10 - DRBD 8.4 support. Depending on the installed DRBD version, Ganeti now uses
11 the correct command syntax. It is possible to use different DRBD versions
12 on different nodes as long as they are compatible to each other. This
13 enables rolling upgrades of DRBD with no downtime. As permanent operation
14 of different DRBD versions within a node group is discouraged,
15 ``gnt-cluster verify`` will emit a warning if it detects such a situation.
16 - hroller now also plans for capacity to move non-redundant instances off
17 any node to be rebooted; the old behavior of completely ignoring any
18 non-redundant instances can be restored by adding the --ignore-non-redundant
20 - The cluster option '--no-lvm-storage' was removed in favor of the new option
21 '--enabled-disk-templates'.
29 Incompatible/important changes
30 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
32 - Instance policy can contain multiple instance specs, as described in
33 the “Constrained instance sizes” section of :doc:`Partitioned Ganeti
34 <design-partitioned>`. As a consequence, it's not possible to partially change
35 or override instance specs. Bounding specs (min and max) can be specified as a
36 whole using the new option ``--ipolicy-bounds-specs``, while standard
37 specs use the new option ``--ipolicy-std-specs``.
38 - The output of the info command of gnt-cluster, gnt-group, gnt-node,
39 gnt-instance is a valid YAML object.
40 - hail now honors network restrictions when allocating nodes. This led to an
41 update of the IAllocator protocol. See the IAllocator documentation for
47 - The :doc:`Remote API <rapi>` daemon now supports a command line flag
48 to always require authentication, ``--require-authentication``. It can
49 be specified in ``$sysconfdir/default/ganeti``.
50 - A new cluster attribute 'enabled_disk_templates' is introduced. It will
51 be used to manage the disk templates to be used by instances in the cluster.
52 Initially, it will be set to a list that includes plain, drbd, if they were
53 enabled by specifying a volume group name, and file and sharedfile, if those
54 were enabled at configure time. Additionally, it will include all disk
55 templates that are currently used by instances. The order of disk templates
56 will be based on Ganeti's history of supporting them. In the future, the
57 first entry of the list will be used as a default disk template on instance
59 - ``cfgupgrade`` now supports a ``--downgrade`` option to bring the
60 configuration back to the previous stable version.
61 - Disk templates in group ipolicy can be restored to the default value.
62 - Initial support for diskless instances and virtual clusters in QA.
63 - More QA and unit tests for instance policies.
64 - Every opcode now contains a reason trail (visible through ``gnt-job info``)
65 describing why the opcode itself was executed.
66 - The monitoring daemon is now available. It allows users to query the cluster
67 for obtaining information about the status of the system. The daemon is only
68 responsible for providing the information over the network: the actual data
69 gathering is performed by data collectors (currently, only the DRBD status
70 collector is available).
71 - In order to help developers work on Ganeti, a new script
72 (``devel/build_chroot``) is provided, for building a chroot that contains all
73 the required development libraries and tools for compiling Ganeti on a Debian
75 - A new tool, ``harep``, for performing self-repair and recreation of instances
76 in Ganeti has been added.
77 - Split queries are enabled for tags, network, exports, cluster info, groups,
79 - New command ``show-ispecs-cmd`` for ``gnt-cluster`` and ``gnt-group``.
80 It prints the command line to set the current policies, to ease
85 The following new dependencies have been added:
88 - The ``curl`` library is not optional anymore for compiling the Haskell code.
89 - ``snap-server`` library (if monitoring is enabled).
92 - The minimum Python version needed to run Ganeti is now 2.6.
93 - ``yaml`` library (only for running the QA).
99 *(Released Fri, 24 May 2013)*
101 Incompatible/important changes
102 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
104 - Instance policies for disk size were documented to be on a per-disk
105 basis, but hail applied them to the sum of all disks. This has been
107 - ``hbal`` will now exit with status 0 if, during job execution over
108 LUXI, early exit has been requested and all jobs are successful;
109 before, exit status 1 was used, which cannot be differentiated from
111 - Compatibility with newer versions of rbd has been fixed
112 - ``gnt-instance batch-create`` has been changed to use the bulk create
113 opcode from Ganeti. This lead to incompatible changes in the format of
114 the JSON file. It's now not a custom dict anymore but a dict
115 compatible with the ``OpInstanceCreate`` opcode.
