisor parameters (issue
387 311); now it's possible to unset all OS-specific hypervisor
389 - Fixed the ``dry-run`` mode for many operations: verification of
390 results was over-zealous but didn't take into account the ``dry-run``
391 operation, resulting in "wrong" failures.
392 - Fixed bash completion in ``gnt-job list`` when the job queue has
393 hundreds of entries; especially with older ``bash`` versions, this
394 results in significant CPU usage.
396 And lastly, a few other improvements have been made:
398 - Added option to force master-failover without voting (issue 282).
399 - Clarified error message on lock conflict (issue 287).
400 - Logging of newly submitted jobs has been improved (issue 290).
401 - Hostname checks have been made uniform between instance rename and
403 - The ``--submit`` option is now supported by ``gnt-debug delay``.
404 - Shutting down the master daemon by sending SIGTERM now stops it from
405 processing jobs waiting for locks; instead, those jobs will be started
406 once again after the master daemon is started the next time (issue
408 - Support for Xen's ``xl`` program has been improved (besides the fixes
410 - Reduced logging noise in the Haskell confd daemon (only show one log
411 entry for each config reload, instead of two).
412 - Several man page updates and typo fixes.
418 *(Released Fri, 12 Oct 2012)*
420 A small bugfix release. Among the bugs fixed:
422 - Fixed double use of ``PRIORITY_OPT`` in ``gnt-node migrate``, that
423 made the command unusable.
424 - Commands that issue many jobs don't fail anymore just because some jobs
425 take so long that other jobs are archived.
426 - Failures during ``gnt-instance reinstall`` are reflected by the exit
428 - Issue 190 fixed. Check for DRBD in cluster verify is enabled only when
430 - When ``always_failover`` is set, ``--allow-failover`` is not required
431 in migrate commands anymore.
432 - ``bash_completion`` works even if extglob is disabled.
433 - Fixed bug with locks that made failover for RDB-based instances fail.
434 - Fixed bug in non-mirrored instance allocation that made Ganeti choose
435 a random node instead of one based on the allocator metric.
436 - Support for newer versions of pylint and pep8.
437 - Hail doesn't fail anymore when trying to add an instance of type
438 ``file``, ``sharedfile`` or ``rbd``.
439 - Added new Makefile target to rebuild the whole distribution, so that
440 all files are included.
446 *(Released Fri, 27 Jul 2012)*
449 .. attention:: The ``LUXI`` protocol has been made more consistent
450 regarding its handling of command arguments. This, however, leads to
451 incompatibility issues with previous versions. Please ensure that you
452 restart Ganeti daemons soon after the upgrade, otherwise most
453 ``LUXI`` calls (job submission, setting/resetting the drain flag,
454 pausing/resuming the watcher, cancelling and archiving jobs, querying
455 the cluster configuration) will fail.
464 The current ``admin_up`` field, which used to denote whether an instance
465 should be running or not, has been removed. Instead, ``admin_state`` is
466 introduced, with 3 possible values -- ``up``, ``down`` and ``offline``.
468 The rational behind this is that an instance being “down” can have
471 - it could be down during a reboot
472 - it could be temporarily be down for a reinstall
473 - or it could be down because it is deprecated and kept just for its
476 The previous Boolean state was making it difficult to do capacity
477 calculations: should Ganeti reserve memory for a down instance? Now, the
478 tri-state field makes it clear:
480 - in ``up`` and ``down`` state, all resources are reserved for the
481 instance, and it can be at any time brought up if it is down
482 - in ``offline`` state, only disk space is reserved for it, but not
485 The field can have an extra use: since the transition between ``up`` and
486 ``down`` and vice-versus is done via ``gnt-instance start/stop``, but
487 transition between ``offline`` and ``down`` is done via ``gnt-instance
488 modify``, it is possible to given different rights to users. For
489 example, owners of an instance could be allowed to start/stop it, but
490 not transition it out of the offline state.
492 Instance policies and specs
493 +++++++++++++++++++++++++++
495 In previous Ganeti versions, an instance creation request was not
496 limited on the minimum size and on the maximum size just by the cluster
497 resources. As such, any policy could be implemented only in third-party
498 clients (RAPI clients, or shell wrappers over ``gnt-*``
499 tools). Furthermore, calculating cluster capacity via ``hspace`` again
500 required external input with regards to instance sizes.
502 In order to improve these workflows and to allow for example better
503 per-node group differentiation, we introduced instance specs, which
506 - minimum instance disk size, disk count, memory size, cpu count
507 - maximum values for the above metrics
508 - and “standard” values (used in ``hspace`` to calculate the standard
511 The minimum/maximum values can be also customised at node-group level,
512 for example allowing more powerful hardware to support bigger instance
515 Beside the instance specs, there are a few other settings belonging to
516 the instance policy framework. It is possible now to customise, per
517 cluster and node-group:
519 - the list of allowed disk templates
520 - the maximum ratio of VCPUs per PCPUs (to control CPU oversubscription)
521 - the maximum ratio of instance to spindles (see below for more
522 information) for local storage
524 All these together should allow all tools that talk to Ganeti to know
525 what are the ranges of allowed values for instances and the
526 over-subscription that is allowed.
528 For the VCPU/PCPU ratio, we already have the VCPU configuration from the
529 instance configuration, and the physical CPU configuration from the
530 node. For the spindle ratios however, we didn't track before these
531 values, so new parameters have been added:
533 - a new node parameter ``spindle_count``, defaults to 1, customisable at
534 node group or node level
535 - at new backend parameter (for instances), ``spindle_use`` defaults to 1
537 Note that spindles in this context doesn't need to mean actual
538 mechanical hard-drives; it's just a relative number for both the node
539 I/O capacity and instance I/O consumption.
541 Instance migration behaviour
542 ++++++++++++++++++++++++++++
544 While live-migration is in general desirable over failover, it is
545 possible that for some workloads it is actually worse, due to the
546 variable time of the “suspend” phase during live migration.
548 To allow the tools to work consistently over such instances (without
549 having to hard-code instance names), a new backend parameter
550 ``always_failover`` has been added to control the migration/failover
551 behaviour. When set to True, all migration requests for an instance will
552 instead fall-back to failover.
554 Instance memory ballooning
555 ++++++++++++++++++++++++++
557 Initial support for memory ballooning has been added. The memory for an
558 instance is no longer fixed (backend parameter ``memory``), but instead
559 can vary between minimum and maximum values (backend parameters
560 ``minmem`` and ``maxmem``). Currently we only change an instance's
563 - live migrating or failing over and instance and the target node
564 doesn't have enough memory
565 - user requests changing the memory via ``gnt-instance modify
571 In order to control the use of specific CPUs by instance, support for
572 controlling CPU pinning has been added for the Xen, HVM and LXC
573 hypervisors. This is controlled by a new hypervisor parameter
574 ``cpu_mask``; details about possible values for this are in the
575 :manpage:`gnt-instance(8)`. Note that use of the most specific (precise
576 VCPU-to-CPU mapping) form will work well only when all nodes in your
577 cluster have the same amount of CPUs.
582 Another area in which Ganeti was not customisable were the parameters
583 used for storage configuration, e.g. how many stripes to use for LVM,
584 DRBD resync configuration, etc.
586 To improve this area, we've added disks parameters, which are
587 customisable at cluster and node group level, and which allow to
588 specify various parameters for disks (DRBD has the most parameters
589 currently), for example:
591 - DRBD resync algorithm and parameters (e.g. speed)
592 - the default VG for meta-data volumes for DRBD
593 - number of stripes for LVM (plain disk template)
596 These parameters can be modified via ``gnt-cluster modify -D …`` and
597 ``gnt-group modify -D …``, and are used at either instance creation (in
598 case of LVM stripes, for example) or at disk “activation” time
601 Rados block device support
602 ++++++++++++++++++++++++++
604 A Rados (http://ceph.com/wiki/Rbd) storage backend has been added,
605 denoted by the ``rbd`` disk template type. This is considered
606 experimental, feedback is welcome. For details on configuring it, see
607 the :doc:`install` document and the :manpage:`gnt-cluster(8)` man page.
612 The existing master IP functionality works well only in simple setups (a
613 single network shared by all nodes); however, if nodes belong to
614 different networks, then the ``/32`` setup and lack of routing
615 information is not enough.
617 To allow the master IP to function well in more complex cases, the
618 system was reworked as follows:
620 - a master IP netmask setting has been added
621 - the master IP activation/turn-down code was moved from the node daemon
623 - whether to run the Ganeti-supplied master IP script or a user-supplied
624 on is a ``gnt-cluster init`` setting
626 Details about the location of the standard and custom setup scripts are
627 in the man page :manpage:`gnt-cluster(8)`; for information about the
628 setup script protocol, look at the Ganeti-supplied script.
633 The `SPICE <http://www.linux-kvm.org/page/SPICE>`_ support has been
636 It is now possible to use TLS-protected connections, and when renewing
637 or changing the cluster certificates (via ``gnt-cluster renew-crypto``,
638 it is now possible to specify spice or spice CA certificates. Also, it
639 is possible to configure a password for SPICE sessions via the
640 hypervisor parameter ``spice_password_file``.
642 There are also new parameters to control the compression and streaming
643 options (e.g. ``spice_image_compression``, ``spice_streaming_video``,
644 etc.). For details, see the man page :manpage:`gnt-instance(8)` and look
645 for the spice parameters.
647 Lastly, it is now possible to see the SPICE connection information via
648 ``gnt-instance console``.
653 A new tool (``tools/ovfconverter``) has been added that supports
654 conversion between Ganeti and the `Open Virtualization Format
655 <http://en.wikipedia.org/wiki/Open_Virtualization_Format>`_ (both to and
658 This relies on the ``qemu-img`` tool to convert the disk formats, so the
659 actual compatibility with other virtualization solutions depends on it.
664 The configuration query daemon (``ganeti-confd``) is now optional, and
665 has been rewritten in Haskell; whether to use the daemon at all, use the
666 Python (default) or the Haskell version is selectable at configure time
667 via the ``--enable-confd`` parameter, which can take one of the
668 ``haskell``, ``python`` or ``no`` values. If not used, disabling the
669 daemon will result in a smaller footprint; for larger systems, we
670 welcome feedback on the Haskell version which might become the default
673 If you want to use ``gnt-node list-drbd`` you need to have the Haskell
674 daemon running. The Python version doesn't implement the new call.
