8 *(Released Mon, 24 Jun 2013)*
10 Incompatible/important changes
11 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
13 - Instance policy can contain multiple instance specs, as described in
14 the “Constrained instance sizes” section of :doc:`Partitioned Ganeti
15 <design-partitioned>`. As a consequence, it's not possible to partially change
16 or override instance specs. Bounding specs (min and max) can be specified as a
17 whole using the new option ``--ipolicy-bounds-specs``, while standard
18 specs use the new option ``--ipolicy-std-specs``.
19 - The output of the info command of gnt-cluster, gnt-group, gnt-node,
20 gnt-instance is a valid YAML object.
21 - hail now honors network restrictions when allocating nodes. This led to an
22 update of the IAllocator protocol. See the IAllocator documentation for
28 - The :doc:`Remote API <rapi>` daemon now supports a command line flag
29 to always require authentication, ``--require-authentication``. It can
30 be specified in ``$sysconfdir/default/ganeti``.
31 - A new cluster attribute 'enabled_disk_templates' is introduced. It will
32 be used to manage the disk templates to be used by instances in the cluster.
33 Initially, it will be set to a list that includes plain, drbd, if they were
34 enabled by specifying a volume group name, and file and sharedfile, if those
35 were enabled at configure time. Additionally, it will include all disk
36 templates that are currently used by instances. The order of disk templates
37 will be based on Ganeti's history of supporting them. In the future, the
38 first entry of the list will be used as a default disk template on instance
40 - ``cfgupgrade`` now supports a ``--downgrade`` option to bring the
41 configuration back to the previous stable version.
42 - Disk templates in group ipolicy can be restored to the default value.
43 - Initial support for diskless instances and virtual clusters in QA.
44 - More QA and unit tests for instance policies.
45 - Every opcode now contains a reason trail (visible through ``gnt-job info``)
46 describing why the opcode itself was executed.
47 - The monitoring daemon is now available. It allows users to query the cluster
48 for obtaining information about the status of the system. The daemon is only
49 responsible for providing the information over the network: the actual data
50 gathering is performed by data collectors (currently, only the DRBD status
51 collector is available).
52 - In order to help developers work on Ganeti, a new script
53 (``devel/build_chroot``) is provided, for building a chroot that contains all
54 the required development libraries and tools for compiling Ganeti on a Debian
56 - A new tool, ``harep``, for performing self-repair and recreation of instances
57 in Ganeti has been added.
58 - Split queries are enabled for tags, network, exports, cluster info, groups,
60 - New command ``show-ispecs-cmd`` for ``gnt-cluster`` and ``gnt-group``.
61 It prints the command line to set the current policies, to ease
66 The following new dependencies have been added:
69 - The ``curl`` library is not optional anymore for compiling the Haskell code.
70 - ``snap-server`` library (if monitoring is enabled).
73 - The minimum Python version needed to run Ganeti is now 2.6.
74 - ``yaml`` library (only for running the QA).
80 *(Released Tue, 25 Jun 2013)*
82 Incompatible/important changes
83 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
85 - Instance policies for disk size were documented to be on a per-disk
86 basis, but hail applied them to the sum of all disks. This has been
88 - ``hbal`` will now exit with status 0 if, during job execution over
89 LUXI, early exit has been requested and all jobs are successful;
90 before, exit status 1 was used, which cannot be differentiated from
92 - Compatibility with newer versions of rbd has been fixed
93 - ``gnt-instance batch-create`` has been changed to use the bulk create
94 opcode from Ganeti. This lead to incompatible changes in the format of
95 the JSON file. It's now not a custom dict anymore but a dict
96 compatible with the ``OpInstanceCreate`` opcode.
97 - Parent directories for file storage need to be listed in
98 ``$sysconfdir/ganeti/file-storage-paths`` now. ``cfgupgrade`` will
99 write the file automatically based on old configuration values, but it
100 can not distribute it across all nodes and the file contents should be
101 verified. Use ``gnt-cluster copyfile
102 $sysconfdir/ganeti/file-storage-paths`` once the cluster has been
103 upgraded. The reason for requiring this list of paths now is that
104 before it would have been possible to inject new paths via RPC,
105 allowing files to be created in arbitrary locations. The RPC protocol
106 is protected using SSL/X.509 certificates, but as a design principle
107 Ganeti does not permit arbitrary paths to be passed.
108 - The parsing of the variants file for OSes (see
109 :manpage:`ganeti-os-interface(7)`) has been slightly changed: now empty
110 lines and comment lines (starting with ``#``) are ignored for better
112 - The ``setup-ssh`` tool added in Ganeti 2.2 has been replaced and is no
113 longer available. ``gnt-node add`` now invokes a new tool on the
114 destination node, named ``prepare-node-join``, to configure the SSH
115 daemon. Paramiko is no longer necessary to configure nodes' SSH
116 daemons via ``gnt-node add``.
117 - Draining (``gnt-cluster queue drain``) and un-draining the job queue
118 (``gnt-cluster queue undrain``) now affects all nodes in a cluster and
119 the flag is not reset after a master failover.
120 - Python 2.4 has *not* been tested with this release. Using 2.6 or above
121 is recommended. 2.6 will be mandatory from the 2.8 series.
127 - New network management functionality to support automatic allocation
128 of IP addresses and managing of network parameters. See
129 :manpage:`gnt-network(8)` for more details.
130 - New external storage backend, to allow managing arbitrary storage
131 systems external to the cluster. See
132 :manpage:`ganeti-extstorage-interface(7)`.
133 - New ``exclusive-storage`` node parameter added, restricted to
134 nodegroup level. When it's set to true, physical disks are assigned in
135 an exclusive fashion to instances, as documented in :doc:`Partitioned
136 Ganeti <design-partitioned>`. Currently, only instances using the
137 ``plain`` disk template are supported.
138 - The KVM hypervisor has been updated with many new hypervisor
139 parameters, including a generic one for passing arbitrary command line
140 values. See a complete list in :manpage:`gnt-instance(8)`. It is now
141 compatible up to qemu 1.4.
142 - A new tool, called ``mon-collector``, is the stand-alone executor of
143 the data collectors for a monitoring system. As of this version, it
144 just includes the DRBD data collector, that can be executed by calling
145 ``mon-collector`` using the ``drbd`` parameter. See
146 :manpage:`mon-collector(7)`.
147 - A new user option, :pyeval:`rapi.RAPI_ACCESS_READ`, has been added
148 for RAPI users. It allows granting permissions to query for
149 information to a specific user without giving
150 :pyeval:`rapi.RAPI_ACCESS_WRITE` permissions.
151 - A new tool named ``node-cleanup`` has been added. It cleans remains of
152 a cluster from a machine by stopping all daemons, removing
153 certificates and ssconf files. Unless the ``--no-backup`` option is
154 given, copies of the certificates are made.
155 - Instance creations now support the use of opportunistic locking,
156 potentially speeding up the (parallel) creation of multiple instances.
157 This feature is currently only available via the :doc:`RAPI
158 <rapi>` interface and when an instance allocator is used. If the
159 ``opportunistic_locking`` parameter is set the opcode will try to
160 acquire as many locks as possible, but will not wait for any locks
161 held by other opcodes. If not enough resources can be found to
162 allocate the instance, the temporary error code
163 :pyeval:`errors.ECODE_TEMP_NORES` is returned. The operation can be
164 retried thereafter, with or without opportunistic locking.
165 - New experimental linux-ha resource scripts.
166 - Restricted-commands support: ganeti can now be asked (via command line
167 or rapi) to perform commands on a node. These are passed via ganeti
168 RPC rather than ssh. This functionality is restricted to commands
169 specified on the ``$sysconfdir/ganeti/restricted-commands`` for security
170 reasons. The file is not copied automatically.
176 - Diskless instances are now externally mirrored (Issue 237). This for
177 now has only been tested in conjunction with explicit target nodes for
179 - Queries not needing locks or RPC access to the node can now be
180 performed by the confd daemon, making them independent from jobs, and
181 thus faster to execute. This is selectable at configure time.
182 - The functionality for allocating multiple instances at once has been
183 overhauled and is now also available through :doc:`RAPI <rapi>`.
187 - Fix permissions on the confd query socket (Issue 477)
188 - Fix permissions on the job archive dir (Issue 498)
189 - Fix handling of an internal exception in replace-disks (Issue 472)
190 - Fix gnt-node info handling of shortened names (Issue 497)
191 - Fix gnt-instance grow-disk when wiping is enabled
192 - Documentation improvements, and support for newer pandoc
193 - Fix hspace honoring ipolicy for disks (Issue 484)
194 - Improve handling of the ``kvm_extra`` HV parameter
200 *(Released Fri, 24 May 2013)*
202 - ``devel/upload`` now works when ``/var/run`` on the target nodes is a
204 - Disks added through ``gnt-instance modify`` or created through
205 ``gnt-instance recreate-disks`` are wiped, if the
206 ``prealloc_wipe_disks`` flag is set.
207 - If wiping newly created disks fails, the disks are removed. Also,
208 partial failures in creating disks through ``gnt-instance modify``
209 triggers a cleanup of the partially-created disks.
210 - Removing the master IP address doesn't fail if the address has been
212 - Fix ownership of the OS log dir
213 - Workaround missing SO_PEERCRED constant (Issue 191)
219 *(Released Fri, 3 May 2013)*
221 This was the first release candidate of the 2.7 series. Since beta3:
223 - Fix kvm compatibility with qemu 1.4 (Issue 389)
224 - Documentation updates (admin guide, upgrade notes, install
225 instructions) (Issue 372)
226 - Fix gnt-group list nodes and instances count (Issue 436)
227 - Fix compilation without non-mandatory libraries (Issue 441)
228 - Fix xen-hvm hypervisor forcing nics to type 'ioemu' (Issue 247)
229 - Make confd logging more verbose at INFO level (Issue 435)
230 - Improve "networks" documentation in :manpage:`gnt-instance(8)`
231 - Fix failure path for instance storage type conversion (Issue 229)
232 - Update htools text backend documentation
233 - Improve the renew-crypto section of :manpage:`gnt-cluster(8)`
234 - Disable inter-cluster instance move for file-based instances, because
235 it is dependant on instance export, which is not supported for
236 file-based instances. (Issue 414)
237 - Fix gnt-job crashes on non-ascii characters (Issue 427)
238 - Fix volume group checks on non-vm-capable nodes (Issue 432)
244 *(Released Mon, 22 Apr 2013)*
246 This was the third beta release of the 2.7 series. Since beta2:
248 - Fix hail to verify disk instance policies on a per-disk basis (Issue 418).
249 - Fix data loss on wrong usage of ``gnt-instance move``
250 - Properly export errors in confd-based job queries
251 - Add ``users-setup`` tool
252 - Fix iallocator protocol to report 0 as a disk size for diskless
253 instances. This avoids hail breaking when a diskless instance is
255 - Fix job queue directory permission problem that made confd job queries
256 fail. This requires running an ``ensure-dirs --full-run`` on upgrade
257 for access to archived jobs (Issue 406).
258 - Limit the sizes of networks supported by ``gnt-network`` to something
259 between a ``/16`` and a ``/30`` to prevent memory bloat and crashes.
260 - Fix bugs in instance disk template conversion
261 - Fix GHC 7 compatibility
262 - Fix ``burnin`` install path (Issue 426).
263 - Allow very small disk grows (Issue 347).
264 - Fix a ``ganeti-noded`` memory bloat introduced in 2.5, by making sure
265 that noded doesn't import masterd code (Issue 419).
266 - Make sure the default metavg at cluster init is the same as the vg, if
267 unspecified (Issue 358).
268 - Fix cleanup of partially created disks (part of Issue 416)
274 *(Released Tue, 2 Apr 2013)*
276 This was the second beta release of the 2.7 series. Since beta1:
278 - Networks no longer have a "type" slot, since this information was
279 unused in Ganeti: instead of it tags should be used.