116 - Parent directories for file storage need to be listed in
117 ``$sysconfdir/ganeti/file-storage-paths`` now. ``cfgupgrade`` will
118 write the file automatically based on old configuration values, but it
119 can not distribute it across all nodes and the file contents should be
120 verified. Use ``gnt-cluster copyfile
121 $sysconfdir/ganeti/file-storage-paths`` once the cluster has been
122 upgraded. The reason for requiring this list of paths now is that
123 before it would have been possible to inject new paths via RPC,
124 allowing files to be created in arbitrary locations. The RPC protocol
125 is protected using SSL/X.509 certificates, but as a design principle
126 Ganeti does not permit arbitrary paths to be passed.
127 - The parsing of the variants file for OSes (see
128 :manpage:`ganeti-os-interface(7)`) has been slightly changed: now empty
129 lines and comment lines (starting with ``#``) are ignored for better
131 - The ``setup-ssh`` tool added in Ganeti 2.2 has been replaced and is no
132 longer available. ``gnt-node add`` now invokes a new tool on the
133 destination node, named ``prepare-node-join``, to configure the SSH
134 daemon. Paramiko is no longer necessary to configure nodes' SSH
135 daemons via ``gnt-node add``.
136 - Draining (``gnt-cluster queue drain``) and un-draining the job queue
137 (``gnt-cluster queue undrain``) now affects all nodes in a cluster and
138 the flag is not reset after a master failover.
139 - Python 2.4 has *not* been tested with this release. Using 2.6 or above
140 is recommended. 2.6 will be mandatory from the 2.8 series.
146 - New network management functionality to support automatic allocation
147 of IP addresses and managing of network parameters. See
148 :manpage:`gnt-network(8)` for more details.
149 - New external storage backend, to allow managing arbitrary storage
150 systems external to the cluster. See
151 :manpage:`ganeti-extstorage-interface(7)`.
152 - New ``exclusive-storage`` node parameter added, restricted to
153 nodegroup level. When it's set to true, physical disks are assigned in
154 an exclusive fashion to instances, as documented in :doc:`Partitioned
155 Ganeti <design-partitioned>`. Currently, only instances using the
156 ``plain`` disk template are supported.
157 - The KVM hypervisor has been updated with many new hypervisor
158 parameters, including a generic one for passing arbitrary command line
159 values. See a complete list in :manpage:`gnt-instance(8)`. It is now
160 compatible up to qemu 1.4.
161 - A new tool, called ``mon-collector``, is the stand-alone executor of
162 the data collectors for a monitoring system. As of this version, it
163 just includes the DRBD data collector, that can be executed by calling
164 ``mon-collector`` using the ``drbd`` parameter. See
165 :manpage:`mon-collector(7)`.
166 - A new user option, :pyeval:`rapi.RAPI_ACCESS_READ`, has been added
167 for RAPI users. It allows granting permissions to query for
168 information to a specific user without giving
169 :pyeval:`rapi.RAPI_ACCESS_WRITE` permissions.
170 - A new tool named ``node-cleanup`` has been added. It cleans remains of
171 a cluster from a machine by stopping all daemons, removing
172 certificates and ssconf files. Unless the ``--no-backup`` option is
173 given, copies of the certificates are made.
174 - Instance creations now support the use of opportunistic locking,
175 potentially speeding up the (parallel) creation of multiple instances.
176 This feature is currently only available via the :doc:`RAPI
177 <rapi>` interface and when an instance allocator is used. If the
178 ``opportunistic_locking`` parameter is set the opcode will try to
179 acquire as many locks as possible, but will not wait for any locks
180 held by other opcodes. If not enough resources can be found to
181 allocate the instance, the temporary error code
182 :pyeval:`errors.ECODE_TEMP_NORES` is returned. The operation can be
183 retried thereafter, with or without opportunistic locking.