677 User interface changes
678 ~~~~~~~~~~~~~~~~~~~~~~
680 We have replaced the ``--disks`` option of ``gnt-instance
681 replace-disks`` with a more flexible ``--disk`` option, which allows
682 adding and removing disks at arbitrary indices (Issue 188). Furthermore,
683 disk size and mode can be changed upon recreation (via ``gnt-instance
684 recreate-disks``, which accepts the same ``--disk`` option).
686 As many people are used to a ``show`` command, we have added that as an
687 alias to ``info`` on all ``gnt-*`` commands.
689 The ``gnt-instance grow-disk`` command has a new mode in which it can
690 accept the target size of the disk, instead of the delta; this can be
691 more safe since two runs in absolute mode will be idempotent, and
692 sometimes it's also easier to specify the desired size directly.
694 Also the handling of instances with regard to offline secondaries has
695 been improved. Instance operations should not fail because one of it's
696 secondary nodes is offline, even though it's safe to proceed.
698 A new command ``list-drbd`` has been added to the ``gnt-node`` script to
699 support debugging of DRBD issues on nodes. It provides a mapping of DRBD
700 minors to instance name.
705 RAPI coverage has improved, with (for example) new resources for
706 recreate-disks, node power-cycle, etc.
711 There is partial support for ``xl`` in the Xen hypervisor; feedback is
714 Python 2.7 is better supported, and after Ganeti 2.6 we will investigate
715 whether to still support Python 2.4 or move to Python 2.6 as minimum
718 Support for Fedora has been slightly improved; the provided example
719 init.d script should work better on it and the INSTALL file should
720 document the needed dependencies.
725 The deprecated ``QueryLocks`` LUXI request has been removed. Use
726 ``Query(what=QR_LOCK, ...)`` instead.
728 The LUXI requests :pyeval:`luxi.REQ_QUERY_JOBS`,
729 :pyeval:`luxi.REQ_QUERY_INSTANCES`, :pyeval:`luxi.REQ_QUERY_NODES`,
730 :pyeval:`luxi.REQ_QUERY_GROUPS`, :pyeval:`luxi.REQ_QUERY_EXPORTS` and
731 :pyeval:`luxi.REQ_QUERY_TAGS` are deprecated and will be removed in a
732 future version. :pyeval:`luxi.REQ_QUERY` should be used instead.
734 RAPI client: ``CertificateError`` now derives from
735 ``GanetiApiError``. This should make it more easy to handle Ganeti
738 Deprecation warnings due to PyCrypto/paramiko import in
739 ``tools/setup-ssh`` have been silenced, as usually they are safe; please
740 make sure to run an up-to-date paramiko version, if you use this tool.
742 The QA scripts now depend on Python 2.5 or above (the main code base
743 still works with Python 2.4).
745 The configuration file (``config.data``) is now written without
746 indentation for performance reasons; if you want to edit it, it can be
747 re-formatted via ``tools/fmtjson``.
749 A number of bugs has been fixed in the cluster merge tool.
751 ``x509`` certification verification (used in import-export) has been
752 changed to allow the same clock skew as permitted by the cluster
753 verification. This will remove some rare but hard to diagnose errors in
760 *(Released Thu, 19 Jul 2012)*
762 Very few changes from rc4 to the final release, only bugfixes:
764 - integrated fixes from release 2.5.2 (fix general boot flag for KVM
765 instance, fix CDROM booting for KVM instances)
766 - fixed node group modification of node parameters
767 - fixed issue in LUClusterVerifyGroup with multi-group clusters
768 - fixed generation of bash completion to ensure a stable ordering
775 *(Released Fri, 13 Jul 2012)*
777 Third release candidate for 2.6. The following changes were done from
780 - Fixed ``UpgradeConfig`` w.r.t. to disk parameters on disk objects.
781 - Fixed an inconsistency in the LUXI protocol with the provided
782 arguments (NOT backwards compatible)
783 - Fixed a bug with node groups ipolicy where ``min`` was greater than
784 the cluster ``std`` value
785 - Implemented a new ``gnt-node list-drbd`` call to list DRBD minors for
786 easier instance debugging on nodes (requires ``hconfd`` to work)
792 *(Released Tue, 03 Jul 2012)*
794 Second release candidate for 2.6. The following changes were done from
797 - Fixed ``gnt-cluster verify`` regarding ``master-ip-script`` on non
799 - Fixed a RAPI regression on missing beparams/memory
800 - Fixed redistribution of files on offline nodes
801 - Added possibility to run activate-disks even though secondaries are
802 offline. With this change it relaxes also the strictness on some other
803 commands which use activate disks internally:
804 * ``gnt-instance start|reboot|rename|backup|export``
805 - Made it possible to remove safely an instance if its secondaries are
807 - Made it possible to reinstall even though secondaries are offline
813 *(Released Mon, 25 Jun 2012)*
815 First release candidate for 2.6. The following changes were done from
818 - Fixed bugs with disk parameters and ``rbd`` templates as well as
820 - Made ``gnt-instance modify`` more consistent regarding new NIC/Disk
821 behaviour. It supports now the modify operation
822 - ``hcheck`` implemented to analyze cluster health and possibility of
823 improving health by rebalance
824 - ``hbal`` has been improved in dealing with split instances
830 *(Released Mon, 11 Jun 2012)*
832 Second beta release of 2.6. The following changes were done from beta2
835 - Fixed ``daemon-util`` with non-root user models
836 - Fixed creation of plain instances with ``--no-wait-for-sync``
837 - Fix wrong iv_names when running ``cfgupgrade``
838 - Export more information in RAPI group queries
839 - Fixed bug when changing instance network interfaces
840 - Extended burnin to do NIC changes
841 - query: Added ``<``, ``>``, ``<=``, ``>=`` comparison operators
842 - Changed default for DRBD barriers
843 - Fixed DRBD error reporting for syncer rate
844 - Verify the options on disk parameters
846 And of course various fixes to documentation and improved unittests and
853 *(Released Wed, 23 May 2012)*
855 First beta release of 2.6. The following changes were done from beta1 to
858 - integrated patch for distributions without ``start-stop-daemon``
859 - adapted example init.d script to work on Fedora
860 - fixed log handling in Haskell daemons
861 - adapted checks in the watcher for pycurl linked against libnss
862 - add partial support for ``xl`` instead of ``xm`` for Xen
863 - fixed a type issue in cluster verification
864 - fixed ssconf handling in the Haskell code (was breaking confd in IPv6
867 Plus integrated fixes from the 2.5 branch:
869 - fixed ``kvm-ifup`` to use ``/bin/bash``
870 - fixed parallel build failures
871 - KVM live migration when using a custom keymap
877 *(Released Tue, 24 Jul 2012)*
879 A small bugfix release, with no new features:
881 - fixed bash-isms in kvm-ifup, for compatibility with systems which use a
882 different default shell (e.g. Debian, Ubuntu)
883 - fixed KVM startup and live migration with a custom keymap (fixes Issue
884 243 and Debian bug #650664)
885 - fixed compatibility with KVM versions that don't support multiple boot
886 devices (fixes Issue 230 and Debian bug #624256)
888 Additionally, a few fixes were done to the build system (fixed parallel
889 build failures) and to the unittests (fixed race condition in test for
890 FileID functions, and the default enable/disable mode for QA test is now
897 *(Released Fri, 11 May 2012)*
899 A small bugfix release.
901 The main issues solved are on the topic of compatibility with newer LVM
904 - fixed parsing of ``lv_attr`` field
905 - adapted to new ``vgreduce --removemissing`` behaviour where sometimes
906 the ``--force`` flag is needed
908 Also on the topic of compatibility, ``tools/lvmstrap`` has been changed
909 to accept kernel 3.x too (was hardcoded to 2.6.*).
911 A regression present in 2.5.0 that broke handling (in the gnt-* scripts)
912 of hook results and that also made display of other errors suboptimal
913 was fixed; the code behaves now like 2.4 and earlier.
915 Another change in 2.5, the cleanup of the OS scripts environment, is too
916 aggressive: it removed even the ``PATH`` variable, which requires the OS
917 scripts to *always* need to export it. Since this is a bit too strict,
918 we now export a minimal PATH, the same that we export for hooks.
920 The fix for issue 201 (Preserve bridge MTU in KVM ifup script) was
921 integrated into this release.
923 Finally, a few other miscellaneous changes were done (no new features,
924 just small improvements):
926 - Fix ``gnt-group --help`` display
927 - Fix hardcoded Xen kernel path
928 - Fix grow-disk handling of invalid units
929 - Update synopsis for ``gnt-cluster repair-disk-sizes``
930 - Accept both PUT and POST in noded (makes future upgrade to 2.6 easier)
936 *(Released Thu, 12 Apr 2012)*
938 Incompatible/important changes and bugfixes
939 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
941 - The default of the ``/2/instances/[instance_name]/rename`` RAPI
942 resource's ``ip_check`` parameter changed from ``True`` to ``False``
943 to match the underlying LUXI interface.
944 - The ``/2/nodes/[node_name]/evacuate`` RAPI resource was changed to use
945 body parameters, see :doc:`RAPI documentation <rapi>`. The server does
946 not maintain backwards-compatibility as the underlying operation
947 changed in an incompatible way. The RAPI client can talk to old
948 servers, but it needs to be told so as the return value changed.
949 - When creating file-based instances via RAPI, the ``file_driver``
950 parameter no longer defaults to ``loop`` and must be specified.
951 - The deprecated ``bridge`` NIC parameter is no longer supported. Use
953 - Support for the undocumented and deprecated RAPI instance creation
954 request format version 0 has been dropped. Use version 1, supported
955 since Ganeti 2.1.3 and :doc:`documented <rapi>`, instead.
956 - Pyparsing 1.4.6 or above is required, see :doc:`installation
957 documentation <install>`.