280 - The rapi client now has a ``target_node`` option to MigrateInstance.
281 - Fix early exit return code for hbal (Issue 386).
282 - Fix ``gnt-instance migrate/failover -n`` (Issue 396).
283 - Fix ``rbd showmapped`` output parsing (Issue 312).
284 - Networks are now referenced indexed by UUID, rather than name. This
285 will require running cfgupgrade, from 2.7.0beta1, if networks are in
287 - The OS environment now includes network information.
288 - Deleting of a network is now disallowed if any instance nic is using
289 it, to prevent dangling references.
290 - External storage is now documented in man pages.
291 - The exclusive_storage flag can now only be set at nodegroup level.
292 - Hbal can now submit an explicit priority with its jobs.
293 - Many network related locking fixes.
294 - Bump up the required pylint version to 0.25.1.
295 - Fix the ``no_remember`` option in RAPI client.
296 - Many ipolicy related tests, qa, and fixes.
297 - Many documentation improvements and fixes.
298 - Fix building with ``--disable-file-storage``.
299 - Fix ``-q`` option in htools, which was broken if passed more than
301 - Some haskell/python interaction improvements and fixes.
302 - Fix iallocator in case of missing LVM storage.
303 - Fix confd config load in case of ``--no-lvm-storage``.
304 - The confd/query functionality is now mentioned in the security
311 *(Released Wed, 6 Feb 2013)*
313 This was the first beta release of the 2.7 series. All important changes
314 are listed in the latest 2.7 entry.
320 *(Released Fri, 21 Dec 2012)*
322 Important behaviour change: hbal won't rebalance anymore instances which
323 have the ``auto_balance`` attribute set to false. This was the intention
324 all along, but until now it only skipped those from the N+1 memory
325 reservation (DRBD-specific).
327 A significant number of bug fixes in this release:
329 - Fixed disk adoption interaction with ipolicy checks.
330 - Fixed networking issues when instances are started, stopped or
331 migrated, by forcing the tap device's MAC prefix to "fe" (issue 217).
332 - Fixed the warning in cluster verify for shared storage instances not
334 - Fixed removal of storage directory on shared file storage (issue 262).
335 - Fixed validation of LVM volume group name in OpClusterSetParams
336 (``gnt-cluster modify``) (issue 285).
337 - Fixed runtime memory increases (``gnt-instance modify -m``).
338 - Fixed live migration under Xen's ``xl`` mode.
339 - Fixed ``gnt-instance console`` with ``xl``.
340 - Fixed building with newer Haskell compiler/libraries.
341 - Fixed PID file writing in Haskell daemons (confd); this prevents
342 restart issues if confd was launched manually (outside of
343 ``daemon-util``) while another copy of it was running
344 - Fixed a type error when doing live migrations with KVM (issue 297) and
345 the error messages for failing migrations have been improved.
346 - Fixed opcode validation for the out-of-band commands (``gnt-node
348 - Fixed a type error when unsetting OS hypervisor parameters (issue
349 311); now it's possible to unset all OS-specific hypervisor
351 - Fixed the ``dry-run`` mode for many operations: verification of
352 results was over-zealous but didn't take into account the ``dry-run``
353 operation, resulting in "wrong" failures.
354 - Fixed bash completion in ``gnt-job list`` when the job queue has
355 hundreds of entries; especially with older ``bash`` versions, this
356 results in significant CPU usage.
358 And lastly, a few other improvements have been made:
360 - Added option to force master-failover without voting (issue 282).
361 - Clarified error message on lock conflict (issue 287).
362 - Logging of newly submitted jobs has been improved (issue 290).
363 - Hostname checks have been made uniform between instance rename and
365 - The ``--submit`` option is now supported by ``gnt-debug delay``.
366 - Shutting down the master daemon by sending SIGTERM now stops it from
367 processing jobs waiting for locks; instead, those jobs will be started
368 once again after the master daemon is started the next time (issue
370 - Support for Xen's ``xl`` program has been improved (besides the fixes
372 - Reduced logging noise in the Haskell confd daemon (only show one log
373 entry for each config reload, instead of two).
374 - Several man page updates and typo fixes.
380 *(Released Fri, 12 Oct 2012)*
382 A small bugfix release. Among the bugs fixed:
384 - Fixed double use of ``PRIORITY_OPT`` in ``gnt-node migrate``, that
385 made the command unusable.
386 - Commands that issue many jobs don't fail anymore just because some jobs
387 take so long that other jobs are archived.
388 - Failures during ``gnt-instance reinstall`` are reflected by the exit
390 - Issue 190 fixed. Check for DRBD in cluster verify is enabled only when
392 - When ``always_failover`` is set, ``--allow-failover`` is not required
393 in migrate commands anymore.
394 - ``bash_completion`` works even if extglob is disabled.
395 - Fixed bug with locks that made failover for RDB-based instances fail.
396 - Fixed bug in non-mirrored instance allocation that made Ganeti choose
397 a random node instead of one based on the allocator metric.
398 - Support for newer versions of pylint and pep8.
399 - Hail doesn't fail anymore when trying to add an instance of type
400 ``file``, ``sharedfile`` or ``rbd``.
401 - Added new Makefile target to rebuild the whole distribution, so that
402 all files are included.
408 *(Released Fri, 27 Jul 2012)*
411 .. attention:: The ``LUXI`` protocol has been made more consistent
412 regarding its handling of command arguments. This, however, leads to
413 incompatibility issues with previous versions. Please ensure that you
414 restart Ganeti daemons soon after the upgrade, otherwise most
415 ``LUXI`` calls (job submission, setting/resetting the drain flag,
416 pausing/resuming the watcher, cancelling and archiving jobs, querying
417 the cluster configuration) will fail.
426 The current ``admin_up`` field, which used to denote whether an instance
427 should be running or not, has been removed. Instead, ``admin_state`` is
428 introduced, with 3 possible values -- ``up``, ``down`` and ``offline``.
430 The rational behind this is that an instance being “down” can have
433 - it could be down during a reboot
434 - it could be temporarily be down for a reinstall
435 - or it could be down because it is deprecated and kept just for its
438 The previous Boolean state was making it difficult to do capacity
439 calculations: should Ganeti reserve memory for a down instance? Now, the
440 tri-state field makes it clear:
442 - in ``up`` and ``down`` state, all resources are reserved for the
443 instance, and it can be at any time brought up if it is down
444 - in ``offline`` state, only disk space is reserved for it, but not
447 The field can have an extra use: since the transition between ``up`` and
448 ``down`` and vice-versus is done via ``gnt-instance start/stop``, but
449 transition between ``offline`` and ``down`` is done via ``gnt-instance
450 modify``, it is possible to given different rights to users. For
451 example, owners of an instance could be allowed to start/stop it, but
452 not transition it out of the offline state.
454 Instance policies and specs
455 +++++++++++++++++++++++++++
457 In previous Ganeti versions, an instance creation request was not
458 limited on the minimum size and on the maximum size just by the cluster
459 resources. As such, any policy could be implemented only in third-party
460 clients (RAPI clients, or shell wrappers over ``gnt-*``
461 tools). Furthermore, calculating cluster capacity via ``hspace`` again
462 required external input with regards to instance sizes.
464 In order to improve these workflows and to allow for example better
465 per-node group differentiation, we introduced instance specs, which
468 - minimum instance disk size, disk count, memory size, cpu count
469 - maximum values for the above metrics
470 - and “standard” values (used in ``hspace`` to calculate the standard
473 The minimum/maximum values can be also customised at node-group level,
474 for example allowing more powerful hardware to support bigger instance
477 Beside the instance specs, there are a few other settings belonging to
478 the instance policy framework. It is possible now to customise, per
479 cluster and node-group:
481 - the list of allowed disk templates
482 - the maximum ratio of VCPUs per PCPUs (to control CPU oversubscription)
483 - the maximum ratio of instance to spindles (see below for more
484 information) for local storage
486 All these together should allow all tools that talk to Ganeti to know
487 what are the ranges of allowed values for instances and the
488 over-subscription that is allowed.
490 For the VCPU/PCPU ratio, we already have the VCPU configuration from the
491 instance configuration, and the physical CPU configuration from the
492 node. For the spindle ratios however, we didn't track before these
493 values, so new parameters have been added:
495 - a new node parameter ``spindle_count``, defaults to 1, customisable at
496 node group or node level
497 - at new backend parameter (for instances), ``spindle_use`` defaults to 1
499 Note that spindles in this context doesn't need to mean actual
500 mechanical hard-drives; it's just a relative number for both the node
501 I/O capacity and instance I/O consumption.
503 Instance migration behaviour
504 ++++++++++++++++++++++++++++
506 While live-migration is in general desirable over failover, it is
507 possible that for some workloads it is actually worse, due to the
508 variable time of the “suspend” phase during live migration.
510 To allow the tools to work consistently over such instances (without
511 having to hard-code instance names), a new backend parameter
512 ``always_failover`` has been added to control the migration/failover
513 behaviour. When set to True, all migration requests for an instance will
514 instead fall-back to failover.
516 Instance memory ballooning
517 ++++++++++++++++++++++++++
519 Initial support for memory ballooning has been added. The memory for an
520 instance is no longer fixed (backend parameter ``memory``), but instead
521 can vary between minimum and maximum values (backend parameters
522 ``minmem`` and ``maxmem``). Currently we only change an instance's
525 - live migrating or failing over and instance and the target node
526 doesn't have enough memory
527 - user requests changing the memory via ``gnt-instance modify
533 In order to control the use of specific CPUs by instance, support for
534 controlling CPU pinning has been added for the Xen, HVM and LXC
535 hypervisors. This is controlled by a new hypervisor parameter
536 ``cpu_mask``; details about possible values for this are in the
537 :manpage:`gnt-instance(8)`. Note that use of the most specific (precise
538 VCPU-to-CPU mapping) form will work well only when all nodes in your
539 cluster have the same amount of CPUs.
544 Another area in which Ganeti was not customisable were the parameters
545 used for storage configuration, e.g. how many stripes to use for LVM,
546 DRBD resync configuration, etc.
548 To improve this area, we've added disks parameters, which are
549 customisable at cluster and node group level, and which allow to
550 specify various parameters for disks (DRBD has the most parameters
551 currently), for example:
553 - DRBD resync algorithm and parameters (e.g. speed)
554 - the default VG for meta-data volumes for DRBD
555 - number of stripes for LVM (plain disk template)
558 These parameters can be modified via ``gnt-cluster modify -D …`` and
559 ``gnt-group modify -D …``, and are used at either instance creation (in
560 case of LVM stripes, for example) or at disk “activation” time
563 Rados block device support
564 ++++++++++++++++++++++++++
566 A Rados (http://ceph.com/wiki/Rbd) storage backend has been added,
567 denoted by the ``rbd`` disk template type. This is considered
568 experimental, feedback is welcome. For details on configuring it, see
569 the :doc:`install` document and the :manpage:`gnt-cluster(8)` man page.
574 The existing master IP functionality works well only in simple setups (a
575 single network shared by all nodes); however, if nodes belong to
576 different networks, then the ``/32`` setup and lack of routing
577 information is not enough.
579 To allow the master IP to function well in more complex cases, the
580 system was reworked as follows:
582 - a master IP netmask setting has been added
583 - the master IP activation/turn-down code was moved from the node daemon
585 - whether to run the Ganeti-supplied master IP script or a user-supplied
586 on is a ``gnt-cluster init`` setting
588 Details about the location of the standard and custom setup scripts are
589 in the man page :manpage:`gnt-cluster(8)`; for information about the
590 setup script protocol, look at the Ganeti-supplied script.
595 The `SPICE <http://www.linux-kvm.org/page/SPICE>`_ support has been
598 It is now possible to use TLS-protected connections, and when renewing
599 or changing the cluster certificates (via ``gnt-cluster renew-crypto``,
600 it is now possible to specify spice or spice CA certificates. Also, it
601 is possible to configure a password for SPICE sessions via the
602 hypervisor parameter ``spice_password_file``.