184 - New experimental linux-ha resource scripts.
185 - Restricted-commands support: ganeti can now be asked (via command line
186 or rapi) to perform commands on a node. These are passed via ganeti
187 RPC rather than ssh. This functionality is restricted to commands
188 specified on the ``$sysconfdir/ganeti/restricted-commands`` for security
189 reasons. The file is not copied automatically.
195 - Diskless instances are now externally mirrored (Issue 237). This for
196 now has only been tested in conjunction with explicit target nodes for
198 - Queries not needing locks or RPC access to the node can now be
199 performed by the confd daemon, making them independent from jobs, and
200 thus faster to execute. This is selectable at configure time.
201 - The functionality for allocating multiple instances at once has been
202 overhauled and is now also available through :doc:`RAPI <rapi>`.
206 - ``devel/upload`` now works when ``/var/run`` on the target nodes is a
208 - Disks added through ``gnt-instance modify`` or created through
209 ``gnt-instance recreate-disks`` are wiped, if the
210 ``prealloc_wipe_disks`` flag is set.
211 - If wiping newly created disks fails, the disks are removed. Also,
212 partial failures in creating disks through ``gnt-instance modify``
213 triggers a cleanup of the partially-created disks.
214 - Removing the master IP address doesn't fail if the address has been
216 - Fix ownership of the OS log dir
217 - Workaround missing SO_PEERCRED constant (Issue 191)
223 *(Released Fri, 3 May 2013)*
225 This was the first release candidate of the 2.7 series. Since beta3:
227 - Fix kvm compatibility with qemu 1.4 (Issue 389)
228 - Documentation updates (admin guide, upgrade notes, install
229 instructions) (Issue 372)
230 - Fix gnt-group list nodes and instances count (Issue 436)
231 - Fix compilation without non-mandatory libraries (Issue 441)
232 - Fix xen-hvm hypervisor forcing nics to type 'ioemu' (Issue 247)
233 - Make confd logging more verbose at INFO level (Issue 435)
234 - Improve "networks" documentation in :manpage:`gnt-instance(8)`
235 - Fix failure path for instance storage type conversion (Issue 229)
236 - Update htools text backend documentation
237 - Improve the renew-crypto section of :manpage:`gnt-cluster(8)`
238 - Disable inter-cluster instance move for file-based instances, because
239 it is dependant on instance export, which is not supported for
240 file-based instances. (Issue 414)
241 - Fix gnt-job crashes on non-ascii characters (Issue 427)
242 - Fix volume group checks on non-vm-capable nodes (Issue 432)
248 *(Released Mon, 22 Apr 2013)*
250 This was the third beta release of the 2.7 series. Since beta2:
252 - Fix hail to verify disk instance policies on a per-disk basis (Issue 418).
253 - Fix data loss on wrong usage of ``gnt-instance move``
254 - Properly export errors in confd-based job queries
255 - Add ``users-setup`` tool
256 - Fix iallocator protocol to report 0 as a disk size for diskless
257 instances. This avoids hail breaking when a diskless instance is
259 - Fix job queue directory permission problem that made confd job queries
260 fail. This requires running an ``ensure-dirs --full-run`` on upgrade
261 for access to archived jobs (Issue 406).
262 - Limit the sizes of networks supported by ``gnt-network`` to something
263 between a ``/16`` and a ``/30`` to prevent memory bloat and crashes.
264 - Fix bugs in instance disk template conversion
265 - Fix GHC 7 compatibility
266 - Fix ``burnin`` install path (Issue 426).
267 - Allow very small disk grows (Issue 347).
268 - Fix a ``ganeti-noded`` memory bloat introduced in 2.5, by making sure
269 that noded doesn't import masterd code (Issue 419).
270 - Make sure the default metavg at cluster init is the same as the vg, if
271 unspecified (Issue 358).
272 - Fix cleanup of partially created disks (part of Issue 416)
278 *(Released Tue, 2 Apr 2013)*
280 This was the second beta release of the 2.7 series. Since beta1:
282 - Networks no longer have a "type" slot, since this information was
283 unused in Ganeti: instead of it tags should be used.