958 - The "cluster-verify" hooks are now executed per group by the
959 ``OP_CLUSTER_VERIFY_GROUP`` opcode. This maintains the same behavior
960 if you just run ``gnt-cluster verify``, which generates one opcode per
962 - The environment as passed to the OS scripts is cleared, and thus no
963 environment variables defined in the node daemon's environment will be
964 inherited by the scripts.
965 - The :doc:`iallocator <iallocator>` mode ``multi-evacuate`` has been
967 - :doc:`New iallocator modes <design-multi-reloc>` have been added to
968 support operations involving multiple node groups.
969 - Offline nodes are ignored when failing over an instance.
970 - Support for KVM version 1.0, which changed the version reporting format
972 - TCP/IP ports used by DRBD disks are returned to a pool upon instance
974 - ``Makefile`` is now compatible with Automake 1.11.2
975 - Includes all bugfixes made in the 2.4 series
980 - The ganeti-htools project has been merged into the ganeti-core source
981 tree and will be built as part of Ganeti (see :doc:`install-quick`).
982 - Implemented support for :doc:`shared storage <design-shared-storage>`.
983 - Add support for disks larger than 2 TB in ``lvmstrap`` by supporting
984 GPT-style partition tables (requires `parted
985 <http://www.gnu.org/s/parted/>`_).
986 - Added support for floppy drive and 2nd CD-ROM drive in KVM hypervisor.
987 - Allowed adding tags on instance creation.
988 - Export instance tags to hooks (``INSTANCE_TAGS``, see :doc:`hooks`)
989 - Allow instances to be started in a paused state, enabling the user to
990 see the complete console output on boot using the console.
991 - Added new hypervisor flag to control default reboot behaviour
992 (``reboot_behavior``).
993 - Added support for KVM keymaps (hypervisor parameter ``keymap``).
994 - Improved out-of-band management support:
996 - Added ``gnt-node health`` command reporting the health status of
998 - Added ``gnt-node power`` command to manage power status of nodes.
999 - Added command for emergency power-off (EPO), ``gnt-cluster epo``.
1001 - Instance migration can fall back to failover if instance is not
1003 - Filters can be used when listing nodes, instances, groups and locks;
1004 see :manpage:`ganeti(7)` manpage.
1005 - Added post-execution status as variables to :doc:`hooks <hooks>`
1007 - Instance tags are exported/imported together with the instance.
1008 - When given an explicit job ID, ``gnt-job info`` will work for archived
1010 - Jobs can define dependencies on other jobs (not yet supported via
1011 RAPI or command line, but used by internal commands and usable via
1014 - Lock monitor (``gnt-debug locks``) shows jobs waiting for
1017 - Instance failover is now available as a RAPI resource
1018 (``/2/instances/[instance_name]/failover``).
1019 - ``gnt-instance info`` defaults to static information if primary node
1021 - Opcodes have a new ``comment`` attribute.
1022 - Added basic SPICE support to KVM hypervisor.
1023 - ``tools/ganeti-listrunner`` allows passing of arguments to executable.
1025 Node group improvements
1026 ~~~~~~~~~~~~~~~~~~~~~~~
1028 - ``gnt-cluster verify`` has been modified to check groups separately,
1029 thereby improving performance.
1030 - Node group support has been added to ``gnt-cluster verify-disks``,
1031 which now operates per node group.
1032 - Watcher has been changed to work better with node groups.
1034 - One process and state file per node group.
1035 - Slow watcher in one group doesn't block other group's watcher.
1037 - Added new command, ``gnt-group evacuate``, to move all instances in a
1038 node group to other groups.
1039 - Added ``gnt-instance change-group`` to move an instance to another
1041 - ``gnt-cluster command`` and ``gnt-cluster copyfile`` now support
1042 per-group operations.
1043 - Node groups can be tagged.
1044 - Some operations switch from an exclusive to a shared lock as soon as
1046 - Instance's primary and secondary nodes' groups are now available as
1047 query fields (``pnode.group``, ``pnode.group.uuid``, ``snodes.group``
1048 and ``snodes.group.uuid``).
1053 - Numerous updates to documentation and manpages.
1055 - :doc:`RAPI <rapi>` documentation now has detailed parameter
1057 - Some opcode/job results are now also documented, see :doc:`RAPI
1060 - A lockset's internal lock is now also visible in lock monitor.
1061 - Log messages from job queue workers now contain information about the
1062 opcode they're processing.
1063 - ``gnt-instance console`` no longer requires the instance lock.
1064 - A short delay when waiting for job changes reduces the number of LUXI
1065 requests significantly.
1066 - DRBD metadata volumes are overwritten with zeros during disk creation.
1067 - Out-of-band commands no longer acquire the cluster lock in exclusive
1069 - ``devel/upload`` now uses correct permissions for directories.
1075 *(Released Fri, 23 Mar 2012)*
1077 This was the sixth release candidate of the 2.5 series.
1083 *(Released Mon, 9 Jan 2012)*
1085 This was the fifth release candidate of the 2.5 series.
1091 *(Released Thu, 27 Oct 2011)*
1093 This was the fourth release candidate of the 2.5 series.
1099 *(Released Wed, 26 Oct 2011)*
1101 This was the third release candidate of the 2.5 series.
1107 *(Released Tue, 18 Oct 2011)*
1109 This was the second release candidate of the 2.5 series.
1115 *(Released Tue, 4 Oct 2011)*
1117 This was the first release candidate of the 2.5 series.
1123 *(Released Wed, 31 Aug 2011)*
1125 This was the third beta release of the 2.5 series.
1131 *(Released Mon, 22 Aug 2011)*
1133 This was the second beta release of the 2.5 series.
1139 *(Released Fri, 12 Aug 2011)*
1141 This was the first beta release of the 2.5 series.
1147 *(Released Thu, 27 Oct 2011)*
1149 - Fixed bug when parsing command line parameter values ending in
1151 - Fixed assertion error after unclean master shutdown
1152 - Disable HTTP client pool for RPC, significantly reducing memory usage
1154 - Fixed queue archive creation with wrong permissions
1160 *(Released Tue, 23 Aug 2011)*
1164 - Fixed documentation for importing with ``--src-dir`` option
1165 - Fixed a bug in ``ensure-dirs`` with queue/archive permissions
1166 - Fixed a parsing issue with DRBD 8.3.11 in the Linux kernel
1172 *(Released Fri, 5 Aug 2011)*
1174 Many bug-fixes and a few small features:
1176 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
1177 issues when you tried to add an instance with two MAC addresses in one
1179 - KVM: fixed per-instance stored UID value
1180 - KVM: configure bridged NICs at migration start
1181 - KVM: Fix a bug where instance will not start with never KVM versions
1183 - Added OS search path to ``gnt-cluster info``
1184 - Fixed an issue with ``file_storage_dir`` where you were forced to
1185 provide an absolute path, but the documentation states it is a
1186 relative path, the documentation was right
1187 - Added a new parameter to instance stop/start called ``--no-remember``
1188 that will make the state change to not be remembered
1189 - Implemented ``no_remember`` at RAPI level
1190 - Improved the documentation
1191 - Node evacuation: don't call IAllocator if node is already empty
1192 - Fixed bug in DRBD8 replace disks on current nodes
1193 - Fixed bug in recreate-disks for DRBD instances
1194 - Moved assertion checking locks in ``gnt-instance replace-disks``
1195 causing it to abort with not owning the right locks for some situation
1196 - Job queue: Fixed potential race condition when cancelling queued jobs
1197 - Fixed off-by-one bug in job serial generation
1198 - ``gnt-node volumes``: Fix instance names
1199 - Fixed aliases in bash completion
1200 - Fixed a bug in reopening log files after being sent a SIGHUP
1201 - Added a flag to burnin to allow specifying VCPU count
1202 - Bugfixes to non-root Ganeti configuration
1208 *(Released Thu, 12 May 2011)*
1210 Many bug-fixes and a few new small features:
1212 - Fixed a bug related to log opening failures
1213 - Fixed a bug in instance listing with orphan instances
1214 - Fixed a bug which prevented resetting the cluster-level node parameter
1215 ``oob_program`` to the default
1216 - Many fixes related to the ``cluster-merge`` tool
1217 - Fixed a race condition in the lock monitor, which caused failures
1218 during (at least) creation of many instances in parallel
1219 - Improved output for gnt-job info
1220 - Removed the quiet flag on some ssh calls which prevented debugging
1222 - Improved the N+1 failure messages in cluster verify by actually
1223 showing the memory values (needed and available)
1224 - Increased lock attempt timeouts so that when executing long operations
1225 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
1226 too early and thus prevent the use of the 'fair' mechanism
1227 - Changed instance query data (``gnt-instance info``) to not acquire
1228 locks unless needed, thus allowing its use on locked instance if only
1229 static information is asked for
1230 - Improved behaviour with filesystems that do not support rename on an
1232 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
1233 kept locks on all nodes during the wipe, which is unneeded
1234 - Fixed ``gnt-watcher`` handling of errors during hooks execution
1235 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
1236 10GiB) which caused the wipe to fail right at the end in some cases
1237 - Fixed master IP activation when doing master failover with no-voting
1238 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
1239 the master node itself
1240 - Fixed potential data-loss in under disk full conditions, where Ganeti
1241 wouldn't check correctly the return code and would consider
1242 partially-written files 'correct'
1243 - Fixed bug related to multiple VGs and DRBD disk replacing
1244 - Added new disk parameter ``metavg`` that allows placement of the meta
1245 device for DRBD in a different volume group
1246 - Fixed error handling in the node daemon when the system libc doesn't
1247 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
1248 - Fixed lock release during replace-disks, which kept cluster-wide locks
1249 when doing disk replaces with an iallocator script
1250 - Added check for missing bridges in cluster verify
1251 - Handle EPIPE errors while writing to the terminal better, so that
1252 piping the output to e.g. ``less`` doesn't cause a backtrace
1253 - Fixed rare case where a ^C during Luxi calls could have been
1254 interpreted as server errors, instead of simply terminating
1255 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
1257 - Added a few more parameters to the KVM hypervisor, allowing a second
1258 CDROM, custom disk type for CDROMs and a floppy image
1259 - Removed redundant message in instance rename when the name is given
1261 - Added option to ``gnt-instance recreate-disks`` to allow creating the
1262 disks on new nodes, allowing recreation when the original instance
1263 nodes are completely gone
1264 - Added option when converting disk templates to DRBD to skip waiting
1265 for the resync, in order to make the instance available sooner
1266 - Added two new variables to the OS scripts environment (containing the
1268 - Made the root_path and optional parameter for the xen-pvm hypervisor,
1269 to allow use of ``pvgrub`` as bootloader
1270 - Changed the instance memory modifications to only check out-of-memory
1271 conditions on memory increases, and turned the secondary node warnings
1272 into errors (they can still be overridden via ``--force``)
1273 - Fixed the handling of a corner case when the Python installation gets
1274 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
1275 to execute a command that doesn't exist
1276 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
1277 node of the instance returned failures during instance shutdown; this
1278 adds the option ``--ignore-consistency`` to gnt-instance move
1280 And as usual, various improvements to the error messages, documentation
1287 *(Released Wed, 09 Mar 2011)*
1289 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
1290 the RAPI users file if run twice (even with ``--dry-run``).