604 There are also new parameters to control the compression and streaming
605 options (e.g. ``spice_image_compression``, ``spice_streaming_video``,
606 etc.). For details, see the man page :manpage:`gnt-instance(8)` and look
607 for the spice parameters.
609 Lastly, it is now possible to see the SPICE connection information via
610 ``gnt-instance console``.
615 A new tool (``tools/ovfconverter``) has been added that supports
616 conversion between Ganeti and the `Open Virtualization Format
617 <http://en.wikipedia.org/wiki/Open_Virtualization_Format>`_ (both to and
620 This relies on the ``qemu-img`` tool to convert the disk formats, so the
621 actual compatibility with other virtualization solutions depends on it.
626 The configuration query daemon (``ganeti-confd``) is now optional, and
627 has been rewritten in Haskell; whether to use the daemon at all, use the
628 Python (default) or the Haskell version is selectable at configure time
629 via the ``--enable-confd`` parameter, which can take one of the
630 ``haskell``, ``python`` or ``no`` values. If not used, disabling the
631 daemon will result in a smaller footprint; for larger systems, we
632 welcome feedback on the Haskell version which might become the default
635 If you want to use ``gnt-node list-drbd`` you need to have the Haskell
636 daemon running. The Python version doesn't implement the new call.
639 User interface changes
640 ~~~~~~~~~~~~~~~~~~~~~~
642 We have replaced the ``--disks`` option of ``gnt-instance
643 replace-disks`` with a more flexible ``--disk`` option, which allows
644 adding and removing disks at arbitrary indices (Issue 188). Furthermore,
645 disk size and mode can be changed upon recreation (via ``gnt-instance
646 recreate-disks``, which accepts the same ``--disk`` option).
648 As many people are used to a ``show`` command, we have added that as an
649 alias to ``info`` on all ``gnt-*`` commands.
651 The ``gnt-instance grow-disk`` command has a new mode in which it can
652 accept the target size of the disk, instead of the delta; this can be
653 more safe since two runs in absolute mode will be idempotent, and
654 sometimes it's also easier to specify the desired size directly.
656 Also the handling of instances with regard to offline secondaries has
657 been improved. Instance operations should not fail because one of it's
658 secondary nodes is offline, even though it's safe to proceed.
660 A new command ``list-drbd`` has been added to the ``gnt-node`` script to
661 support debugging of DRBD issues on nodes. It provides a mapping of DRBD
662 minors to instance name.
667 RAPI coverage has improved, with (for example) new resources for
668 recreate-disks, node power-cycle, etc.
673 There is partial support for ``xl`` in the Xen hypervisor; feedback is
676 Python 2.7 is better supported, and after Ganeti 2.6 we will investigate
677 whether to still support Python 2.4 or move to Python 2.6 as minimum
680 Support for Fedora has been slightly improved; the provided example
681 init.d script should work better on it and the INSTALL file should
682 document the needed dependencies.
687 The deprecated ``QueryLocks`` LUXI request has been removed. Use
688 ``Query(what=QR_LOCK, ...)`` instead.
690 The LUXI requests :pyeval:`luxi.REQ_QUERY_JOBS`,
691 :pyeval:`luxi.REQ_QUERY_INSTANCES`, :pyeval:`luxi.REQ_QUERY_NODES`,
692 :pyeval:`luxi.REQ_QUERY_GROUPS`, :pyeval:`luxi.REQ_QUERY_EXPORTS` and
693 :pyeval:`luxi.REQ_QUERY_TAGS` are deprecated and will be removed in a
694 future version. :pyeval:`luxi.REQ_QUERY` should be used instead.
696 RAPI client: ``CertificateError`` now derives from
697 ``GanetiApiError``. This should make it more easy to handle Ganeti
700 Deprecation warnings due to PyCrypto/paramiko import in
701 ``tools/setup-ssh`` have been silenced, as usually they are safe; please
702 make sure to run an up-to-date paramiko version, if you use this tool.
704 The QA scripts now depend on Python 2.5 or above (the main code base
705 still works with Python 2.4).
707 The configuration file (``config.data``) is now written without
708 indentation for performance reasons; if you want to edit it, it can be
709 re-formatted via ``tools/fmtjson``.
711 A number of bugs has been fixed in the cluster merge tool.
713 ``x509`` certification verification (used in import-export) has been
714 changed to allow the same clock skew as permitted by the cluster
715 verification. This will remove some rare but hard to diagnose errors in
722 *(Released Thu, 19 Jul 2012)*
724 Very few changes from rc4 to the final release, only bugfixes:
726 - integrated fixes from release 2.5.2 (fix general boot flag for KVM
727 instance, fix CDROM booting for KVM instances)
728 - fixed node group modification of node parameters
729 - fixed issue in LUClusterVerifyGroup with multi-group clusters
730 - fixed generation of bash completion to ensure a stable ordering
737 *(Released Fri, 13 Jul 2012)*
739 Third release candidate for 2.6. The following changes were done from
742 - Fixed ``UpgradeConfig`` w.r.t. to disk parameters on disk objects.
743 - Fixed an inconsistency in the LUXI protocol with the provided
744 arguments (NOT backwards compatible)
745 - Fixed a bug with node groups ipolicy where ``min`` was greater than
746 the cluster ``std`` value
747 - Implemented a new ``gnt-node list-drbd`` call to list DRBD minors for
748 easier instance debugging on nodes (requires ``hconfd`` to work)
754 *(Released Tue, 03 Jul 2012)*
756 Second release candidate for 2.6. The following changes were done from
759 - Fixed ``gnt-cluster verify`` regarding ``master-ip-script`` on non
761 - Fixed a RAPI regression on missing beparams/memory
762 - Fixed redistribution of files on offline nodes
763 - Added possibility to run activate-disks even though secondaries are
764 offline. With this change it relaxes also the strictness on some other
765 commands which use activate disks internally:
766 * ``gnt-instance start|reboot|rename|backup|export``
767 - Made it possible to remove safely an instance if its secondaries are
769 - Made it possible to reinstall even though secondaries are offline
775 *(Released Mon, 25 Jun 2012)*
777 First release candidate for 2.6. The following changes were done from
780 - Fixed bugs with disk parameters and ``rbd`` templates as well as
782 - Made ``gnt-instance modify`` more consistent regarding new NIC/Disk
783 behaviour. It supports now the modify operation
784 - ``hcheck`` implemented to analyze cluster health and possibility of
785 improving health by rebalance
786 - ``hbal`` has been improved in dealing with split instances
792 *(Released Mon, 11 Jun 2012)*
794 Second beta release of 2.6. The following changes were done from beta2
797 - Fixed ``daemon-util`` with non-root user models
798 - Fixed creation of plain instances with ``--no-wait-for-sync``
799 - Fix wrong iv_names when running ``cfgupgrade``
800 - Export more information in RAPI group queries
801 - Fixed bug when changing instance network interfaces
802 - Extended burnin to do NIC changes
803 - query: Added ``<``, ``>``, ``<=``, ``>=`` comparison operators
804 - Changed default for DRBD barriers
805 - Fixed DRBD error reporting for syncer rate
806 - Verify the options on disk parameters
808 And of course various fixes to documentation and improved unittests and
815 *(Released Wed, 23 May 2012)*
817 First beta release of 2.6. The following changes were done from beta1 to
820 - integrated patch for distributions without ``start-stop-daemon``
821 - adapted example init.d script to work on Fedora
822 - fixed log handling in Haskell daemons
823 - adapted checks in the watcher for pycurl linked against libnss
824 - add partial support for ``xl`` instead of ``xm`` for Xen
825 - fixed a type issue in cluster verification
826 - fixed ssconf handling in the Haskell code (was breaking confd in IPv6
829 Plus integrated fixes from the 2.5 branch:
831 - fixed ``kvm-ifup`` to use ``/bin/bash``
832 - fixed parallel build failures
833 - KVM live migration when using a custom keymap
839 *(Released Tue, 24 Jul 2012)*
841 A small bugfix release, with no new features:
843 - fixed bash-isms in kvm-ifup, for compatibility with systems which use a
844 different default shell (e.g. Debian, Ubuntu)
845 - fixed KVM startup and live migration with a custom keymap (fixes Issue
846 243 and Debian bug #650664)
847 - fixed compatibility with KVM versions that don't support multiple boot
848 devices (fixes Issue 230 and Debian bug #624256)
850 Additionally, a few fixes were done to the build system (fixed parallel
851 build failures) and to the unittests (fixed race condition in test for
852 FileID functions, and the default enable/disable mode for QA test is now
859 *(Released Fri, 11 May 2012)*
861 A small bugfix release.
863 The main issues solved are on the topic of compatibility with newer LVM
866 - fixed parsing of ``lv_attr`` field
867 - adapted to new ``vgreduce --removemissing`` behaviour where sometimes
868 the ``--force`` flag is needed
870 Also on the topic of compatibility, ``tools/lvmstrap`` has been changed
871 to accept kernel 3.x too (was hardcoded to 2.6.*).
873 A regression present in 2.5.0 that broke handling (in the gnt-* scripts)
874 of hook results and that also made display of other errors suboptimal
875 was fixed; the code behaves now like 2.4 and earlier.
877 Another change in 2.5, the cleanup of the OS scripts environment, is too
878 aggressive: it removed even the ``PATH`` variable, which requires the OS
879 scripts to *always* need to export it. Since this is a bit too strict,
880 we now export a minimal PATH, the same that we export for hooks.
882 The fix for issue 201 (Preserve bridge MTU in KVM ifup script) was
883 integrated into this release.
885 Finally, a few other miscellaneous changes were done (no new features,
886 just small improvements):
888 - Fix ``gnt-group --help`` display
889 - Fix hardcoded Xen kernel path
890 - Fix grow-disk handling of invalid units
891 - Update synopsis for ``gnt-cluster repair-disk-sizes``
892 - Accept both PUT and POST in noded (makes future upgrade to 2.6 easier)
898 *(Released Thu, 12 Apr 2012)*
900 Incompatible/important changes and bugfixes
901 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
903 - The default of the ``/2/instances/[instance_name]/rename`` RAPI
904 resource's ``ip_check`` parameter changed from ``True`` to ``False``
905 to match the underlying LUXI interface.
906 - The ``/2/nodes/[node_name]/evacuate`` RAPI resource was changed to use
907 body parameters, see :doc:`RAPI documentation <rapi>`. The server does
908 not maintain backwards-compatibility as the underlying operation
909 changed in an incompatible way. The RAPI client can talk to old
910 servers, but it needs to be told so as the return value changed.
911 - When creating file-based instances via RAPI, the ``file_driver``
912 parameter no longer defaults to ``loop`` and must be specified.
913 - The deprecated ``bridge`` NIC parameter is no longer supported. Use
915 - Support for the undocumented and deprecated RAPI instance creation
916 request format version 0 has been dropped. Use version 1, supported
917 since Ganeti 2.1.3 and :doc:`documented <rapi>`, instead.
918 - Pyparsing 1.4.6 or above is required, see :doc:`installation
919 documentation <install>`.
920 - The "cluster-verify" hooks are now executed per group by the
921 ``OP_CLUSTER_VERIFY_GROUP`` opcode. This maintains the same behavior
922 if you just run ``gnt-cluster verify``, which generates one opcode per
924 - The environment as passed to the OS scripts is cleared, and thus no
925 environment variables defined in the node daemon's environment will be
926 inherited by the scripts.
927 - The :doc:`iallocator <iallocator>` mode ``multi-evacuate`` has been
929 - :doc:`New iallocator modes <design-multi-reloc>` have been added to
930 support operations involving multiple node groups.
931 - Offline nodes are ignored when failing over an instance.
932 - Support for KVM version 1.0, which changed the version reporting format
934 - TCP/IP ports used by DRBD disks are returned to a pool upon instance
936 - ``Makefile`` is now compatible with Automake 1.11.2
937 - Includes all bugfixes made in the 2.4 series
942 - The ganeti-htools project has been merged into the ganeti-core source
943 tree and will be built as part of Ganeti (see :doc:`install-quick`).