284 - The rapi client now has a ``target_node`` option to MigrateInstance.
285 - Fix early exit return code for hbal (Issue 386).
286 - Fix ``gnt-instance migrate/failover -n`` (Issue 396).
287 - Fix ``rbd showmapped`` output parsing (Issue 312).
288 - Networks are now referenced indexed by UUID, rather than name. This
289 will require running cfgupgrade, from 2.7.0beta1, if networks are in
291 - The OS environment now includes network information.
292 - Deleting of a network is now disallowed if any instance nic is using
293 it, to prevent dangling references.
294 - External storage is now documented in man pages.
295 - The exclusive_storage flag can now only be set at nodegroup level.
296 - Hbal can now submit an explicit priority with its jobs.
297 - Many network related locking fixes.
298 - Bump up the required pylint version to 0.25.1.
299 - Fix the ``no_remember`` option in RAPI client.
300 - Many ipolicy related tests, qa, and fixes.
301 - Many documentation improvements and fixes.
302 - Fix building with ``--disable-file-storage``.
303 - Fix ``-q`` option in htools, which was broken if passed more than
305 - Some haskell/python interaction improvements and fixes.
306 - Fix iallocator in case of missing LVM storage.
307 - Fix confd config load in case of ``--no-lvm-storage``.
308 - The confd/query functionality is now mentioned in the security
315 *(Released Wed, 6 Feb 2013)*
317 This was the first beta release of the 2.7 series. All important changes
318 are listed in the latest 2.7 entry.
324 *(Released Fri, 21 Dec 2012)*
326 Important behaviour change: hbal won't rebalance anymore instances which
327 have the ``auto_balance`` attribute set to false. This was the intention
328 all along, but until now it only skipped those from the N+1 memory
329 reservation (DRBD-specific).
331 A significant number of bug fixes in this release:
333 - Fixed disk adoption interaction with ipolicy checks.
334 - Fixed networking issues when instances are started, stopped or
335 migrated, by forcing the tap device's MAC prefix to "fe" (issue 217).
336 - Fixed the warning in cluster verify for shared storage instances not
338 - Fixed removal of storage directory on shared file storage (issue 262).
339 - Fixed validation of LVM volume group name in OpClusterSetParams
340 (``gnt-cluster modify``) (issue 285).
341 - Fixed runtime memory increases (``gnt-instance modify -m``).
342 - Fixed live migration under Xen's ``xl`` mode.
343 - Fixed ``gnt-instance console`` with ``xl``.
344 - Fixed building with newer Haskell compiler/libraries.
345 - Fixed PID file writing in Haskell daemons (confd); this prevents
346 restart issues if confd was launched manually (outside of
347 ``daemon-util``) while another copy of it was running
348 - Fixed a type error when doing live migrations with KVM (issue 297) and
349 the error messages for failing migrations have been improved.
350 - Fixed opcode validation for the out-of-band commands (``gnt-node
352 - Fixed a type error when unsetting OS hypervisor parameters (issue
353 311); now it's possible to unset all OS-specific hypervisor
355 - Fixed the ``dry-run`` mode for many operations: verification of
356 results was over-zealous but didn't take into account the ``dry-run``
357 operation, resulting in "wrong" failures.
358 - Fixed bash completion in ``gnt-job list`` when the job queue has
359 hundreds of entries; especially with older ``bash`` versions, this
360 results in significant CPU usage.
362 And lastly, a few other improvements have been made:
364 - Added option to force master-failover without voting (issue 282).
365 - Clarified error message on lock conflict (issue 287).
366 - Logging of newly submitted jobs has been improved (issue 290).
367 - Hostname checks have been made uniform between instance rename and
369 - The ``--submit`` option is now supported by ``gnt-debug delay``.