1292 The release fixes that bug (nothing else changed).
1298 *(Released Mon, 07 Mar 2011)*
1300 Final 2.4.0 release. Just a few small fixes:
1302 - Fixed RAPI node evacuate
1303 - Fixed the kvm-ifup script
1304 - Fixed internal error handling for special job cases
1305 - Updated man page to specify the escaping feature for options
1311 *(Released Mon, 28 Feb 2011)*
1313 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
1314 that this feature wiped the disks of the wrong instance, leading to loss
1319 - Fixed title of query field containing instance name
1320 - Expanded the glossary in the documentation
1321 - Fixed one unittest (internal issue)
1327 *(Released Mon, 21 Feb 2011)*
1329 A number of bug fixes plus just a couple functionality changes.
1331 On the user-visible side, the ``gnt-* list`` command output has changed
1332 with respect to "special" field states. The current rc1 style of display
1333 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
1334 the default output mode special fields states are displayed as follows:
1336 - Offline resource: ``*``
1337 - Unavailable/not applicable: ``-``
1338 - Data missing (RPC failure): ``?``
1339 - Unknown field: ``??``
1341 Another user-visible change is the addition of ``--force-join`` to
1346 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
1347 - Fixed regression in RAPI/instance reinstall where all parameters were
1348 required (instead of optional)
1349 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
1350 - Fixed iallocator usage (offline nodes were not considered offline)
1351 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
1352 - Fixed hypervisor and OS parameter validation with respect to
1353 non-vm_capable nodes
1354 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
1356 - Fixed ``tools/listrunner`` with respect to agent-based usage
1362 *(Released Fri, 4 Feb 2011)*
1364 Many changes and fixes since the beta1 release. While there were some
1365 internal changes, the code has been mostly stabilised for the RC
1368 Note: the dumb allocator was removed in this release, as it was not kept
1369 up-to-date with the IAllocator protocol changes. It is recommended to
1370 use the ``hail`` command from the ganeti-htools package.
1372 Note: the 2.4 and up versions of Ganeti are not compatible with the
1373 0.2.x branch of ganeti-htools. You need to upgrade to
1374 ganeti-htools-0.3.0 (or later).
1376 Regressions fixed from 2.3
1377 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1379 - Fixed the ``gnt-cluster verify-disks`` command
1380 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
1382 - Fixed disk adoption breakage
1383 - Fixed wrong headers in instance listing for field aliases
1388 - Fixed corner case in KVM handling of NICs
1389 - Fixed many cases of wrong handling of non-vm_capable nodes
1390 - Fixed a bug where a missing instance symlink was not possible to
1391 recreate with any ``gnt-*`` command (now ``gnt-instance
1392 activate-disks`` does it)
1393 - Fixed the volume group name as reported by ``gnt-cluster
1395 - Increased timeouts for the import-export code, hopefully leading to
1396 fewer aborts due network or instance timeouts
1397 - Fixed bug in ``gnt-node list-storage``
1398 - Fixed bug where not all daemons were started on cluster
1399 initialisation, but only at the first watcher run
1400 - Fixed many bugs in the OOB implementation
1401 - Fixed watcher behaviour in presence of instances with offline
1403 - Fixed instance list output for instances running on the wrong node
1404 - a few fixes to the cluster-merge tool, but it still cannot merge
1405 multi-node groups (currently it is not recommended to use this tool)
1411 - Improved network configuration for the KVM hypervisor
1412 - Added e1000 as a supported NIC for Xen-HVM
1413 - Improved the lvmstrap tool to also be able to use partitions, as
1414 opposed to full disks
1415 - Improved speed of disk wiping (the cluster parameter
1416 ``prealloc_wipe_disks``, so that it has a low impact on the total time
1417 of instance creations
1418 - Added documentation for the OS parameters
1419 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
1420 hypervisor is not responding
1421 - Added display of blacklisted and hidden OS information in
1422 ``gnt-cluster info``
1423 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
1424 NIC and node parameters, which might create problems with currently
1425 invalid (but undetected) configuration files, but prevents validation
1426 failures when unrelated parameters are modified
1427 - Changed cluster initialisation to wait for the master daemon to become
1429 - Expanded the RAPI interface:
1431 - Added config redistribution resource
1432 - Added activation/deactivation of instance disks
1433 - Added export of console information
1435 - Implemented log file reopening on SIGHUP, which allows using
1436 logrotate(8) for the Ganeti log files
1437 - Added a basic OOB helper script as an example
1443 *(Released Fri, 14 Jan 2011)*
1448 - Fixed timezone issues when formatting timestamps
1449 - Added support for node groups, available via ``gnt-group`` and other
1451 - Added out-of-band framework and management, see :doc:`design
1452 document <design-oob>`
1453 - Removed support for roman numbers from ``gnt-node list`` and
1454 ``gnt-instance list``.
1455 - Allowed modification of master network interface via ``gnt-cluster
1456 modify --master-netdev``
1457 - Accept offline secondaries while shutting down instance disks
1458 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
1459 - Added support for multiple LVM volume groups
1460 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
1462 - Added commands to list available fields:
1464 - ``gnt-node list-fields``
1465 - ``gnt-group list-fields``
1466 - ``gnt-instance list-fields``
1468 - Updated documentation and man pages
1473 - Moved ``rapi_users`` file into separate directory, now named
1474 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
1476 - Added new tool for running commands on many machines,
1477 ``tools/ganeti-listrunner``
1478 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
1479 improving the ``gnt-instance console`` output
1480 - Allowed customisation of disk index separator at ``configure`` time
1481 - Export node group allocation policy to :doc:`iallocator <iallocator>`
1482 - Added support for non-partitioned md disks in ``lvmstrap``
1483 - Added script to gracefully power off KVM instances
1484 - Split ``utils`` module into smaller parts
1485 - Changed query operations to return more detailed information, e.g.
1486 whether an information is unavailable due to an offline node. To use
1487 this new functionality, the LUXI call ``Query`` must be used. Field
1488 information is now stored by the master daemon and can be retrieved
1489 using ``QueryFields``. Instances, nodes and groups can also be queried
1490 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
1491 exposed via RAPI). The following commands make use of this
1492 infrastructure change:
1494 - ``gnt-group list``
1495 - ``gnt-group list-fields``
1497 - ``gnt-node list-fields``
1498 - ``gnt-instance list``
1499 - ``gnt-instance list-fields``
1500 - ``gnt-debug locks``
1505 - New RAPI resources (see :doc:`rapi`):
1509 - ``/2/groups/[group_name]``
1510 - ``/2/groups/[group_name]/assign-nodes``
1511 - ``/2/groups/[group_name]/modify``
1512 - ``/2/groups/[group_name]/rename``
1513 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
1517 - Implemented ``no_install`` for instance creation
1518 - Implemented OS parameters for instance reinstallation, allowing
1519 use of special settings on reinstallation (e.g. for preserving data)
1524 - Added IPv6 support in import/export
1525 - Pause DRBD synchronization while wiping disks on instance creation
1526 - Updated unittests and QA scripts
1527 - Improved network parameters passed to KVM
1528 - Converted man pages from docbook to reStructuredText
1534 *(Released Mon, 20 Dec 2010)*
1536 Released version 2.3.1~rc1 without any changes.
1542 *(Released Wed, 1 Dec 2010)*
1544 - impexpd: Disable OpenSSL compression in socat if possible (backport
1545 from master, commit e90739d625b, see :doc:`installation guide
1546 <install-quick>` for details)