944 - Implemented support for :doc:`shared storage <design-shared-storage>`.
945 - Add support for disks larger than 2 TB in ``lvmstrap`` by supporting
946 GPT-style partition tables (requires `parted
947 <http://www.gnu.org/s/parted/>`_).
948 - Added support for floppy drive and 2nd CD-ROM drive in KVM hypervisor.
949 - Allowed adding tags on instance creation.
950 - Export instance tags to hooks (``INSTANCE_TAGS``, see :doc:`hooks`)
951 - Allow instances to be started in a paused state, enabling the user to
952 see the complete console output on boot using the console.
953 - Added new hypervisor flag to control default reboot behaviour
954 (``reboot_behavior``).
955 - Added support for KVM keymaps (hypervisor parameter ``keymap``).
956 - Improved out-of-band management support:
958 - Added ``gnt-node health`` command reporting the health status of
960 - Added ``gnt-node power`` command to manage power status of nodes.
961 - Added command for emergency power-off (EPO), ``gnt-cluster epo``.
963 - Instance migration can fall back to failover if instance is not
965 - Filters can be used when listing nodes, instances, groups and locks;
966 see :manpage:`ganeti(7)` manpage.
967 - Added post-execution status as variables to :doc:`hooks <hooks>`
969 - Instance tags are exported/imported together with the instance.
970 - When given an explicit job ID, ``gnt-job info`` will work for archived
972 - Jobs can define dependencies on other jobs (not yet supported via
973 RAPI or command line, but used by internal commands and usable via
976 - Lock monitor (``gnt-debug locks``) shows jobs waiting for
979 - Instance failover is now available as a RAPI resource
980 (``/2/instances/[instance_name]/failover``).
981 - ``gnt-instance info`` defaults to static information if primary node
983 - Opcodes have a new ``comment`` attribute.
984 - Added basic SPICE support to KVM hypervisor.
985 - ``tools/ganeti-listrunner`` allows passing of arguments to executable.
987 Node group improvements
988 ~~~~~~~~~~~~~~~~~~~~~~~
990 - ``gnt-cluster verify`` has been modified to check groups separately,
991 thereby improving performance.
992 - Node group support has been added to ``gnt-cluster verify-disks``,
993 which now operates per node group.
994 - Watcher has been changed to work better with node groups.
996 - One process and state file per node group.
997 - Slow watcher in one group doesn't block other group's watcher.
999 - Added new command, ``gnt-group evacuate``, to move all instances in a
1000 node group to other groups.
1001 - Added ``gnt-instance change-group`` to move an instance to another
1003 - ``gnt-cluster command`` and ``gnt-cluster copyfile`` now support
1004 per-group operations.
1005 - Node groups can be tagged.
1006 - Some operations switch from an exclusive to a shared lock as soon as
1008 - Instance's primary and secondary nodes' groups are now available as
1009 query fields (``pnode.group``, ``pnode.group.uuid``, ``snodes.group``
1010 and ``snodes.group.uuid``).
1015 - Numerous updates to documentation and manpages.
1017 - :doc:`RAPI <rapi>` documentation now has detailed parameter
1019 - Some opcode/job results are now also documented, see :doc:`RAPI
1022 - A lockset's internal lock is now also visible in lock monitor.
1023 - Log messages from job queue workers now contain information about the
1024 opcode they're processing.
1025 - ``gnt-instance console`` no longer requires the instance lock.
1026 - A short delay when waiting for job changes reduces the number of LUXI
1027 requests significantly.
1028 - DRBD metadata volumes are overwritten with zeros during disk creation.
1029 - Out-of-band commands no longer acquire the cluster lock in exclusive
1031 - ``devel/upload`` now uses correct permissions for directories.
1037 *(Released Fri, 23 Mar 2012)*
1039 This was the sixth release candidate of the 2.5 series.
1045 *(Released Mon, 9 Jan 2012)*
1047 This was the fifth release candidate of the 2.5 series.
1053 *(Released Thu, 27 Oct 2011)*
1055 This was the fourth release candidate of the 2.5 series.
1061 *(Released Wed, 26 Oct 2011)*
1063 This was the third release candidate of the 2.5 series.
1069 *(Released Tue, 18 Oct 2011)*
1071 This was the second release candidate of the 2.5 series.
1077 *(Released Tue, 4 Oct 2011)*
1079 This was the first release candidate of the 2.5 series.
1085 *(Released Wed, 31 Aug 2011)*
1087 This was the third beta release of the 2.5 series.
1093 *(Released Mon, 22 Aug 2011)*
1095 This was the second beta release of the 2.5 series.
1101 *(Released Fri, 12 Aug 2011)*
1103 This was the first beta release of the 2.5 series.
1109 *(Released Thu, 27 Oct 2011)*
1111 - Fixed bug when parsing command line parameter values ending in
1113 - Fixed assertion error after unclean master shutdown
1114 - Disable HTTP client pool for RPC, significantly reducing memory usage
1116 - Fixed queue archive creation with wrong permissions
1122 *(Released Tue, 23 Aug 2011)*
1126 - Fixed documentation for importing with ``--src-dir`` option
1127 - Fixed a bug in ``ensure-dirs`` with queue/archive permissions
1128 - Fixed a parsing issue with DRBD 8.3.11 in the Linux kernel
1134 *(Released Fri, 5 Aug 2011)*
1136 Many bug-fixes and a few small features:
1138 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
1139 issues when you tried to add an instance with two MAC addresses in one
1141 - KVM: fixed per-instance stored UID value
1142 - KVM: configure bridged NICs at migration start
1143 - KVM: Fix a bug where instance will not start with never KVM versions
1145 - Added OS search path to ``gnt-cluster info``
1146 - Fixed an issue with ``file_storage_dir`` where you were forced to
1147 provide an absolute path, but the documentation states it is a
1148 relative path, the documentation was right
1149 - Added a new parameter to instance stop/start called ``--no-remember``
1150 that will make the state change to not be remembered
1151 - Implemented ``no_remember`` at RAPI level
1152 - Improved the documentation
1153 - Node evacuation: don't call IAllocator if node is already empty
1154 - Fixed bug in DRBD8 replace disks on current nodes
1155 - Fixed bug in recreate-disks for DRBD instances
1156 - Moved assertion checking locks in ``gnt-instance replace-disks``
1157 causing it to abort with not owning the right locks for some situation
1158 - Job queue: Fixed potential race condition when cancelling queued jobs
1159 - Fixed off-by-one bug in job serial generation
1160 - ``gnt-node volumes``: Fix instance names
1161 - Fixed aliases in bash completion
1162 - Fixed a bug in reopening log files after being sent a SIGHUP
1163 - Added a flag to burnin to allow specifying VCPU count
1164 - Bugfixes to non-root Ganeti configuration
1170 *(Released Thu, 12 May 2011)*
1172 Many bug-fixes and a few new small features:
1174 - Fixed a bug related to log opening failures
1175 - Fixed a bug in instance listing with orphan instances
1176 - Fixed a bug which prevented resetting the cluster-level node parameter
1177 ``oob_program`` to the default
1178 - Many fixes related to the ``cluster-merge`` tool
1179 - Fixed a race condition in the lock monitor, which caused failures
1180 during (at least) creation of many instances in parallel
1181 - Improved output for gnt-job info
1182 - Removed the quiet flag on some ssh calls which prevented debugging
1184 - Improved the N+1 failure messages in cluster verify by actually
1185 showing the memory values (needed and available)
1186 - Increased lock attempt timeouts so that when executing long operations
1187 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
1188 too early and thus prevent the use of the 'fair' mechanism
1189 - Changed instance query data (``gnt-instance info``) to not acquire
1190 locks unless needed, thus allowing its use on locked instance if only
1191 static information is asked for
1192 - Improved behaviour with filesystems that do not support rename on an
1194 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
1195 kept locks on all nodes during the wipe, which is unneeded
1196 - Fixed ``gnt-watcher`` handling of errors during hooks execution
1197 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
1198 10GiB) which caused the wipe to fail right at the end in some cases
1199 - Fixed master IP activation when doing master failover with no-voting
1200 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
1201 the master node itself
1202 - Fixed potential data-loss in under disk full conditions, where Ganeti
1203 wouldn't check correctly the return code and would consider
1204 partially-written files 'correct'
1205 - Fixed bug related to multiple VGs and DRBD disk replacing
1206 - Added new disk parameter ``metavg`` that allows placement of the meta
1207 device for DRBD in a different volume group
1208 - Fixed error handling in the node daemon when the system libc doesn't
1209 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
1210 - Fixed lock release during replace-disks, which kept cluster-wide locks
1211 when doing disk replaces with an iallocator script
1212 - Added check for missing bridges in cluster verify
1213 - Handle EPIPE errors while writing to the terminal better, so that
1214 piping the output to e.g. ``less`` doesn't cause a backtrace
1215 - Fixed rare case where a ^C during Luxi calls could have been
1216 interpreted as server errors, instead of simply terminating
1217 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
1219 - Added a few more parameters to the KVM hypervisor, allowing a second
1220 CDROM, custom disk type for CDROMs and a floppy image
1221 - Removed redundant message in instance rename when the name is given
1223 - Added option to ``gnt-instance recreate-disks`` to allow creating the
1224 disks on new nodes, allowing recreation when the original instance
1225 nodes are completely gone
1226 - Added option when converting disk templates to DRBD to skip waiting
1227 for the resync, in order to make the instance available sooner
1228 - Added two new variables to the OS scripts environment (containing the
1230 - Made the root_path and optional parameter for the xen-pvm hypervisor,
1231 to allow use of ``pvgrub`` as bootloader
1232 - Changed the instance memory modifications to only check out-of-memory
1233 conditions on memory increases, and turned the secondary node warnings
1234 into errors (they can still be overridden via ``--force``)
1235 - Fixed the handling of a corner case when the Python installation gets
1236 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
1237 to execute a command that doesn't exist
1238 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
1239 node of the instance returned failures during instance shutdown; this
1240 adds the option ``--ignore-consistency`` to gnt-instance move
1242 And as usual, various improvements to the error messages, documentation
1249 *(Released Wed, 09 Mar 2011)*
1251 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
1252 the RAPI users file if run twice (even with ``--dry-run``).
1254 The release fixes that bug (nothing else changed).
1260 *(Released Mon, 07 Mar 2011)*
1262 Final 2.4.0 release. Just a few small fixes:
1264 - Fixed RAPI node evacuate
1265 - Fixed the kvm-ifup script
1266 - Fixed internal error handling for special job cases
1267 - Updated man page to specify the escaping feature for options
1273 *(Released Mon, 28 Feb 2011)*
1275 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
1276 that this feature wiped the disks of the wrong instance, leading to loss
1281 - Fixed title of query field containing instance name
1282 - Expanded the glossary in the documentation
1283 - Fixed one unittest (internal issue)
1289 *(Released Mon, 21 Feb 2011)*
1291 A number of bug fixes plus just a couple functionality changes.
1293 On the user-visible side, the ``gnt-* list`` command output has changed
1294 with respect to "special" field states. The current rc1 style of display
1295 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
1296 the default output mode special fields states are displayed as follows:
1298 - Offline resource: ``*``
1299 - Unavailable/not applicable: ``-``
1300 - Data missing (RPC failure): ``?``
1301 - Unknown field: ``??``
1303 Another user-visible change is the addition of ``--force-join`` to
1308 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
1309 - Fixed regression in RAPI/instance reinstall where all parameters were
1310 required (instead of optional)
1311 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
1312 - Fixed iallocator usage (offline nodes were not considered offline)
1313 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
1314 - Fixed hypervisor and OS parameter validation with respect to
1315 non-vm_capable nodes
1316 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
1318 - Fixed ``tools/listrunner`` with respect to agent-based usage
1324 *(Released Fri, 4 Feb 2011)*
1326 Many changes and fixes since the beta1 release. While there were some
1327 internal changes, the code has been mostly stabilised for the RC
1330 Note: the dumb allocator was removed in this release, as it was not kept
1331 up-to-date with the IAllocator protocol changes. It is recommended to
1332 use the ``hail`` command from the ganeti-htools package.