370 - Shutting down the master daemon by sending SIGTERM now stops it from
371 processing jobs waiting for locks; instead, those jobs will be started
372 once again after the master daemon is started the next time (issue
374 - Support for Xen's ``xl`` program has been improved (besides the fixes
376 - Reduced logging noise in the Haskell confd daemon (only show one log
377 entry for each config reload, instead of two).
378 - Several man page updates and typo fixes.
384 *(Released Fri, 12 Oct 2012)*
386 A small bugfix release. Among the bugs fixed:
388 - Fixed double use of ``PRIORITY_OPT`` in ``gnt-node migrate``, that
389 made the command unusable.
390 - Commands that issue many jobs don't fail anymore just because some jobs
391 take so long that other jobs are archived.
392 - Failures during ``gnt-instance reinstall`` are reflected by the exit
394 - Issue 190 fixed. Check for DRBD in cluster verify is enabled only when
396 - When ``always_failover`` is set, ``--allow-failover`` is not required
397 in migrate commands anymore.
398 - ``bash_completion`` works even if extglob is disabled.
399 - Fixed bug with locks that made failover for RDB-based instances fail.
400 - Fixed bug in non-mirrored instance allocation that made Ganeti choose
401 a random node instead of one based on the allocator metric.
402 - Support for newer versions of pylint and pep8.
403 - Hail doesn't fail anymore when trying to add an instance of type
404 ``file``, ``sharedfile`` or ``rbd``.
405 - Added new Makefile target to rebuild the whole distribution, so that
406 all files are included.
412 *(Released Fri, 27 Jul 2012)*
415 .. attention:: The ``LUXI`` protocol has been made more consistent
416 regarding its handling of command arguments. This, however, leads to
417 incompatibility issues with previous versions. Please ensure that you
418 restart Ganeti daemons soon after the upgrade, otherwise most
419 ``LUXI`` calls (job submission, setting/resetting the drain flag,
420 pausing/resuming the watcher, cancelling and archiving jobs, querying
421 the cluster configuration) will fail.
430 The current ``admin_up`` field, which used to denote whether an instance
431 should be running or not, has been removed. Instead, ``admin_state`` is
432 introduced, with 3 possible values -- ``up``, ``down`` and ``offline``.
434 The rational behind this is that an instance being “down” can have
437 - it could be down during a reboot
438 - it could be temporarily be down for a reinstall
439 - or it could be down because it is deprecated and kept just for its
442 The previous Boolean state was making it difficult to do capacity
443 calculations: should Ganeti reserve memory for a down instance? Now, the
444 tri-state field makes it clear:
446 - in ``up`` and ``down`` state, all resources are reserved for the
447 instance, and it can be at any time brought up if it is down
448 - in ``offline`` state, only disk space is reserved for it, but not
451 The field can have an extra use: since the transition between ``up`` and
452 ``down`` and vice-versus is done via ``gnt-instance start/stop``, but
453 transition between ``offline`` and ``down`` is done via ``gnt-instance
454 modify``, it is possible to given different rights to users. For
455 example, owners of an instance could be allowed to start/stop it, but
456 not transition it out of the offline state.
458 Instance policies and specs
459 +++++++++++++++++++++++++++
461 In previous Ganeti versions, an instance creation request was not
462 limited on the minimum size and on the maximum size just by the cluster
463 resources. As such, any policy could be implemented only in third-party
464 clients (RAPI clients, or shell wrappers over ``gnt-*``
465 tools). Furthermore, calculating cluster capacity via ``hspace`` again
466 required external input with regards to instance sizes.
468 In order to improve these workflows and to allow for example better
469 per-node group differentiation, we introduced instance specs, which
472 - minimum instance disk size, disk count, memory size, cpu count
473 - maximum values for the above metrics
474 - and “standard” values (used in ``hspace`` to calculate the standard
477 The minimum/maximum values can be also customised at node-group level,
478 for example allowing more powerful hardware to support bigger instance
481 Beside the instance specs, there are a few other settings belonging to
482 the instance policy framework. It is possible now to customise, per
483 cluster and node-group:
485 - the list of allowed disk templates
486 - the maximum ratio of VCPUs per PCPUs (to control CPU oversubscription)
487 - the maximum ratio of instance to spindles (see below for more
488 information) for local storage
490 All these together should allow all tools that talk to Ganeti to know
491 what are the ranges of allowed values for instances and the
492 over-subscription that is allowed.