1547 - Changed unittest coverage report to exclude test scripts
1548 - Added script to check version format
1554 *(Released Wed, 1 Dec 2010)*
1556 Released version 2.3.0~rc1 without any changes.
1562 *(Released Fri, 19 Nov 2010)*
1564 A number of bugfixes and documentation updates:
1566 - Update ganeti-os-interface documentation
1567 - Fixed a bug related to duplicate MACs or similar items which should be
1569 - Fix breakage in OS state modify
1570 - Reinstall instance: disallow offline secondaries (fixes bug related to
1571 OS changing but reinstall failing)
1572 - plus all the other fixes between 2.2.1 and 2.2.2
1578 *(Released Tue, 2 Nov 2010)*
1580 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
1581 - Fixed master failover race with watcher
1582 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
1584 - Accept previously stopped instance for export with instance removal
1585 - Simplify and extend the environment variables for instance OS scripts
1586 - Added new node flags, ``master_capable`` and ``vm_capable``
1587 - Added optional instance disk wiping prior during allocation. This is a
1588 cluster-wide option and can be set/modified using
1589 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
1590 - Added IPv6 support, see :doc:`design document <design-2.3>` and
1591 :doc:`install-quick`
1592 - Added a new watcher option (``--ignore-pause``)
1593 - Added option to ignore offline node on instance start/stop
1594 (``--ignore-offline``)
1595 - Allow overriding OS parameters with ``gnt-instance reinstall``
1596 - Added ability to change node's secondary IP address using ``gnt-node
1598 - Implemented privilege separation for all daemons except
1599 ``ganeti-noded``, see ``configure`` options
1600 - Complain if an instance's disk is marked faulty in ``gnt-cluster
1602 - Implemented job priorities (see ``ganeti(7)`` manpage)
1603 - Ignore failures while shutting down instances during failover from
1605 - Exit daemon's bootstrap process only once daemon is ready
1606 - Export more information via ``LUInstanceQuery``/remote API
1607 - Improved documentation, QA and unittests
1608 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
1609 restart if the file was created or changed
1610 - Added LUXI protocol version sent with each request and response,
1611 allowing detection of server/client mismatches
1612 - Moved the Python scripts among gnt-* and ganeti-* into modules
1613 - Moved all code related to setting up SSH to an external script,
1615 - Infrastructure changes for node group support in future versions
1621 *(Released Fri, 19 Nov 2010)*
1623 A few small bugs fixed, and some improvements to the build system:
1625 - Fix documentation regarding conversion to drbd
1626 - Fix validation of parameters in cluster modify (``gnt-cluster modify
1628 - Fix error handling in node modify with multiple changes
1629 - Allow remote imports without checked names
1635 *(Released Tue, 19 Oct 2010)*
1637 - Disable SSL session ID cache in RPC client
1643 *(Released Thu, 14 Oct 2010)*
1645 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
1646 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
1648 - Fix problems with interaction between Curl and Python's HTTP server,
1649 resulting in increased speed in many RPC calls
1650 - Improve our release script to prevent breakage with older aclocal and
1657 *(Released Thu, 7 Oct 2010)*
1659 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
1660 value retrieved from master
1661 - Added support for blacklisted or hidden OS definitions
1662 - Added simple lock monitor (accessible via (``gnt-debug locks``)
1663 - Added support for -mem-path in KVM hypervisor abstraction layer
1664 - Allow overriding instance parameters in tool for inter-cluster
1665 instance moves (``tools/move-instance``)
1666 - Improved opcode summaries (e.g. in ``gnt-job list``)
1667 - Improve consistency of OS listing by sorting it
1668 - Documentation updates
1674 *(Released Fri, 8 Oct 2010)*
1676 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
1682 *(Released Mon, 4 Oct 2010)*
1684 - Fixed regression in ``gnt-instance rename``
1690 *(Released Wed, 22 Sep 2010)*
1692 - Fixed OS_VARIANT variable for OS scripts
1693 - Fixed cluster tag operations via RAPI
1694 - Made ``setup-ssh`` exit with non-zero code if an error occurred
1695 - Disabled RAPI CA checks in watcher
1701 *(Released Mon, 23 Aug 2010)*
1703 - Support DRBD versions of the format "a.b.c.d"
1705 - Re-introduce support for usage from multiple threads in RAPI client
1706 - Instance renames and modify via RAPI
1707 - Work around race condition between processing and archival in job
1709 - Mark opcodes following failed one as failed, too
1710 - Job field ``lock_status`` was removed due to difficulties making it
1711 work with the changed job queue in Ganeti 2.2; a better way to monitor
1712 locks is expected for a later 2.2.x release
1713 - Fixed dry-run behaviour with many commands
1714 - Support ``ssh-agent`` again when adding nodes
1715 - Many additional bugfixes
1721 *(Released Fri, 30 Jul 2010)*
1723 Important change: the internal RPC mechanism between Ganeti nodes has
1724 changed from using a home-grown http library (based on the Python base
1725 libraries) to use the PycURL library. This requires that PycURL is
1726 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
1727 against GnuTLS by default. cURL's support for GnuTLS had known issues
1728 before cURL 7.21.0 and we recommend using the latest cURL release or
1729 linking against OpenSSL. Most other distributions already link PycURL
1730 and cURL against OpenSSL. The command::
1732 python -c 'import pycurl; print pycurl.version'
1734 can be used to determine the libraries PycURL and cURL are linked
1737 Other significant changes:
1739 - Rewrote much of the internals of the job queue, in order to achieve
1740 better parallelism; this decouples job query operations from the job
1741 processing, and it should allow much nicer behaviour of the master
1742 daemon under load, and it also has uncovered some long-standing bugs
1743 related to the job serialisation (now fixed)
1744 - Added a default iallocator setting to the cluster parameters,
1745 eliminating the need to always pass nodes or an iallocator for
1746 operations that require selection of new node(s)
1747 - Added experimental support for the LXC virtualization method
1748 - Added support for OS parameters, which allows the installation of
1749 instances to pass parameter to OS scripts in order to customise the
1751 - Added a hypervisor parameter controlling the migration type (live or
1752 non-live), since hypervisors have various levels of reliability; this
1753 has renamed the 'live' parameter to 'mode'
1754 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
1755 logical volumes, meaning that cluster verify will ignore them and not
1756 flag their presence as errors
1757 - The watcher will now reset the error count for failed instances after
1758 8 hours, thus allowing self-healing if the problem that caused the
1759 instances to be down/fail to start has cleared in the meantime
1760 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
1761 check for, and warn, if the drbd module parameter ``usermode_helper``
1762 is not consistent with the cluster-wide setting; this is needed to
1763 make diagnose easier of failed drbd creations
1764 - Started adding base IPv6 support, but this is not yet
1765 enabled/available for use
1766 - Rename operations (cluster, instance) will now return the new name,
1767 which is especially useful if a short name was passed in
1768 - Added support for instance migration in RAPI
1769 - Added a tool to pre-configure nodes for the SSH setup, before joining
1770 them to the cluster; this will allow in the future a simplified model
1771 for node joining (but not yet fully enabled in 2.2); this needs the
1772 paramiko python library
1773 - Fixed handling of name-resolving errors
1774 - Fixed consistency of job results on the error path
1775 - Fixed master-failover race condition when executed multiple times in
1777 - Fixed many bugs related to the job queue (mostly introduced during the
1778 2.2 development cycle, so not all are impacting 2.1)
1779 - Fixed instance migration with missing disk symlinks
1780 - Fixed handling of unknown jobs in ``gnt-job archive``
1781 - And many other small fixes/improvements
1785 - Enhanced both the unittest and the QA coverage
1786 - Switched the opcode validation to a generic model, and extended the
1787 validation to all opcode parameters
1788 - Changed more parts of the code that write shell scripts to use the
1790 - Switched the master daemon to use the asyncore library for the Luxi
1797 *(Released Thu, 17 Jun 2010)*
1799 - Added tool (``move-instance``) and infrastructure to move instances
1800 between separate clusters (see :doc:`separate documentation
1801 <move-instance>` and :doc:`design document <design-2.2>`)
1802 - Added per-request RPC timeout
1803 - RAPI now requires a Content-Type header for requests with a body (e.g.
1804 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
1805 :rfc:`2616` (HTTP/1.1), section 7.2.1)
1806 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
1808 - Implemented initial support for running Ganeti daemons as separate
1809 users, see configure-time flags ``--with-user-prefix`` and
1810 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
1812 - Instances can be removed after export (``gnt-backup export
1813 --remove-instance``)
1814 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
1815 key (instead of 1024 bit)
1816 - Added new cluster configuration file for cluster domain secret
1817 - Import/export now use SSL instead of SSH
1818 - Added support for showing estimated time when exporting an instance,
1819 see the ``ganeti-os-interface(7)`` manpage and look for
1826 *(Released Tue, 16 Nov 2010)*
1828 Some more bugfixes. Unless critical bugs occur, this will be the last
1831 - Fix case of MAC special-values
1832 - Fix mac checker regex
1833 - backend: Fix typo causing "out of range" error
1834 - Add missing --units in gnt-instance list man page
1840 *(Released Tue, 24 Aug 2010)*
1843 - Don't ignore secondary node silently on non-mirrored disk templates
1845 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
1846 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
1847 - Properly document the usb_mouse parameter
1848 - Fix path in ganeti-rapi(8) (issue 116)
1849 - Adjust error message when the ganeti user's .ssh directory is
1851 - Add same-node-check when changing the disk template to drbd
1857 *(Released Fri, 16 Jul 2010)*
1860 - Add an option to only select some reboot types during qa/burnin.
1861 (on some hypervisors consequent reboots are not supported)
1862 - Fix infrequent race condition in master failover. Sometimes the old
1863 master ip address would be still detected as up for a short time
1864 after it was removed, causing failover to fail.
1865 - Decrease mlockall warnings when the ctypes module is missing. On
1866 Python 2.4 we support running even if no ctypes module is installed,
1867 but we were too verbose about this issue.
1868 - Fix building on old distributions, on which man doesn't have a
1870 - Fix RAPI not to ignore the MAC address on instance creation
1871 - Implement the old instance creation format in the RAPI client.
1877 *(Released Thu, 01 Jul 2010)*
1879 A small bugfix release:
1880 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
1881 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
1882 a non-existing option
1883 - Fix instance create: the --force-variant option was ignored
1884 - Improve pylint 0.21 compatibility and warnings with Python 2.6
1885 - Fix modify node storage with non-FQDN arguments
1886 - Fix RAPI client to authenticate under Python 2.6 when used
1887 for more than 5 requests needing authentication
1888 - Fix gnt-instance modify -t (storage) giving a wrong error message
1889 when converting a non-shutdown drbd instance to plain
1895 *(Released Fri, 18 Jun 2010)*
1897 A small bugfix release:
1899 - Fix live migration of KVM instances started with older Ganeti
1900 versions which had fewer hypervisor parameters
1901 - Fix gnt-instance grow-disk on down instances
1902 - Fix an error-reporting bug during instance migration
1903 - Better checking of the ``--net`` and ``--disk`` values, to avoid
1904 silently ignoring broken ones
1905 - Fix an RPC error reporting bug affecting, for example, RAPI client
1907 - Fix bug triggered by different API version os-es on different nodes
1908 - Fix a bug in instance startup with custom hvparams: OS level
1909 parameters would fail to be applied.
1910 - Fix the RAPI client under Python 2.6 (but more work is needed to
1911 make it work completely well with OpenSSL)
1912 - Fix handling of errors when resolving names from DNS
1918 *(Released Thu, 3 Jun 2010)*
1920 A medium sized development cycle. Some new features, and some
1921 fixes/small improvements/cleanups.