1334 Note: the 2.4 and up versions of Ganeti are not compatible with the
1335 0.2.x branch of ganeti-htools. You need to upgrade to
1336 ganeti-htools-0.3.0 (or later).
1338 Regressions fixed from 2.3
1339 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1341 - Fixed the ``gnt-cluster verify-disks`` command
1342 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
1344 - Fixed disk adoption breakage
1345 - Fixed wrong headers in instance listing for field aliases
1350 - Fixed corner case in KVM handling of NICs
1351 - Fixed many cases of wrong handling of non-vm_capable nodes
1352 - Fixed a bug where a missing instance symlink was not possible to
1353 recreate with any ``gnt-*`` command (now ``gnt-instance
1354 activate-disks`` does it)
1355 - Fixed the volume group name as reported by ``gnt-cluster
1357 - Increased timeouts for the import-export code, hopefully leading to
1358 fewer aborts due network or instance timeouts
1359 - Fixed bug in ``gnt-node list-storage``
1360 - Fixed bug where not all daemons were started on cluster
1361 initialisation, but only at the first watcher run
1362 - Fixed many bugs in the OOB implementation
1363 - Fixed watcher behaviour in presence of instances with offline
1365 - Fixed instance list output for instances running on the wrong node
1366 - a few fixes to the cluster-merge tool, but it still cannot merge
1367 multi-node groups (currently it is not recommended to use this tool)
1373 - Improved network configuration for the KVM hypervisor
1374 - Added e1000 as a supported NIC for Xen-HVM
1375 - Improved the lvmstrap tool to also be able to use partitions, as
1376 opposed to full disks
1377 - Improved speed of disk wiping (the cluster parameter
1378 ``prealloc_wipe_disks``, so that it has a low impact on the total time
1379 of instance creations
1380 - Added documentation for the OS parameters
1381 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
1382 hypervisor is not responding
1383 - Added display of blacklisted and hidden OS information in
1384 ``gnt-cluster info``
1385 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
1386 NIC and node parameters, which might create problems with currently
1387 invalid (but undetected) configuration files, but prevents validation
1388 failures when unrelated parameters are modified
1389 - Changed cluster initialisation to wait for the master daemon to become
1391 - Expanded the RAPI interface:
1393 - Added config redistribution resource
1394 - Added activation/deactivation of instance disks
1395 - Added export of console information
1397 - Implemented log file reopening on SIGHUP, which allows using
1398 logrotate(8) for the Ganeti log files
1399 - Added a basic OOB helper script as an example
1405 *(Released Fri, 14 Jan 2011)*
1410 - Fixed timezone issues when formatting timestamps
1411 - Added support for node groups, available via ``gnt-group`` and other
1413 - Added out-of-band framework and management, see :doc:`design
1414 document <design-oob>`
1415 - Removed support for roman numbers from ``gnt-node list`` and
1416 ``gnt-instance list``.
1417 - Allowed modification of master network interface via ``gnt-cluster
1418 modify --master-netdev``
1419 - Accept offline secondaries while shutting down instance disks
1420 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
1421 - Added support for multiple LVM volume groups
1422 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
1424 - Added commands to list available fields:
1426 - ``gnt-node list-fields``
1427 - ``gnt-group list-fields``
1428 - ``gnt-instance list-fields``
1430 - Updated documentation and man pages
1435 - Moved ``rapi_users`` file into separate directory, now named
1436 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
1438 - Added new tool for running commands on many machines,
1439 ``tools/ganeti-listrunner``
1440 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
1441 improving the ``gnt-instance console`` output
1442 - Allowed customisation of disk index separator at ``configure`` time
1443 - Export node group allocation policy to :doc:`iallocator <iallocator>`
1444 - Added support for non-partitioned md disks in ``lvmstrap``
1445 - Added script to gracefully power off KVM instances
1446 - Split ``utils`` module into smaller parts
1447 - Changed query operations to return more detailed information, e.g.
1448 whether an information is unavailable due to an offline node. To use
1449 this new functionality, the LUXI call ``Query`` must be used. Field
1450 information is now stored by the master daemon and can be retrieved
1451 using ``QueryFields``. Instances, nodes and groups can also be queried
1452 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
1453 exposed via RAPI). The following commands make use of this
1454 infrastructure change:
1456 - ``gnt-group list``
1457 - ``gnt-group list-fields``
1459 - ``gnt-node list-fields``
1460 - ``gnt-instance list``
1461 - ``gnt-instance list-fields``
1462 - ``gnt-debug locks``
1467 - New RAPI resources (see :doc:`rapi`):
1471 - ``/2/groups/[group_name]``
1472 - ``/2/groups/[group_name]/assign-nodes``
1473 - ``/2/groups/[group_name]/modify``
1474 - ``/2/groups/[group_name]/rename``
1475 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
1479 - Implemented ``no_install`` for instance creation
1480 - Implemented OS parameters for instance reinstallation, allowing
1481 use of special settings on reinstallation (e.g. for preserving data)
1486 - Added IPv6 support in import/export
1487 - Pause DRBD synchronization while wiping disks on instance creation
1488 - Updated unittests and QA scripts
1489 - Improved network parameters passed to KVM
1490 - Converted man pages from docbook to reStructuredText
1496 *(Released Mon, 20 Dec 2010)*
1498 Released version 2.3.1~rc1 without any changes.
1504 *(Released Wed, 1 Dec 2010)*
1506 - impexpd: Disable OpenSSL compression in socat if possible (backport
1507 from master, commit e90739d625b, see :doc:`installation guide
1508 <install-quick>` for details)
1509 - Changed unittest coverage report to exclude test scripts
1510 - Added script to check version format
1516 *(Released Wed, 1 Dec 2010)*
1518 Released version 2.3.0~rc1 without any changes.
1524 *(Released Fri, 19 Nov 2010)*
1526 A number of bugfixes and documentation updates:
1528 - Update ganeti-os-interface documentation
1529 - Fixed a bug related to duplicate MACs or similar items which should be
1531 - Fix breakage in OS state modify
1532 - Reinstall instance: disallow offline secondaries (fixes bug related to
1533 OS changing but reinstall failing)
1534 - plus all the other fixes between 2.2.1 and 2.2.2
1540 *(Released Tue, 2 Nov 2010)*
1542 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
1543 - Fixed master failover race with watcher
1544 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
1546 - Accept previously stopped instance for export with instance removal
1547 - Simplify and extend the environment variables for instance OS scripts
1548 - Added new node flags, ``master_capable`` and ``vm_capable``
1549 - Added optional instance disk wiping prior during allocation. This is a
1550 cluster-wide option and can be set/modified using
1551 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
1552 - Added IPv6 support, see :doc:`design document <design-2.3>` and
1553 :doc:`install-quick`
1554 - Added a new watcher option (``--ignore-pause``)
1555 - Added option to ignore offline node on instance start/stop
1556 (``--ignore-offline``)
1557 - Allow overriding OS parameters with ``gnt-instance reinstall``
1558 - Added ability to change node's secondary IP address using ``gnt-node
1560 - Implemented privilege separation for all daemons except
1561 ``ganeti-noded``, see ``configure`` options
1562 - Complain if an instance's disk is marked faulty in ``gnt-cluster
1564 - Implemented job priorities (see ``ganeti(7)`` manpage)
1565 - Ignore failures while shutting down instances during failover from
1567 - Exit daemon's bootstrap process only once daemon is ready
1568 - Export more information via ``LUInstanceQuery``/remote API
1569 - Improved documentation, QA and unittests
1570 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
1571 restart if the file was created or changed
1572 - Added LUXI protocol version sent with each request and response,
1573 allowing detection of server/client mismatches
1574 - Moved the Python scripts among gnt-* and ganeti-* into modules
1575 - Moved all code related to setting up SSH to an external script,
1577 - Infrastructure changes for node group support in future versions
1583 *(Released Fri, 19 Nov 2010)*
1585 A few small bugs fixed, and some improvements to the build system:
1587 - Fix documentation regarding conversion to drbd
1588 - Fix validation of parameters in cluster modify (``gnt-cluster modify
1590 - Fix error handling in node modify with multiple changes
1591 - Allow remote imports without checked names
1597 *(Released Tue, 19 Oct 2010)*
1599 - Disable SSL session ID cache in RPC client
1605 *(Released Thu, 14 Oct 2010)*
1607 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
1608 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
1610 - Fix problems with interaction between Curl and Python's HTTP server,
1611 resulting in increased speed in many RPC calls
1612 - Improve our release script to prevent breakage with older aclocal and
1619 *(Released Thu, 7 Oct 2010)*
1621 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
1622 value retrieved from master
1623 - Added support for blacklisted or hidden OS definitions
1624 - Added simple lock monitor (accessible via (``gnt-debug locks``)
1625 - Added support for -mem-path in KVM hypervisor abstraction layer
1626 - Allow overriding instance parameters in tool for inter-cluster
1627 instance moves (``tools/move-instance``)
1628 - Improved opcode summaries (e.g. in ``gnt-job list``)
1629 - Improve consistency of OS listing by sorting it
1630 - Documentation updates
1636 *(Released Fri, 8 Oct 2010)*
1638 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
1644 *(Released Mon, 4 Oct 2010)*
1646 - Fixed regression in ``gnt-instance rename``
1652 *(Released Wed, 22 Sep 2010)*
1654 - Fixed OS_VARIANT variable for OS scripts
1655 - Fixed cluster tag operations via RAPI
1656 - Made ``setup-ssh`` exit with non-zero code if an error occurred
1657 - Disabled RAPI CA checks in watcher
1663 *(Released Mon, 23 Aug 2010)*
1665 - Support DRBD versions of the format "a.b.c.d"
1667 - Re-introduce support for usage from multiple threads in RAPI client
1668 - Instance renames and modify via RAPI
1669 - Work around race condition between processing and archival in job
1671 - Mark opcodes following failed one as failed, too
1672 - Job field ``lock_status`` was removed due to difficulties making it
1673 work with the changed job queue in Ganeti 2.2; a better way to monitor
1674 locks is expected for a later 2.2.x release
1675 - Fixed dry-run behaviour with many commands
1676 - Support ``ssh-agent`` again when adding nodes
1677 - Many additional bugfixes
1683 *(Released Fri, 30 Jul 2010)*
1685 Important change: the internal RPC mechanism between Ganeti nodes has
1686 changed from using a home-grown http library (based on the Python base
1687 libraries) to use the PycURL library. This requires that PycURL is
1688 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
1689 against GnuTLS by default. cURL's support for GnuTLS had known issues
1690 before cURL 7.21.0 and we recommend using the latest cURL release or
1691 linking against OpenSSL. Most other distributions already link PycURL
1692 and cURL against OpenSSL. The command::
1694 python -c 'import pycurl; print pycurl.version'
1696 can be used to determine the libraries PycURL and cURL are linked
1699 Other significant changes:
1701 - Rewrote much of the internals of the job queue, in order to achieve
1702 better parallelism; this decouples job query operations from the job
1703 processing, and it should allow much nicer behaviour of the master
1704 daemon under load, and it also has uncovered some long-standing bugs
1705 related to the job serialisation (now fixed)
1706 - Added a default iallocator setting to the cluster parameters,
1707 eliminating the need to always pass nodes or an iallocator for
1708 operations that require selection of new node(s)
1709 - Added experimental support for the LXC virtualization method
1710 - Added support for OS parameters, which allows the installation of
1711 instances to pass parameter to OS scripts in order to customise the
1713 - Added a hypervisor parameter controlling the migration type (live or
1714 non-live), since hypervisors have various levels of reliability; this
1715 has renamed the 'live' parameter to 'mode'
1716 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
1717 logical volumes, meaning that cluster verify will ignore them and not
1718 flag their presence as errors
1719 - The watcher will now reset the error count for failed instances after
1720 8 hours, thus allowing self-healing if the problem that caused the
1721 instances to be down/fail to start has cleared in the meantime
1722 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
1723 check for, and warn, if the drbd module parameter ``usermode_helper``
1724 is not consistent with the cluster-wide setting; this is needed to
1725 make diagnose easier of failed drbd creations
1726 - Started adding base IPv6 support, but this is not yet
1727 enabled/available for use
1728 - Rename operations (cluster, instance) will now return the new name,
1729 which is especially useful if a short name was passed in
1730 - Added support for instance migration in RAPI
1731 - Added a tool to pre-configure nodes for the SSH setup, before joining
1732 them to the cluster; this will allow in the future a simplified model
1733 for node joining (but not yet fully enabled in 2.2); this needs the
1734 paramiko python library
1735 - Fixed handling of name-resolving errors
1736 - Fixed consistency of job results on the error path
1737 - Fixed master-failover race condition when executed multiple times in
1739 - Fixed many bugs related to the job queue (mostly introduced during the
1740 2.2 development cycle, so not all are impacting 2.1)
1741 - Fixed instance migration with missing disk symlinks
1742 - Fixed handling of unknown jobs in ``gnt-job archive``
1743 - And many other small fixes/improvements
1747 - Enhanced both the unittest and the QA coverage
1748 - Switched the opcode validation to a generic model, and extended the
1749 validation to all opcode parameters
1750 - Changed more parts of the code that write shell scripts to use the
1752 - Switched the master daemon to use the asyncore library for the Luxi
1759 *(Released Thu, 17 Jun 2010)*
1761 - Added tool (``move-instance``) and infrastructure to move instances
1762 between separate clusters (see :doc:`separate documentation
1763 <move-instance>` and :doc:`design document <design-2.2>`)
1764 - Added per-request RPC timeout
1765 - RAPI now requires a Content-Type header for requests with a body (e.g.