494 For the VCPU/PCPU ratio, we already have the VCPU configuration from the
495 instance configuration, and the physical CPU configuration from the
496 node. For the spindle ratios however, we didn't track before these
497 values, so new parameters have been added:
499 - a new node parameter ``spindle_count``, defaults to 1, customisable at
500 node group or node level
501 - at new backend parameter (for instances), ``spindle_use`` defaults to 1
503 Note that spindles in this context doesn't need to mean actual
504 mechanical hard-drives; it's just a relative number for both the node
505 I/O capacity and instance I/O consumption.
507 Instance migration behaviour
508 ++++++++++++++++++++++++++++
510 While live-migration is in general desirable over failover, it is
511 possible that for some workloads it is actually worse, due to the
512 variable time of the “suspend” phase during live migration.
514 To allow the tools to work consistently over such instances (without
515 having to hard-code instance names), a new backend parameter
516 ``always_failover`` has been added to control the migration/failover
517 behaviour. When set to True, all migration requests for an instance will
518 instead fall-back to failover.
520 Instance memory ballooning
521 ++++++++++++++++++++++++++
523 Initial support for memory ballooning has been added. The memory for an
524 instance is no longer fixed (backend parameter ``memory``), but instead
525 can vary between minimum and maximum values (backend parameters
526 ``minmem`` and ``maxmem``). Currently we only change an instance's
529 - live migrating or failing over and instance and the target node
530 doesn't have enough memory
531 - user requests changing the memory via ``gnt-instance modify
537 In order to control the use of specific CPUs by instance, support for
538 controlling CPU pinning has been added for the Xen, HVM and LXC
539 hypervisors. This is controlled by a new hypervisor parameter
540 ``cpu_mask``; details about possible values for this are in the
541 :manpage:`gnt-instance(8)`. Note that use of the most specific (precise
542 VCPU-to-CPU mapping) form will work well only when all nodes in your
543 cluster have the same amount of CPUs.
548 Another area in which Ganeti was not customisable were the parameters
549 used for storage configuration, e.g. how many stripes to use for LVM,
550 DRBD resync configuration, etc.
552 To improve this area, we've added disks parameters, which are
553 customisable at cluster and node group level, and which allow to
554 specify various parameters for disks (DRBD has the most parameters
555 currently), for example:
557 - DRBD resync algorithm and parameters (e.g. speed)
558 - the default VG for meta-data volumes for DRBD
559 - number of stripes for LVM (plain disk template)
562 These parameters can be modified via ``gnt-cluster modify -D …`` and
563 ``gnt-group modify -D …``, and are used at either instance creation (in
564 case of LVM stripes, for example) or at disk “activation” time
567 Rados block device support
568 ++++++++++++++++++++++++++
570 A Rados (http://ceph.com/wiki/Rbd) storage backend has been added,
571 denoted by the ``rbd`` disk template type. This is considered
572 experimental, feedback is welcome. For details on configuring it, see
573 the :doc:`install` document and the :manpage:`gnt-cluster(8)` man page.
578 The existing master IP functionality works well only in simple setups (a
579 single network shared by all nodes); however, if nodes belong to
580 different networks, then the ``/32`` setup and lack of routing
581 information is not enough.
583 To allow the master IP to function well in more complex cases, the
584 system was reworked as follows:
586 - a master IP netmask setting has been added
587 - the master IP activation/turn-down code was moved from the node daemon
589 - whether to run the Ganeti-supplied master IP script or a user-supplied
590 on is a ``gnt-cluster init`` setting
592 Details about the location of the standard and custom setup scripts are
593 in the man page :manpage:`gnt-cluster(8)`; for information about the
594 setup script protocol, look at the Ganeti-supplied script.