1923 Significant features
1924 ~~~~~~~~~~~~~~~~~~~~
1926 The node deamon now tries to mlock itself into memory, unless the
1927 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
1928 its logs, and falls back to console logging. This allows emergency
1929 features such as ``gnt-node powercycle`` to work even in the event of a
1930 broken node disk (tested offlining the disk hosting the node's
1931 filesystem and dropping its memory caches; don't try this at home)
1933 KVM: add vhost-net acceleration support. It can be tested with a new
1934 enough version of the kernel and of qemu-kvm.
1936 KVM: Add instance chrooting feature. If you use privilege dropping for
1937 your VMs you can also now force them to chroot to an empty directory,
1938 before starting the emulated guest.
1940 KVM: Add maximum migration bandwith and maximum downtime tweaking
1941 support (requires a new-enough version of qemu-kvm).
1943 Cluster verify will now warn if the master node doesn't have the master
1944 ip configured on it.
1946 Add a new (incompatible) instance creation request format to RAPI which
1947 supports all parameters (previously only a subset was supported, and it
1948 wasn't possible to extend the old format to accomodate all the new
1949 features. The old format is still supported, and a client can check for
1950 this feature, before using it, by checking for its presence in the
1951 ``features`` RAPI resource.
1953 Now with ancient latin support. Try it passing the ``--roman`` option to
1954 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
1955 (requires the python-roman module to be installed, in order to work).
1960 As usual many internal code refactorings, documentation updates, and
1963 - Lots of improvements and cleanups to the experimental Remote API
1964 (RAPI) client library.
1965 - A new unit test suite for the core daemon libraries.
1966 - A fix to creating missing directories makes sure the umask is not
1967 applied anymore. This enforces the same directory permissions
1969 - Better handling terminating daemons with ctrl+c (used when running
1970 them in debugging mode).
1971 - Fix a race condition in live migrating a KVM instance, when stat()
1972 on the old proc status file returned EINVAL, which is an unexpected
1974 - Fixed manpage checking with newer man and utf-8 charachters. But now
1975 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
1981 *(Released Fri, 7 May 2010)*
1983 Fix a bug which prevented untagged KVM instances from starting.
1989 *(Released Fri, 7 May 2010)*
1991 Another release with a long development cycle, during which many
1992 different features were added.
1994 Significant features
1995 ~~~~~~~~~~~~~~~~~~~~
1997 The KVM hypervisor now can run the individual instances as non-root, to
1998 reduce the impact of a VM being hijacked due to bugs in the
1999 hypervisor. It is possible to run all instances as a single (non-root)
2000 user, to manually specify a user for each instance, or to dynamically
2001 allocate a user out of a cluster-wide pool to each instance, with the
2002 guarantee that no two instances will run under the same user ID on any
2005 An experimental RAPI client library, that can be used standalone
2006 (without the other Ganeti libraries), is provided in the source tree as
2007 ``lib/rapi/client.py``. Note this client might change its interface in
2008 the future, as we iterate on its capabilities.
2010 A new command, ``gnt-cluster renew-crypto`` has been added to easily
2011 replace the cluster's certificates and crypto keys. This might help in
2012 case they have been compromised, or have simply expired.
2014 A new disk option for instance creation has been added that allows one
2015 to "adopt" currently existing logical volumes, with data
2016 preservation. This should allow easier migration to Ganeti from
2017 unmanaged (or managed via other software) instances.
2019 Another disk improvement is the possibility to convert between redundant
2020 (DRBD) and plain (LVM) disk configuration for an instance. This should
2021 allow better scalability (starting with one node and growing the
2022 cluster, or shrinking a two-node cluster to one node).
2024 A new feature that could help with automated node failovers has been
2025 implemented: if a node sees itself as offline (by querying the master
2026 candidates), it will try to shutdown (hard) all instances and any active
2027 DRBD devices. This reduces the risk of duplicate instances if an
2028 external script automatically failovers the instances on such nodes. To
2029 enable this, the cluster parameter ``maintain_node_health`` should be
2030 enabled; in the future this option (per the name) will enable other
2031 automatic maintenance features.
2033 Instance export/import now will reuse the original instance
2034 specifications for all parameters; that means exporting an instance,
2035 deleting it and the importing it back should give an almost identical
2036 instance. Note that the default import behaviour has changed from
2037 before, where it created only one NIC; now it recreates the original
2040 Cluster verify has added a few new checks: SSL certificates validity,
2041 /etc/hosts consistency across the cluster, etc.
2046 As usual, many internal changes were done, documentation fixes,
2049 - Fixed cluster initialization with disabled cluster storage (regression
2050 introduced in 2.1.1)
2051 - File-based storage supports growing the disks
2052 - Fixed behaviour of node role changes
2053 - Fixed cluster verify for some corner cases, plus a general rewrite of
2054 cluster verify to allow future extension with more checks
2055 - Fixed log spamming by watcher and node daemon (regression introduced
2057 - Fixed possible validation issues when changing the list of enabled
2059 - Fixed cleanup of /etc/hosts during node removal
2060 - Fixed RAPI response for invalid methods
2061 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
2062 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
2063 from ide disks, etc.)
2064 - Allow OS changes without re-installation (to record a changed OS
2065 outside of Ganeti, or to allow OS renames)
2066 - Allow instance creation without OS installation (useful for example if
2067 the OS will be installed manually, or restored from a backup not in
2069 - Implemented option to make cluster ``copyfile`` use the replication
2071 - Added list of enabled hypervisors to ssconf (possibly useful for
2073 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
2075 - A partial form of node re-IP is possible via node readd, which now
2076 allows changed node primary IP
2077 - Command line utilities now show an informational message if the job is
2079 - The logs of the master daemon now show the PID/UID/GID of the
2086 *(Released Fri, 12 Mar 2010)*
2088 During the 2.1.0 long release candidate cycle, a lot of improvements and
2089 changes have accumulated with were released later as 2.1.1.
2094 The node evacuate command (``gnt-node evacuate``) was significantly
2095 rewritten, and as such the IAllocator protocol was changed - a new
2096 request type has been added. This unfortunate change during a stable
2097 series is designed to improve performance of node evacuations; on
2098 clusters with more than about five nodes and which are well-balanced,
2099 evacuation should proceed in parallel for all instances of the node
2100 being evacuated. As such, any existing IAllocator scripts need to be
2101 updated, otherwise the above command will fail due to the unknown
2102 request. The provided "dumb" allocator has not been updated; but the
2103 ganeti-htools package supports the new protocol since version 0.2.4.
2105 Another important change is increased validation of node and instance
2106 names. This might create problems in special cases, if invalid host
2107 names are being used.
2109 Also, a new layer of hypervisor parameters has been added, that sits at
2110 OS level between the cluster defaults and the instance ones. This allows
2111 customisation of virtualization parameters depending on the installed
2112 OS. For example instances with OS 'X' may have a different KVM kernel
2113 (or any other parameter) than the cluster defaults. This is intended to
2114 help managing a multiple OSes on the same cluster, without manual
2115 modification of each instance's parameters.
2117 A tool for merging clusters, ``cluster-merge``, has been added in the
2118 tools sub-directory.