1766 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
1767 :rfc:`2616` (HTTP/1.1), section 7.2.1)
1768 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
1770 - Implemented initial support for running Ganeti daemons as separate
1771 users, see configure-time flags ``--with-user-prefix`` and
1772 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
1774 - Instances can be removed after export (``gnt-backup export
1775 --remove-instance``)
1776 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
1777 key (instead of 1024 bit)
1778 - Added new cluster configuration file for cluster domain secret
1779 - Import/export now use SSL instead of SSH
1780 - Added support for showing estimated time when exporting an instance,
1781 see the ``ganeti-os-interface(7)`` manpage and look for
1788 *(Released Tue, 16 Nov 2010)*
1790 Some more bugfixes. Unless critical bugs occur, this will be the last
1793 - Fix case of MAC special-values
1794 - Fix mac checker regex
1795 - backend: Fix typo causing "out of range" error
1796 - Add missing --units in gnt-instance list man page
1802 *(Released Tue, 24 Aug 2010)*
1805 - Don't ignore secondary node silently on non-mirrored disk templates
1807 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
1808 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
1809 - Properly document the usb_mouse parameter
1810 - Fix path in ganeti-rapi(8) (issue 116)
1811 - Adjust error message when the ganeti user's .ssh directory is
1813 - Add same-node-check when changing the disk template to drbd
1819 *(Released Fri, 16 Jul 2010)*
1822 - Add an option to only select some reboot types during qa/burnin.
1823 (on some hypervisors consequent reboots are not supported)
1824 - Fix infrequent race condition in master failover. Sometimes the old
1825 master ip address would be still detected as up for a short time
1826 after it was removed, causing failover to fail.
1827 - Decrease mlockall warnings when the ctypes module is missing. On
1828 Python 2.4 we support running even if no ctypes module is installed,
1829 but we were too verbose about this issue.
1830 - Fix building on old distributions, on which man doesn't have a
1832 - Fix RAPI not to ignore the MAC address on instance creation
1833 - Implement the old instance creation format in the RAPI client.
1839 *(Released Thu, 01 Jul 2010)*
1841 A small bugfix release:
1842 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
1843 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
1844 a non-existing option
1845 - Fix instance create: the --force-variant option was ignored
1846 - Improve pylint 0.21 compatibility and warnings with Python 2.6
1847 - Fix modify node storage with non-FQDN arguments
1848 - Fix RAPI client to authenticate under Python 2.6 when used
1849 for more than 5 requests needing authentication
1850 - Fix gnt-instance modify -t (storage) giving a wrong error message
1851 when converting a non-shutdown drbd instance to plain
1857 *(Released Fri, 18 Jun 2010)*
1859 A small bugfix release:
1861 - Fix live migration of KVM instances started with older Ganeti
1862 versions which had fewer hypervisor parameters
1863 - Fix gnt-instance grow-disk on down instances
1864 - Fix an error-reporting bug during instance migration
1865 - Better checking of the ``--net`` and ``--disk`` values, to avoid
1866 silently ignoring broken ones
1867 - Fix an RPC error reporting bug affecting, for example, RAPI client
1869 - Fix bug triggered by different API version os-es on different nodes
1870 - Fix a bug in instance startup with custom hvparams: OS level
1871 parameters would fail to be applied.
1872 - Fix the RAPI client under Python 2.6 (but more work is needed to
1873 make it work completely well with OpenSSL)
1874 - Fix handling of errors when resolving names from DNS
1880 *(Released Thu, 3 Jun 2010)*
1882 A medium sized development cycle. Some new features, and some
1883 fixes/small improvements/cleanups.
1885 Significant features
1886 ~~~~~~~~~~~~~~~~~~~~
1888 The node deamon now tries to mlock itself into memory, unless the
1889 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
1890 its logs, and falls back to console logging. This allows emergency
1891 features such as ``gnt-node powercycle`` to work even in the event of a
1892 broken node disk (tested offlining the disk hosting the node's
1893 filesystem and dropping its memory caches; don't try this at home)
1895 KVM: add vhost-net acceleration support. It can be tested with a new
1896 enough version of the kernel and of qemu-kvm.
1898 KVM: Add instance chrooting feature. If you use privilege dropping for
1899 your VMs you can also now force them to chroot to an empty directory,
1900 before starting the emulated guest.
1902 KVM: Add maximum migration bandwith and maximum downtime tweaking
1903 support (requires a new-enough version of qemu-kvm).
1905 Cluster verify will now warn if the master node doesn't have the master
1906 ip configured on it.
1908 Add a new (incompatible) instance creation request format to RAPI which
1909 supports all parameters (previously only a subset was supported, and it
1910 wasn't possible to extend the old format to accomodate all the new
1911 features. The old format is still supported, and a client can check for
1912 this feature, before using it, by checking for its presence in the
1913 ``features`` RAPI resource.
1915 Now with ancient latin support. Try it passing the ``--roman`` option to
1916 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
1917 (requires the python-roman module to be installed, in order to work).
1922 As usual many internal code refactorings, documentation updates, and
1925 - Lots of improvements and cleanups to the experimental Remote API
1926 (RAPI) client library.
1927 - A new unit test suite for the core daemon libraries.
1928 - A fix to creating missing directories makes sure the umask is not
1929 applied anymore. This enforces the same directory permissions
1931 - Better handling terminating daemons with ctrl+c (used when running
1932 them in debugging mode).
1933 - Fix a race condition in live migrating a KVM instance, when stat()
1934 on the old proc status file returned EINVAL, which is an unexpected
1936 - Fixed manpage checking with newer man and utf-8 charachters. But now
1937 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
1943 *(Released Fri, 7 May 2010)*
1945 Fix a bug which prevented untagged KVM instances from starting.
1951 *(Released Fri, 7 May 2010)*
1953 Another release with a long development cycle, during which many
1954 different features were added.
1956 Significant features
1957 ~~~~~~~~~~~~~~~~~~~~
1959 The KVM hypervisor now can run the individual instances as non-root, to
1960 reduce the impact of a VM being hijacked due to bugs in the
1961 hypervisor. It is possible to run all instances as a single (non-root)
1962 user, to manually specify a user for each instance, or to dynamically
1963 allocate a user out of a cluster-wide pool to each instance, with the
1964 guarantee that no two instances will run under the same user ID on any
1967 An experimental RAPI client library, that can be used standalone
1968 (without the other Ganeti libraries), is provided in the source tree as
1969 ``lib/rapi/client.py``. Note this client might change its interface in
1970 the future, as we iterate on its capabilities.
1972 A new command, ``gnt-cluster renew-crypto`` has been added to easily
1973 replace the cluster's certificates and crypto keys. This might help in
1974 case they have been compromised, or have simply expired.
1976 A new disk option for instance creation has been added that allows one
1977 to "adopt" currently existing logical volumes, with data
1978 preservation. This should allow easier migration to Ganeti from
1979 unmanaged (or managed via other software) instances.
1981 Another disk improvement is the possibility to convert between redundant
1982 (DRBD) and plain (LVM) disk configuration for an instance. This should
1983 allow better scalability (starting with one node and growing the
1984 cluster, or shrinking a two-node cluster to one node).
1986 A new feature that could help with automated node failovers has been
1987 implemented: if a node sees itself as offline (by querying the master
1988 candidates), it will try to shutdown (hard) all instances and any active
1989 DRBD devices. This reduces the risk of duplicate instances if an
1990 external script automatically failovers the instances on such nodes. To
1991 enable this, the cluster parameter ``maintain_node_health`` should be
1992 enabled; in the future this option (per the name) will enable other
1993 automatic maintenance features.
1995 Instance export/import now will reuse the original instance
1996 specifications for all parameters; that means exporting an instance,
1997 deleting it and the importing it back should give an almost identical
1998 instance. Note that the default import behaviour has changed from
1999 before, where it created only one NIC; now it recreates the original
2002 Cluster verify has added a few new checks: SSL certificates validity,
2003 /etc/hosts consistency across the cluster, etc.
2008 As usual, many internal changes were done, documentation fixes,
2011 - Fixed cluster initialization with disabled cluster storage (regression
2012 introduced in 2.1.1)
2013 - File-based storage supports growing the disks
2014 - Fixed behaviour of node role changes
2015 - Fixed cluster verify for some corner cases, plus a general rewrite of
2016 cluster verify to allow future extension with more checks
2017 - Fixed log spamming by watcher and node daemon (regression introduced
2019 - Fixed possible validation issues when changing the list of enabled
2021 - Fixed cleanup of /etc/hosts during node removal
2022 - Fixed RAPI response for invalid methods
2023 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
2024 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
2025 from ide disks, etc.)
2026 - Allow OS changes without re-installation (to record a changed OS
2027 outside of Ganeti, or to allow OS renames)
2028 - Allow instance creation without OS installation (useful for example if
2029 the OS will be installed manually, or restored from a backup not in
2031 - Implemented option to make cluster ``copyfile`` use the replication
2033 - Added list of enabled hypervisors to ssconf (possibly useful for
2035 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
2037 - A partial form of node re-IP is possible via node readd, which now
2038 allows changed node primary IP
2039 - Command line utilities now show an informational message if the job is
2041 - The logs of the master daemon now show the PID/UID/GID of the
2048 *(Released Fri, 12 Mar 2010)*
2050 During the 2.1.0 long release candidate cycle, a lot of improvements and
2051 changes have accumulated with were released later as 2.1.1.
2056 The node evacuate command (``gnt-node evacuate``) was significantly
2057 rewritten, and as such the IAllocator protocol was changed - a new
2058 request type has been added. This unfortunate change during a stable
2059 series is designed to improve performance of node evacuations; on
2060 clusters with more than about five nodes and which are well-balanced,
2061 evacuation should proceed in parallel for all instances of the node
2062 being evacuated. As such, any existing IAllocator scripts need to be
2063 updated, otherwise the above command will fail due to the unknown
2064 request. The provided "dumb" allocator has not been updated; but the
2065 ganeti-htools package supports the new protocol since version 0.2.4.