599 The `SPICE <http://www.linux-kvm.org/page/SPICE>`_ support has been
602 It is now possible to use TLS-protected connections, and when renewing
603 or changing the cluster certificates (via ``gnt-cluster renew-crypto``,
604 it is now possible to specify spice or spice CA certificates. Also, it
605 is possible to configure a password for SPICE sessions via the
606 hypervisor parameter ``spice_password_file``.
608 There are also new parameters to control the compression and streaming
609 options (e.g. ``spice_image_compression``, ``spice_streaming_video``,
610 etc.). For details, see the man page :manpage:`gnt-instance(8)` and look
611 for the spice parameters.
613 Lastly, it is now possible to see the SPICE connection information via
614 ``gnt-instance console``.
619 A new tool (``tools/ovfconverter``) has been added that supports
620 conversion between Ganeti and the `Open Virtualization Format
621 <http://en.wikipedia.org/wiki/Open_Virtualization_Format>`_ (both to and
624 This relies on the ``qemu-img`` tool to convert the disk formats, so the
625 actual compatibility with other virtualization solutions depends on it.
630 The configuration query daemon (``ganeti-confd``) is now optional, and
631 has been rewritten in Haskell; whether to use the daemon at all, use the
632 Python (default) or the Haskell version is selectable at configure time
633 via the ``--enable-confd`` parameter, which can take one of the
634 ``haskell``, ``python`` or ``no`` values. If not used, disabling the
635 daemon will result in a smaller footprint; for larger systems, we
636 welcome feedback on the Haskell version which might become the default
639 If you want to use ``gnt-node list-drbd`` you need to have the Haskell
640 daemon running. The Python version doesn't implement the new call.
643 User interface changes
644 ~~~~~~~~~~~~~~~~~~~~~~
646 We have replaced the ``--disks`` option of ``gnt-instance
647 replace-disks`` with a more flexible ``--disk`` option, which allows
648 adding and removing disks at arbitrary indices (Issue 188). Furthermore,
649 disk size and mode can be changed upon recreation (via ``gnt-instance
650 recreate-disks``, which accepts the same ``--disk`` option).
652 As many people are used to a ``show`` command, we have added that as an
653 alias to ``info`` on all ``gnt-*`` commands.
655 The ``gnt-instance grow-disk`` command has a new mode in which it can
656 accept the target size of the disk, instead of the delta; this can be
657 more safe since two runs in absolute mode will be idempotent, and
658 sometimes it's also easier to specify the desired size directly.
660 Also the handling of instances with regard to offline secondaries has
661 been improved. Instance operations should not fail because one of it's
662 secondary nodes is offline, even though it's safe to proceed.
664 A new command ``list-drbd`` has been added to the ``gnt-node`` script to
665 support debugging of DRBD issues on nodes. It provides a mapping of DRBD
666 minors to instance name.
671 RAPI coverage has improved, with (for example) new resources for
672 recreate-disks, node power-cycle, etc.
677 There is partial support for ``xl`` in the Xen hypervisor; feedback is
680 Python 2.7 is better supported, and after Ganeti 2.6 we will investigate
681 whether to still support Python 2.4 or move to Python 2.6 as minimum
684 Support for Fedora has been slightly improved; the provided example
685 init.d script should work better on it and the INSTALL file should
686 document the needed dependencies.
691 The deprecated ``QueryLocks`` LUXI request has been removed. Use
692 ``Query(what=QR_LOCK, ...)`` instead.
694 The LUXI requests :pyeval:`luxi.REQ_QUERY_JOBS`,
695 :pyeval:`luxi.REQ_QUERY_INSTANCES`, :pyeval:`luxi.REQ_QUERY_NODES`,
696 :pyeval:`luxi.REQ_QUERY_GROUPS`, :pyeval:`luxi.REQ_QUERY_EXPORTS` and
697 :pyeval:`luxi.REQ_QUERY_TAGS` are deprecated and will be removed in a
698 future version. :pyeval:`luxi.REQ_QUERY` should be used instead.
700 RAPI client: ``CertificateError`` now derives from
701 ``GanetiApiError``. This should make it more easy to handle Ganeti
704 Deprecation warnings due to PyCrypto/paramiko import in
705 ``tools/setup-ssh`` have been silenced, as usually they are safe; please
706 make sure to run an up-to-date paramiko version, if you use this tool.