2123 - Improved the int/float conversions that should make the code more
2124 robust in face of errors from the node daemons
2125 - Fixed the remove node code in case of internal configuration errors
2126 - Fixed the node daemon behaviour in face of inconsistent queue
2127 directory (e.g. read-only file-system where we can't open the files
2129 - Fixed the behaviour of gnt-node modify for master candidate demotion;
2130 now it either aborts cleanly or, if given the new "auto_promote"
2131 parameter, will automatically promote other nodes as needed
2132 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
2133 completely prevent Ganeti from working
2134 - Fixed bug for instance export when not all disks were successfully
2136 - Fixed behaviour of node add when the new node is slow in starting up
2138 - Fixed handling of signals in the LUXI client, which should improve
2139 behaviour of command-line scripts
2140 - Added checks for invalid node/instance names in the configuration (now
2141 flagged during cluster verify)
2142 - Fixed watcher behaviour for disk activation errors
2143 - Fixed two potentially endless loops in http library, which led to the
2144 RAPI daemon hanging and consuming 100% CPU in some cases
2145 - Fixed bug in RAPI daemon related to hashed passwords
2146 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
2148 - Enhanced compatibility with non-Debian OSes, but not using absolute
2149 path in some commands and allowing customisation of the ssh
2150 configuration directory
2151 - Fixed possible future issue with new Python versions by abiding to the
2152 proper use of ``__slots__`` attribute on classes
2153 - Added checks that should prevent directory traversal attacks
2154 - Many documentation fixes based on feedback from users
2159 - Added an "early_release" more for instance replace disks and node
2160 evacuate, where we release locks earlier and thus allow higher
2161 parallelism within the cluster
2162 - Added watcher hooks, intended to allow the watcher to restart other
2163 daemons (e.g. from the ganeti-nbma project), but they can be used of
2164 course for any other purpose
2165 - Added a compile-time disable for DRBD barriers, to increase
2166 performance if the administrator trusts the power supply or the
2167 storage system to not lose writes
2168 - Added the option of using syslog for logging instead of, or in
2169 addition to, Ganeti's own log files
2170 - Removed boot restriction for paravirtual NICs for KVM, recent versions
2171 can indeed boot from a paravirtual NIC
2172 - Added a generic debug level for many operations; while this is not
2173 used widely yet, it allows one to pass the debug value all the way to
2175 - Enhanced the hooks environment for instance moves (failovers,
2176 migrations) where the primary/secondary nodes changed during the
2177 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
2178 - Enhanced data validations for many user-supplied values; one important
2179 item is the restrictions imposed on instance and node names, which
2180 might reject some (invalid) host names
2181 - Add a configure-time option to disable file-based storage, if it's not
2182 needed; this allows greater security separation between the master
2183 node and the other nodes from the point of view of the inter-node RPC
2185 - Added user notification in interactive tools if job is waiting in the
2186 job queue or trying to acquire locks
2187 - Added log messages when a job is waiting for locks
2188 - Added filtering by node tags in instance operations which admit
2189 multiple instances (start, stop, reboot, reinstall)
2190 - Added a new tool for cluster mergers, ``cluster-merge``
2191 - Parameters from command line which are of the form ``a=b,c=d`` can now
2192 use backslash escapes to pass in values which contain commas,
2193 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
2195 - For KVM, the instance name is the first parameter passed to KVM, so
2196 that it's more visible in the process list
2202 *(Released Tue, 2 Mar 2010)*
2204 Ganeti 2.1 brings many improvements with it. Major changes:
2206 - Added infrastructure to ease automated disk repairs
2207 - Added new daemon to export configuration data in a cheaper way than
2208 using the remote API
2209 - Instance NICs can now be routed instead of being associated with a
2211 - Improved job locking logic to reduce impact of jobs acquiring multiple
2212 locks waiting for other long-running jobs
2214 In-depth implementation details can be found in the Ganeti 2.1 design
2220 - Added chroot hypervisor
2221 - Added more options to xen-hvm hypervisor (``kernel_path`` and
2223 - Added more options to xen-pvm hypervisor (``use_bootloader``,
2224 ``bootloader_path`` and ``bootloader_args``)
2225 - Added the ``use_localtime`` option for the xen-hvm and kvm
2226 hypervisors, and the default value for this has changed to false (in
2227 2.0 xen-hvm always enabled it)
2228 - Added luxi call to submit multiple jobs in one go
2229 - Added cluster initialization option to not modify ``/etc/hosts``
2231 - Added network interface parameters
2232 - Added dry run mode to some LUs
2233 - Added RAPI resources:
2235 - ``/2/instances/[instance_name]/info``
2236 - ``/2/instances/[instance_name]/replace-disks``
2237 - ``/2/nodes/[node_name]/evacuate``
2238 - ``/2/nodes/[node_name]/migrate``
2239 - ``/2/nodes/[node_name]/role``
2240 - ``/2/nodes/[node_name]/storage``
2241 - ``/2/nodes/[node_name]/storage/modify``
2242 - ``/2/nodes/[node_name]/storage/repair``
2244 - Added OpCodes to evacuate or migrate all instances on a node
2245 - Added new command to list storage elements on nodes (``gnt-node
2246 list-storage``) and modify them (``gnt-node modify-storage``)
2247 - Added new ssconf files with master candidate IP address
2248 (``ssconf_master_candidates_ips``), node primary IP address
2249 (``ssconf_node_primary_ips``) and node secondary IP address
2250 (``ssconf_node_secondary_ips``)
2251 - Added ``ganeti-confd`` and a client library to query the Ganeti
2252 configuration via UDP
2253 - Added ability to run hooks after cluster initialization and before
2255 - Added automatic mode for disk replace (``gnt-instance replace-disks
2257 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
2258 after catastrophic data-loss
2259 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
2261 - Added ``gnt-instance move`` command to move instances
2262 - Added ``gnt-cluster watcher`` command to control watcher
2263 - Added ``gnt-node powercycle`` command to powercycle nodes
2264 - Added new job status field ``lock_status``
2265 - Added parseable error codes to cluster verification (``gnt-cluster
2266 verify --error-codes``) and made output less verbose (use
2267 ``--verbose`` to restore previous behaviour)
2268 - Added UUIDs to the main config entities (cluster, nodes, instances)
2269 - Added support for OS variants
2270 - Added support for hashed passwords in the Ganeti remote API users file
2272 - Added option to specify maximum timeout on instance shutdown
2273 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
2274 - Added new helper script to start and stop Ganeti daemons
2275 (``daemon-util``), with the intent to reduce the work necessary to
2276 adjust Ganeti for non-Debian distributions and to start/stop daemons
2278 - Added more unittests
2279 - Fixed critical bug in ganeti-masterd startup
2280 - Removed the configure-time ``kvm-migration-port`` parameter, this is
2281 now customisable at the cluster level for both the KVM and Xen
2282 hypervisors using the new ``migration_port`` parameter
2283 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
2284 reinstalling an instance
2285 - Allowed ``@`` in tag names
2286 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
2287 - Many documentation updates
2288 - Distribute hypervisor files on ``gnt-cluster redist-conf``
2289 - ``gnt-instance reinstall`` can now reinstall multiple instances
2290 - Updated many command line parameters
2291 - Introduced new OS API version 15
2292 - No longer support a default hypervisor
2293 - Treat virtual LVs as inexistent
2294 - Improved job locking logic to reduce lock contention
2295 - Match instance and node names case insensitively
2296 - Reimplemented bash completion script to be more complete
2303 *(Released Thu, 4 Feb 2010)*
2305 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
2306 - Fix a string formatting bug
2307 - Improve safety of the code in some error paths
2308 - Improve data validation in the master of values returned from nodes
2314 *(Released Thu, 17 Dec 2009)*
2316 - Fix security issue due to missing validation of iallocator names; this
2317 allows local and remote execution of arbitrary executables
2318 - Fix failure of gnt-node list during instance removal
2319 - Ship the RAPI documentation in the archive
2325 *(Released Wed, 30 Sep 2009)*
2327 - Fixed many wrong messages
2328 - Fixed a few bugs related to the locking library
2329 - Fixed MAC checking at instance creation time
2330 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
2331 - Fixed a few issues related to signal handling in both daemons and
2333 - Fixed the example startup script provided
2334 - Fixed insserv dependencies in the example startup script (patch from
2336 - Fixed handling of drained nodes in the iallocator framework
2337 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
2339 - Fixed error related to invalid job IDs in job polling
2340 - Fixed job/opcode persistence on unclean master shutdown
2341 - Fixed handling of partial job processing after unclean master
2343 - Fixed error reporting from LUs, previously all errors were converted
2344 into execution errors
2345 - Fixed error reporting from burnin
2346 - Decreased significantly the memory usage of the job queue
2347 - Optimised slightly multi-job submission
2348 - Optimised slightly opcode loading
2349 - Backported the multi-job submit framework from the development
2350 branch; multi-instance start and stop should be faster
2351 - Added script to clean archived jobs after 21 days; this will reduce
2352 the size of the queue directory
2353 - Added some extra checks in disk size tracking
2354 - Added an example ethers hook script
2355 - Added a cluster parameter that prevents Ganeti from modifying of
2357 - Added more node information to RAPI responses
2358 - Added a ``gnt-job watch`` command that allows following the ouput of a
2360 - Added a bind-address option to ganeti-rapi
2361 - Added more checks to the configuration verify
2362 - Enhanced the burnin script such that some operations can be retried
2364 - Converted instance reinstall to multi-instance model
2370 *(Released Fri, 7 Aug 2009)*
2372 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
2373 to allow using the pre-2.0.2 behaviour in activation, if any existing
2374 instances have mismatched disk sizes in the configuration
2375 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
2376 any configuration mismatches for disk sizes
2377 - Added ``gnt-master cluste-failover --no-voting`` to allow master
2378 failover to work on two-node clusters
2379 - Fixed the ``--net`` option of ``gnt-backup import``, which was
2381 - Fixed detection of OS script errors in ``gnt-backup export``
2382 - Fixed exit code of ``gnt-backup export``
2388 *(Released Fri, 17 Jul 2009)*
2390 - Added experimental support for stripped logical volumes; this should
2391 enhance performance but comes with a higher complexity in the block
2392 device handling; stripping is only enabled when passing
2393 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
2394 affected even in the non-stripped mode
2395 - Improved resiliency against transient failures at the end of DRBD
2396 resyncs, and in general of DRBD resync checks
2397 - Fixed a couple of issues with exports and snapshot errors
2398 - Fixed a couple of issues in instance listing
2399 - Added display of the disk size in ``gnt-instance info``
2400 - Fixed checking for valid OSes in instance creation
2401 - Fixed handling of the "vcpus" parameter in instance listing and in
2402 general of invalid parameters
2403 - Fixed http server library, and thus RAPI, to handle invalid
2404 username/password combinations correctly; this means that now they
2405 report unauthorized for queries too, not only for modifications,
2406 allowing earlier detect of configuration problems
2407 - Added a new "role" node list field, equivalent to the master/master
2408 candidate/drained/offline flags combinations
2409 - Fixed cluster modify and changes of candidate pool size
2410 - Fixed cluster verify error messages for wrong files on regular nodes
2411 - Fixed a couple of issues with node demotion from master candidate role
2412 - Fixed node readd issues
2413 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
2414 - Added a new ``--no-voting`` option for masterfailover to fix failover
2415 on two-nodes clusters when the former master node is unreachable
2416 - Added instance reinstall over RAPI
2422 *(Released Tue, 16 Jun 2009)*
2424 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
2425 allow re-adding the start in single-user option (regression from 1.2)
2426 - the watcher writes the instance status to a file, to allow monitoring
2427 to report the instance status (from the master) based on cached
2428 results of the watcher's queries; while this can get stale if the
2429 watcher is being locked due to other work on the cluster, this is
2430 still an improvement