2067 Another important change is increased validation of node and instance
2068 names. This might create problems in special cases, if invalid host
2069 names are being used.
2071 Also, a new layer of hypervisor parameters has been added, that sits at
2072 OS level between the cluster defaults and the instance ones. This allows
2073 customisation of virtualization parameters depending on the installed
2074 OS. For example instances with OS 'X' may have a different KVM kernel
2075 (or any other parameter) than the cluster defaults. This is intended to
2076 help managing a multiple OSes on the same cluster, without manual
2077 modification of each instance's parameters.
2079 A tool for merging clusters, ``cluster-merge``, has been added in the
2080 tools sub-directory.
2085 - Improved the int/float conversions that should make the code more
2086 robust in face of errors from the node daemons
2087 - Fixed the remove node code in case of internal configuration errors
2088 - Fixed the node daemon behaviour in face of inconsistent queue
2089 directory (e.g. read-only file-system where we can't open the files
2091 - Fixed the behaviour of gnt-node modify for master candidate demotion;
2092 now it either aborts cleanly or, if given the new "auto_promote"
2093 parameter, will automatically promote other nodes as needed
2094 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
2095 completely prevent Ganeti from working
2096 - Fixed bug for instance export when not all disks were successfully
2098 - Fixed behaviour of node add when the new node is slow in starting up
2100 - Fixed handling of signals in the LUXI client, which should improve
2101 behaviour of command-line scripts
2102 - Added checks for invalid node/instance names in the configuration (now
2103 flagged during cluster verify)
2104 - Fixed watcher behaviour for disk activation errors
2105 - Fixed two potentially endless loops in http library, which led to the
2106 RAPI daemon hanging and consuming 100% CPU in some cases
2107 - Fixed bug in RAPI daemon related to hashed passwords
2108 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
2110 - Enhanced compatibility with non-Debian OSes, but not using absolute
2111 path in some commands and allowing customisation of the ssh
2112 configuration directory
2113 - Fixed possible future issue with new Python versions by abiding to the
2114 proper use of ``__slots__`` attribute on classes
2115 - Added checks that should prevent directory traversal attacks
2116 - Many documentation fixes based on feedback from users
2121 - Added an "early_release" more for instance replace disks and node
2122 evacuate, where we release locks earlier and thus allow higher
2123 parallelism within the cluster
2124 - Added watcher hooks, intended to allow the watcher to restart other
2125 daemons (e.g. from the ganeti-nbma project), but they can be used of
2126 course for any other purpose
2127 - Added a compile-time disable for DRBD barriers, to increase
2128 performance if the administrator trusts the power supply or the
2129 storage system to not lose writes
2130 - Added the option of using syslog for logging instead of, or in
2131 addition to, Ganeti's own log files
2132 - Removed boot restriction for paravirtual NICs for KVM, recent versions
2133 can indeed boot from a paravirtual NIC
2134 - Added a generic debug level for many operations; while this is not
2135 used widely yet, it allows one to pass the debug value all the way to
2137 - Enhanced the hooks environment for instance moves (failovers,
2138 migrations) where the primary/secondary nodes changed during the
2139 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
2140 - Enhanced data validations for many user-supplied values; one important
2141 item is the restrictions imposed on instance and node names, which
2142 might reject some (invalid) host names
2143 - Add a configure-time option to disable file-based storage, if it's not
2144 needed; this allows greater security separation between the master
2145 node and the other nodes from the point of view of the inter-node RPC
2147 - Added user notification in interactive tools if job is waiting in the
2148 job queue or trying to acquire locks
2149 - Added log messages when a job is waiting for locks
2150 - Added filtering by node tags in instance operations which admit
2151 multiple instances (start, stop, reboot, reinstall)
2152 - Added a new tool for cluster mergers, ``cluster-merge``
2153 - Parameters from command line which are of the form ``a=b,c=d`` can now
2154 use backslash escapes to pass in values which contain commas,
2155 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
2157 - For KVM, the instance name is the first parameter passed to KVM, so
2158 that it's more visible in the process list
2164 *(Released Tue, 2 Mar 2010)*
2166 Ganeti 2.1 brings many improvements with it. Major changes:
2168 - Added infrastructure to ease automated disk repairs
2169 - Added new daemon to export configuration data in a cheaper way than
2170 using the remote API
2171 - Instance NICs can now be routed instead of being associated with a
2173 - Improved job locking logic to reduce impact of jobs acquiring multiple
2174 locks waiting for other long-running jobs
2176 In-depth implementation details can be found in the Ganeti 2.1 design
2182 - Added chroot hypervisor
2183 - Added more options to xen-hvm hypervisor (``kernel_path`` and
2185 - Added more options to xen-pvm hypervisor (``use_bootloader``,
2186 ``bootloader_path`` and ``bootloader_args``)
2187 - Added the ``use_localtime`` option for the xen-hvm and kvm
2188 hypervisors, and the default value for this has changed to false (in
2189 2.0 xen-hvm always enabled it)
2190 - Added luxi call to submit multiple jobs in one go
2191 - Added cluster initialization option to not modify ``/etc/hosts``
2193 - Added network interface parameters
2194 - Added dry run mode to some LUs
2195 - Added RAPI resources:
2197 - ``/2/instances/[instance_name]/info``
2198 - ``/2/instances/[instance_name]/replace-disks``
2199 - ``/2/nodes/[node_name]/evacuate``
2200 - ``/2/nodes/[node_name]/migrate``
2201 - ``/2/nodes/[node_name]/role``
2202 - ``/2/nodes/[node_name]/storage``
2203 - ``/2/nodes/[node_name]/storage/modify``
2204 - ``/2/nodes/[node_name]/storage/repair``
2206 - Added OpCodes to evacuate or migrate all instances on a node
2207 - Added new command to list storage elements on nodes (``gnt-node
2208 list-storage``) and modify them (``gnt-node modify-storage``)
2209 - Added new ssconf files with master candidate IP address
2210 (``ssconf_master_candidates_ips``), node primary IP address
2211 (``ssconf_node_primary_ips``) and node secondary IP address
2212 (``ssconf_node_secondary_ips``)
2213 - Added ``ganeti-confd`` and a client library to query the Ganeti
2214 configuration via UDP
2215 - Added ability to run hooks after cluster initialization and before
2217 - Added automatic mode for disk replace (``gnt-instance replace-disks
2219 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
2220 after catastrophic data-loss
2221 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
2223 - Added ``gnt-instance move`` command to move instances
2224 - Added ``gnt-cluster watcher`` command to control watcher
2225 - Added ``gnt-node powercycle`` command to powercycle nodes
2226 - Added new job status field ``lock_status``
2227 - Added parseable error codes to cluster verification (``gnt-cluster
2228 verify --error-codes``) and made output less verbose (use
2229 ``--verbose`` to restore previous behaviour)
2230 - Added UUIDs to the main config entities (cluster, nodes, instances)
2231 - Added support for OS variants
2232 - Added support for hashed passwords in the Ganeti remote API users file
2234 - Added option to specify maximum timeout on instance shutdown
2235 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
2236 - Added new helper script to start and stop Ganeti daemons
2237 (``daemon-util``), with the intent to reduce the work necessary to
2238 adjust Ganeti for non-Debian distributions and to start/stop daemons
2240 - Added more unittests
2241 - Fixed critical bug in ganeti-masterd startup
2242 - Removed the configure-time ``kvm-migration-port`` parameter, this is
2243 now customisable at the cluster level for both the KVM and Xen
2244 hypervisors using the new ``migration_port`` parameter
2245 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
2246 reinstalling an instance
2247 - Allowed ``@`` in tag names
2248 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
2249 - Many documentation updates
2250 - Distribute hypervisor files on ``gnt-cluster redist-conf``
2251 - ``gnt-instance reinstall`` can now reinstall multiple instances
2252 - Updated many command line parameters
2253 - Introduced new OS API version 15
2254 - No longer support a default hypervisor
2255 - Treat virtual LVs as inexistent
2256 - Improved job locking logic to reduce lock contention
2257 - Match instance and node names case insensitively
2258 - Reimplemented bash completion script to be more complete
2265 *(Released Thu, 4 Feb 2010)*
2267 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
2268 - Fix a string formatting bug
2269 - Improve safety of the code in some error paths
2270 - Improve data validation in the master of values returned from nodes
2276 *(Released Thu, 17 Dec 2009)*
2278 - Fix security issue due to missing validation of iallocator names; this
2279 allows local and remote execution of arbitrary executables
2280 - Fix failure of gnt-node list during instance removal
2281 - Ship the RAPI documentation in the archive
2287 *(Released Wed, 30 Sep 2009)*
2289 - Fixed many wrong messages
2290 - Fixed a few bugs related to the locking library
2291 - Fixed MAC checking at instance creation time
2292 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
2293 - Fixed a few issues related to signal handling in both daemons and
2295 - Fixed the example startup script provided
2296 - Fixed insserv dependencies in the example startup script (patch from
2298 - Fixed handling of drained nodes in the iallocator framework
2299 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
2301 - Fixed error related to invalid job IDs in job polling
2302 - Fixed job/opcode persistence on unclean master shutdown
2303 - Fixed handling of partial job processing after unclean master
2305 - Fixed error reporting from LUs, previously all errors were converted
2306 into execution errors
2307 - Fixed error reporting from burnin
2308 - Decreased significantly the memory usage of the job queue
2309 - Optimised slightly multi-job submission
2310 - Optimised slightly opcode loading
2311 - Backported the multi-job submit framework from the development
2312 branch; multi-instance start and stop should be faster
2313 - Added script to clean archived jobs after 21 days; this will reduce
2314 the size of the queue directory
2315 - Added some extra checks in disk size tracking
2316 - Added an example ethers hook script
2317 - Added a cluster parameter that prevents Ganeti from modifying of
2319 - Added more node information to RAPI responses
2320 - Added a ``gnt-job watch`` command that allows following the ouput of a
2322 - Added a bind-address option to ganeti-rapi
2323 - Added more checks to the configuration verify
2324 - Enhanced the burnin script such that some operations can be retried
2326 - Converted instance reinstall to multi-instance model
2332 *(Released Fri, 7 Aug 2009)*
2334 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
2335 to allow using the pre-2.0.2 behaviour in activation, if any existing
2336 instances have mismatched disk sizes in the configuration
2337 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
2338 any configuration mismatches for disk sizes
2339 - Added ``gnt-master cluste-failover --no-voting`` to allow master
2340 failover to work on two-node clusters
2341 - Fixed the ``--net`` option of ``gnt-backup import``, which was
2343 - Fixed detection of OS script errors in ``gnt-backup export``
2344 - Fixed exit code of ``gnt-backup export``
2350 *(Released Fri, 17 Jul 2009)*
2352 - Added experimental support for stripped logical volumes; this should
2353 enhance performance but comes with a higher complexity in the block
2354 device handling; stripping is only enabled when passing
2355 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
2356 affected even in the non-stripped mode
2357 - Improved resiliency against transient failures at the end of DRBD
2358 resyncs, and in general of DRBD resync checks
2359 - Fixed a couple of issues with exports and snapshot errors
2360 - Fixed a couple of issues in instance listing
2361 - Added display of the disk size in ``gnt-instance info``
2362 - Fixed checking for valid OSes in instance creation
2363 - Fixed handling of the "vcpus" parameter in instance listing and in
2364 general of invalid parameters
2365 - Fixed http server library, and thus RAPI, to handle invalid
2366 username/password combinations correctly; this means that now they
2367 report unauthorized for queries too, not only for modifications,
2368 allowing earlier detect of configuration problems
2369 - Added a new "role" node list field, equivalent to the master/master
2370 candidate/drained/offline flags combinations
2371 - Fixed cluster modify and changes of candidate pool size
2372 - Fixed cluster verify error messages for wrong files on regular nodes
2373 - Fixed a couple of issues with node demotion from master candidate role
2374 - Fixed node readd issues
2375 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
2376 - Added a new ``--no-voting`` option for masterfailover to fix failover
2377 on two-nodes clusters when the former master node is unreachable
2378 - Added instance reinstall over RAPI
2384 *(Released Tue, 16 Jun 2009)*
2386 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
2387 allow re-adding the start in single-user option (regression from 1.2)
2388 - the watcher writes the instance status to a file, to allow monitoring
2389 to report the instance status (from the master) based on cached
2390 results of the watcher's queries; while this can get stale if the
2391 watcher is being locked due to other work on the cluster, this is
2392 still an improvement
2393 - the watcher now also restarts the node daemon and the rapi daemon if
2395 - fixed the watcher to handle full and drained queue cases
2396 - hooks export more instance data in the environment, which helps if
2397 hook scripts need to take action based on the instance's properties
2398 (no longer need to query back into ganeti)
2399 - instance failovers when the instance is stopped do not check for free
2400 RAM, so that failing over a stopped instance is possible in low memory
2402 - rapi uses queries for tags instead of jobs (for less job traffic), and
2403 for cluster tags it won't talk to masterd at all but read them from
2405 - a couple of error handling fixes in RAPI
2406 - drbd handling: improved the error handling of inconsistent disks after
2407 resync to reduce the frequency of "there are some degraded disks for
2408 this instance" messages
2409 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
2410 error handling path called a wrong function name)
2416 *(Released Wed, 27 May 2009)*
2418 - no changes from rc5
2424 *(Released Wed, 20 May 2009)*
2426 - fix a couple of bugs (validation, argument checks)
2427 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
2428 - some small improvements to RAPI and IAllocator
2429 - make watcher automatically start the master daemon if down
2435 *(Released Mon, 27 Apr 2009)*
2437 - change the OS list to not require locks; this helps with big clusters
2438 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
2439 volume group is broken
2440 - ``gnt-instance info``, without any arguments, doesn't run for all
2441 instances anymore; either pass ``--all`` or pass the desired
2442 instances; this helps against mistakes on big clusters where listing
2443 the information for all instances takes a long time
2444 - miscellaneous doc and man pages fixes
2450 *(Released Wed, 8 Apr 2009)*
2452 - Change the internal locking model of some ``gnt-node`` commands, in
2453 order to reduce contention (and blocking of master daemon) when
2454 batching many creation/reinstall jobs
2455 - Fixes to Xen soft reboot
2456 - No longer build documentation at build time, instead distribute it in
2457 the archive, in order to reduce the need for the whole docbook/rst
2464 *(Released Fri, 27 Mar 2009)*
2466 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
2467 - Fix watcher startup sequence, improves the behaviour of busy clusters
2468 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
2469 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
2470 bind address and other places
2471 - Some documentation fixes and updates
2477 *(Released Mon, 2 Mar 2009)*
2479 - More documentation updates, now all docs should be more-or-less
2481 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
2483 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
2485 - Fix an issue related to $libdir/run/ganeti and cluster creation
2491 *(Released Thu, 19 Feb 2009)*
2493 - Xen PVM and KVM have switched the default value for the instance root
2494 disk to the first partition on the first drive, instead of the whole
2495 drive; this means that the OS installation scripts must be changed
2497 - Man pages have been updated
2498 - RAPI has been switched by default to HTTPS, and the exported functions
2499 should all work correctly
2500 - RAPI v1 has been removed
2501 - Many improvements to the KVM hypervisor
2502 - Block device errors are now better reported
2503 - Many other bugfixes and small improvements
2509 *(Released Mon, 26 Jan 2009)*
2511 - Version 2 is a general rewrite of the code and therefore the
2512 differences are too many to list, see the design document for 2.0 in
2513 the ``doc/`` subdirectory for more details
2514 - In this beta version there is not yet a migration path from 1.2 (there
2515 will be one in the final 2.0 release)
2516 - A few significant changes are:
2518 - all commands are executed by a daemon (``ganeti-masterd``) and the
2519 various ``gnt-*`` commands are just front-ends to it
2520 - all the commands are entered into, and executed from a job queue,
2521 see the ``gnt-job(8)`` manpage
2522 - the RAPI daemon supports read-write operations, secured by basic
2523 HTTP authentication on top of HTTPS
2524 - DRBD version 0.7 support has been removed, DRBD 8 is the only
2525 supported version (when migrating from Ganeti 1.2 to 2.0, you need
2526 to migrate to DRBD 8 first while still running Ganeti 1.2)
2527 - DRBD devices are using statically allocated minor numbers, which
2528 will be assigned to existing instances during the migration process
2529 - there is support for both Xen PVM and Xen HVM instances running on
2531 - KVM virtualization is supported too
2532 - file-based storage has been implemented, which means that it is
2533 possible to run the cluster without LVM and DRBD storage, for
2534 example using a shared filesystem exported from shared storage (and
2535 still have live migration)
2541 *(Released Tue, 13 Jan 2009)*
2543 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
2544 - Reuse the old instance mac address by default on instance import, if
2545 the instance name is the same.
2546 - Handle situations in which the node info rpc returns incomplete
2548 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
2549 - Improved version of batcher:
2551 - state file support
2552 - instance mac address support
2553 - support for HVM clusters/instances
2555 - Add an option to show the number of cpu sockets and nodes in
2557 - Support OSes that handle more than one version of the OS api (but do
2558 not change the current API in any other way)
2559 - Fix ``gnt-node migrate``
2560 - ``gnt-debug`` man page
2561 - Fixes various more typos and small issues
2562 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
2568 *(Released Wed, 24 Sep 2008)*
2570 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
2571 type of disk exported to fully virtualized instances.
2572 - provide access to the serial console of HVM instances
2573 - instance auto_balance flag, set by default. If turned off it will
2574 avoid warnings on cluster verify if there is not enough memory to fail
2575 over an instance. in the future it will prevent automatically failing
2576 it over when we will support that.
2577 - batcher tool for instance creation, see ``tools/README.batcher``
2578 - ``gnt-instance reinstall --select-os`` to interactively select a new
2579 operating system when reinstalling an instance.
2580 - when changing the memory amount on instance modify a check has been
2581 added that the instance will be able to start. also warnings are
2582 emitted if the instance will not be able to fail over, if auto_balance
2584 - documentation fixes
2585 - sync fields between ``gnt-instance list/modify/add/import``
2586 - fix a race condition in drbd when the sync speed was set after giving
2587 the device a remote peer.
2593 *(Released Tue, 22 Jul 2008)*
2595 - note: the allowed size and number of tags per object were reduced
2596 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
2597 - fixed twisted 8.x compatibility
2598 - fixed ``gnt-instance replace-disks`` with iallocator
2599 - add TCP keepalives on twisted connections to detect restarted nodes
2600 - disk increase support, see ``gnt-instance grow-disk``
2601 - implement bulk node/instance query for RAPI
2602 - add tags in node/instance listing (optional)
2603 - experimental migration (and live migration) support, read the man page
2604 for ``gnt-instance migrate``
2605 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
2606 has some small improvements in handling its state file
2612 *(Released Fri, 13 Jun 2008)*
2614 - Experimental readonly, REST-based remote API implementation;
2615 automatically started on master node, TCP port 5080, if enabled by
2616 ``--enable-rapi`` parameter to configure script.
2617 - Instance allocator support. Add and import instance accept a
2618 ``--iallocator`` parameter, and call that instance allocator to decide
2619 which node to use for the instance. The iallocator document describes
2620 what's expected from an allocator script.
2621 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
2622 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
2623 node is lost there is still enough memory to fail over the instances
2625 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
2626 ``gnt-cluster verify``, to check for site-specific compliance. All the
2627 hooks will run, and their output, if any, will be displayed. Any
2628 failing hook will make the verification return an error value.
2629 - ``gnt-cluster verify`` now checks that its peers are reachable on the
2630 primary and secondary interfaces
2631 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
2632 that is still declared as part of the cluster and has failed.
2633 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
2634 specifying output fields, that just adds the chosen fields to the
2636 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
2637 export from the filesystem.
2638 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
2639 have been added. Using them you can enable/disable acpi and pae
2640 support, and specify a path for a cd image to be exported to the
2641 instance. These parameters as the name suggest only work on HVM
2643 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
2644 PAE support will be set to the previously hardcoded values, but the
2645 (previously hardcoded) path to the CDROM ISO image will be unset and
2646 if required, needs to be set manually with ``gnt-instance modify``
2648 - The address to which an instance's VNC console is bound is now
2649 selectable per-instance, rather than being cluster wide. Of course
2650 this only applies to instances controlled via VNC, so currently just
2651 applies to HVM clusters.
2657 *(Released Mon, 18 Feb 2008)*
2659 - more tweaks to the disk activation code (especially helpful for DRBD)
2660 - change the default ``gnt-instance list`` output format, now there is
2661 one combined status field (see the manpage for the exact values this
2663 - some more fixes for the mac export to hooks change
2664 - make Ganeti not break with DRBD 8.2.x (which changed the version
2665 format in ``/proc/drbd``) (issue 24)
2666 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
2667 template, allowing migration from DRBD0.7+MD to DRBD8
2673 *(Released Wed, 30 Jan 2008)*
2675 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
2677 - add command aliases infrastructure and a few aliases
2678 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
2679 man pages and the ``--help`` option of ``gnt-node
2680 list``/``gnt-instance list``
2681 - fix ``gnt-backup list`` with down nodes (issue 21)
2682 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
2683 - fix the dist archive and add a check for including svn/git files in
2685 - some developer-related changes: improve the burnin and the QA suite,
2686 add an upload script for testing during development
2692 *(Released Wed, 16 Jan 2008)*
2694 - experimental HVM support, read the install document, section
2695 "Initializing the cluster"
2696 - allow for the PVM hypervisor per-instance kernel and initrd paths
2697 - add a new command ``gnt-cluster verify-disks`` which uses a new
2698 algorithm to improve the reconnection of the DRBD pairs if the device
2699 on the secondary node has gone away
2700 - make logical volume code auto-activate LVs at disk activation time
2701 - slightly improve the speed of activating disks
2702 - allow specification of the MAC address at instance creation time, and
2703 changing it later via ``gnt-instance modify``
2704 - fix handling of external commands that generate lots of output on
2706 - update documentation with regard to minimum version of DRBD8 supported
2712 *(Released Tue, 4 Dec 2007)*
2714 - Log the ``xm create`` output to the node daemon log on failure (to
2715 help diagnosing the error)
2716 - In debug mode, log all external commands output if failed to the logs
2717 - Change parsing of lvm commands to ignore stderr
2723 *(Released Wed, 28 Nov 2007)*
2725 - Another round of updates to the DRBD 8 code to deal with more failures
2726 in the replace secondary node operation
2727 - Some more logging of failures in disk operations (lvm, drbd)
2728 - A few documentation updates
2735 *(Released Tue, 13 Nov 2007)*
2737 - Change configuration file format from Python's Pickle to JSON.
2738 Upgrading is possible using the cfgupgrade utility.
2739 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
2740 faster replace disks and is more stable (DRBD 8 has many improvements
2741 compared to DRBD 0.7)
2742 - Added command line tags support (see man pages for ``gnt-instance``,
2743 ``gnt-node``, ``gnt-cluster``)
2744 - Added instance rename support
2745 - Added multi-instance startup/shutdown
2746 - Added cluster rename support
2747 - Added ``gnt-node evacuate`` to simplify some node operations
2748 - Added instance reboot operation that can speedup reboot as compared to
2750 - Soften the requirement that hostnames are in FQDN format
2751 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
2753 - Removed dependency on debian's patched fping that uses the
2754 non-standard ``-S`` option
2755 - Now the OS definitions are searched for in multiple, configurable
2756 paths (easier for distros to package)
2757 - Some changes to the hooks infrastructure (especially the new
2758 post-configuration update hook)
2759 - Other small bugfixes
2761 .. vim: set textwidth=72 syntax=rst :