2431 - the watcher now also restarts the node daemon and the rapi daemon if
2433 - fixed the watcher to handle full and drained queue cases
2434 - hooks export more instance data in the environment, which helps if
2435 hook scripts need to take action based on the instance's properties
2436 (no longer need to query back into ganeti)
2437 - instance failovers when the instance is stopped do not check for free
2438 RAM, so that failing over a stopped instance is possible in low memory
2440 - rapi uses queries for tags instead of jobs (for less job traffic), and
2441 for cluster tags it won't talk to masterd at all but read them from
2443 - a couple of error handling fixes in RAPI
2444 - drbd handling: improved the error handling of inconsistent disks after
2445 resync to reduce the frequency of "there are some degraded disks for
2446 this instance" messages
2447 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
2448 error handling path called a wrong function name)
2454 *(Released Wed, 27 May 2009)*
2456 - no changes from rc5
2462 *(Released Wed, 20 May 2009)*
2464 - fix a couple of bugs (validation, argument checks)
2465 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
2466 - some small improvements to RAPI and IAllocator
2467 - make watcher automatically start the master daemon if down
2473 *(Released Mon, 27 Apr 2009)*
2475 - change the OS list to not require locks; this helps with big clusters
2476 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
2477 volume group is broken
2478 - ``gnt-instance info``, without any arguments, doesn't run for all
2479 instances anymore; either pass ``--all`` or pass the desired
2480 instances; this helps against mistakes on big clusters where listing
2481 the information for all instances takes a long time
2482 - miscellaneous doc and man pages fixes
2488 *(Released Wed, 8 Apr 2009)*
2490 - Change the internal locking model of some ``gnt-node`` commands, in
2491 order to reduce contention (and blocking of master daemon) when
2492 batching many creation/reinstall jobs
2493 - Fixes to Xen soft reboot
2494 - No longer build documentation at build time, instead distribute it in
2495 the archive, in order to reduce the need for the whole docbook/rst
2502 *(Released Fri, 27 Mar 2009)*
2504 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
2505 - Fix watcher startup sequence, improves the behaviour of busy clusters
2506 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
2507 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
2508 bind address and other places
2509 - Some documentation fixes and updates
2515 *(Released Mon, 2 Mar 2009)*
2517 - More documentation updates, now all docs should be more-or-less
2519 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
2521 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
2523 - Fix an issue related to $libdir/run/ganeti and cluster creation
2529 *(Released Thu, 19 Feb 2009)*
2531 - Xen PVM and KVM have switched the default value for the instance root
2532 disk to the first partition on the first drive, instead of the whole
2533 drive; this means that the OS installation scripts must be changed
2535 - Man pages have been updated
2536 - RAPI has been switched by default to HTTPS, and the exported functions
2537 should all work correctly
2538 - RAPI v1 has been removed
2539 - Many improvements to the KVM hypervisor
2540 - Block device errors are now better reported
2541 - Many other bugfixes and small improvements
2547 *(Released Mon, 26 Jan 2009)*
2549 - Version 2 is a general rewrite of the code and therefore the
2550 differences are too many to list, see the design document for 2.0 in
2551 the ``doc/`` subdirectory for more details
2552 - In this beta version there is not yet a migration path from 1.2 (there
2553 will be one in the final 2.0 release)
2554 - A few significant changes are:
2556 - all commands are executed by a daemon (``ganeti-masterd``) and the
2557 various ``gnt-*`` commands are just front-ends to it
2558 - all the commands are entered into, and executed from a job queue,
2559 see the ``gnt-job(8)`` manpage
2560 - the RAPI daemon supports read-write operations, secured by basic
2561 HTTP authentication on top of HTTPS
2562 - DRBD version 0.7 support has been removed, DRBD 8 is the only
2563 supported version (when migrating from Ganeti 1.2 to 2.0, you need
2564 to migrate to DRBD 8 first while still running Ganeti 1.2)
2565 - DRBD devices are using statically allocated minor numbers, which
2566 will be assigned to existing instances during the migration process
2567 - there is support for both Xen PVM and Xen HVM instances running on
2569 - KVM virtualization is supported too
2570 - file-based storage has been implemented, which means that it is
2571 possible to run the cluster without LVM and DRBD storage, for
2572 example using a shared filesystem exported from shared storage (and
2573 still have live migration)
2579 *(Released Tue, 13 Jan 2009)*
2581 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
2582 - Reuse the old instance mac address by default on instance import, if
2583 the instance name is the same.
2584 - Handle situations in which the node info rpc returns incomplete
2586 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
2587 - Improved version of batcher:
2589 - state file support
2590 - instance mac address support
2591 - support for HVM clusters/instances
2593 - Add an option to show the number of cpu sockets and nodes in
2595 - Support OSes that handle more than one version of the OS api (but do
2596 not change the current API in any other way)
2597 - Fix ``gnt-node migrate``
2598 - ``gnt-debug`` man page
2599 - Fixes various more typos and small issues
2600 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
2606 *(Released Wed, 24 Sep 2008)*
2608 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
2609 type of disk exported to fully virtualized instances.
2610 - provide access to the serial console of HVM instances
2611 - instance auto_balance flag, set by default. If turned off it will
2612 avoid warnings on cluster verify if there is not enough memory to fail
2613 over an instance. in the future it will prevent automatically failing
2614 it over when we will support that.
2615 - batcher tool for instance creation, see ``tools/README.batcher``
2616 - ``gnt-instance reinstall --select-os`` to interactively select a new
2617 operating system when reinstalling an instance.
2618 - when changing the memory amount on instance modify a check has been
2619 added that the instance will be able to start. also warnings are
2620 emitted if the instance will not be able to fail over, if auto_balance
2622 - documentation fixes
2623 - sync fields between ``gnt-instance list/modify/add/import``
2624 - fix a race condition in drbd when the sync speed was set after giving
2625 the device a remote peer.
2631 *(Released Tue, 22 Jul 2008)*
2633 - note: the allowed size and number of tags per object were reduced
2634 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
2635 - fixed twisted 8.x compatibility
2636 - fixed ``gnt-instance replace-disks`` with iallocator
2637 - add TCP keepalives on twisted connections to detect restarted nodes
2638 - disk increase support, see ``gnt-instance grow-disk``
2639 - implement bulk node/instance query for RAPI
2640 - add tags in node/instance listing (optional)
2641 - experimental migration (and live migration) support, read the man page
2642 for ``gnt-instance migrate``
2643 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
2644 has some small improvements in handling its state file
2650 *(Released Fri, 13 Jun 2008)*
2652 - Experimental readonly, REST-based remote API implementation;
2653 automatically started on master node, TCP port 5080, if enabled by
2654 ``--enable-rapi`` parameter to configure script.
2655 - Instance allocator support. Add and import instance accept a
2656 ``--iallocator`` parameter, and call that instance allocator to decide
2657 which node to use for the instance. The iallocator document describes
2658 what's expected from an allocator script.
2659 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
2660 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
2661 node is lost there is still enough memory to fail over the instances
2663 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
2664 ``gnt-cluster verify``, to check for site-specific compliance. All the
2665 hooks will run, and their output, if any, will be displayed. Any
2666 failing hook will make the verification return an error value.
2667 - ``gnt-cluster verify`` now checks that its peers are reachable on the
2668 primary and secondary interfaces
2669 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
2670 that is still declared as part of the cluster and has failed.
2671 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
2672 specifying output fields, that just adds the chosen fields to the
2674 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
2675 export from the filesystem.
2676 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
2677 have been added. Using them you can enable/disable acpi and pae
2678 support, and specify a path for a cd image to be exported to the
2679 instance. These parameters as the name suggest only work on HVM
2681 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
2682 PAE support will be set to the previously hardcoded values, but the
2683 (previously hardcoded) path to the CDROM ISO image will be unset and
2684 if required, needs to be set manually with ``gnt-instance modify``
2686 - The address to which an instance's VNC console is bound is now
2687 selectable per-instance, rather than being cluster wide. Of course
2688 this only applies to instances controlled via VNC, so currently just
2689 applies to HVM clusters.
2695 *(Released Mon, 18 Feb 2008)*
2697 - more tweaks to the disk activation code (especially helpful for DRBD)
2698 - change the default ``gnt-instance list`` output format, now there is
2699 one combined status field (see the manpage for the exact values this
2701 - some more fixes for the mac export to hooks change
2702 - make Ganeti not break with DRBD 8.2.x (which changed the version
2703 format in ``/proc/drbd``) (issue 24)
2704 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
2705 template, allowing migration from DRBD0.7+MD to DRBD8
2711 *(Released Wed, 30 Jan 2008)*
2713 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
2715 - add command aliases infrastructure and a few aliases
2716 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
2717 man pages and the ``--help`` option of ``gnt-node
2718 list``/``gnt-instance list``
2719 - fix ``gnt-backup list`` with down nodes (issue 21)
2720 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
2721 - fix the dist archive and add a check for including svn/git files in
2723 - some developer-related changes: improve the burnin and the QA suite,
2724 add an upload script for testing during development
2730 *(Released Wed, 16 Jan 2008)*
2732 - experimental HVM support, read the install document, section
2733 "Initializing the cluster"
2734 - allow for the PVM hypervisor per-instance kernel and initrd paths
2735 - add a new command ``gnt-cluster verify-disks`` which uses a new
2736 algorithm to improve the reconnection of the DRBD pairs if the device
2737 on the secondary node has gone away
2738 - make logical volume code auto-activate LVs at disk activation time
2739 - slightly improve the speed of activating disks
2740 - allow specification of the MAC address at instance creation time, and
2741 changing it later via ``gnt-instance modify``
2742 - fix handling of external commands that generate lots of output on
2744 - update documentation with regard to minimum version of DRBD8 supported
2750 *(Released Tue, 4 Dec 2007)*
2752 - Log the ``xm create`` output to the node daemon log on failure (to
2753 help diagnosing the error)
2754 - In debug mode, log all external commands output if failed to the logs
2755 - Change parsing of lvm commands to ignore stderr
2761 *(Released Wed, 28 Nov 2007)*
2763 - Another round of updates to the DRBD 8 code to deal with more failures
2764 in the replace secondary node operation
2765 - Some more logging of failures in disk operations (lvm, drbd)
2766 - A few documentation updates
2773 *(Released Tue, 13 Nov 2007)*
2775 - Change configuration file format from Python's Pickle to JSON.
2776 Upgrading is possible using the cfgupgrade utility.
2777 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
2778 faster replace disks and is more stable (DRBD 8 has many improvements
2779 compared to DRBD 0.7)
2780 - Added command line tags support (see man pages for ``gnt-instance``,
2781 ``gnt-node``, ``gnt-cluster``)
2782 - Added instance rename support
2783 - Added multi-instance startup/shutdown
2784 - Added cluster rename support
2785 - Added ``gnt-node evacuate`` to simplify some node operations
2786 - Added instance reboot operation that can speedup reboot as compared to
2788 - Soften the requirement that hostnames are in FQDN format
2789 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
2791 - Removed dependency on debian's patched fping that uses the
2792 non-standard ``-S`` option
2793 - Now the OS definitions are searched for in multiple, configurable
2794 paths (easier for distros to package)
2795 - Some changes to the hooks infrastructure (especially the new
2796 post-configuration update hook)
2797 - Other small bugfixes
2799 .. vim: set textwidth=72 syntax=rst :