708 The QA scripts now depend on Python 2.5 or above (the main code base
709 still works with Python 2.4).
711 The configuration file (``config.data``) is now written without
712 indentation for performance reasons; if you want to edit it, it can be
713 re-formatted via ``tools/fmtjson``.
715 A number of bugs has been fixed in the cluster merge tool.
717 ``x509`` certification verification (used in import-export) has been
718 changed to allow the same clock skew as permitted by the cluster
719 verification. This will remove some rare but hard to diagnose errors in
726 *(Released Thu, 19 Jul 2012)*
728 Very few changes from rc4 to the final release, only bugfixes:
730 - integrated fixes from release 2.5.2 (fix general boot flag for KVM
731 instance, fix CDROM booting for KVM instances)
732 - fixed node group modification of node parameters
733 - fixed issue in LUClusterVerifyGroup with multi-group clusters
734 - fixed generation of bash completion to ensure a stable ordering
741 *(Released Fri, 13 Jul 2012)*
743 Third release candidate for 2.6. The following changes were done from
746 - Fixed ``UpgradeConfig`` w.r.t. to disk parameters on disk objects.
747 - Fixed an inconsistency in the LUXI protocol with the provided
748 arguments (NOT backwards compatible)
749 - Fixed a bug with node groups ipolicy where ``min`` was greater than
750 the cluster ``std`` value
751 - Implemented a new ``gnt-node list-drbd`` call to list DRBD minors for
752 easier instance debugging on nodes (requires ``hconfd`` to work)
758 *(Released Tue, 03 Jul 2012)*
760 Second release candidate for 2.6. The following changes were done from
763 - Fixed ``gnt-cluster verify`` regarding ``master-ip-script`` on non
765 - Fixed a RAPI regression on missing beparams/memory
766 - Fixed redistribution of files on offline nodes
767 - Added possibility to run activate-disks even though secondaries are
768 offline. With this change it relaxes also the strictness on some other
769 commands which use activate disks internally:
770 * ``gnt-instance start|reboot|rename|backup|export``
771 - Made it possible to remove safely an instance if its secondaries are
773 - Made it possible to reinstall even though secondaries are offline
779 *(Released Mon, 25 Jun 2012)*
781 First release candidate for 2.6. The following changes were done from
784 - Fixed bugs with disk parameters and ``rbd`` templates as well as
786 - Made ``gnt-instance modify`` more consistent regarding new NIC/Disk
787 behaviour. It supports now the modify operation
788 - ``hcheck`` implemented to analyze cluster health and possibility of
789 improving health by rebalance
790 - ``hbal`` has been improved in dealing with split instances
796 *(Released Mon, 11 Jun 2012)*
798 Second beta release of 2.6. The following changes were done from beta2
801 - Fixed ``daemon-util`` with non-root user models
802 - Fixed creation of plain instances with ``--no-wait-for-sync``
803 - Fix wrong iv_names when running ``cfgupgrade``
804 - Export more information in RAPI group queries
805 - Fixed bug when changing instance network interfaces
806 - Extended burnin to do NIC changes
807 - query: Added ``<``, ``>``, ``<=``, ``>=`` comparison operators
808 - Changed default for DRBD barriers
809 - Fixed DRBD error reporting for syncer rate
810 - Verify the options on disk parameters
812 And of course various fixes to documentation and improved unittests and
819 *(Released Wed, 23 May 2012)*
821 First beta release of 2.6. The following changes were done from beta1 to
824 - integrated patch for distributions without ``start-stop-daemon``
825 - adapted example init.d script to work on Fedora
826 - fixed log handling in Haskell daemons
827 - adapted checks in the watcher for pycurl linked against libnss
828 - add partial support for ``xl`` instead of ``xm`` for Xen
829 - fixed a type issue in cluster verification
830 - fixed ssconf handling in the Haskell code (was breaking confd in IPv6
833 Plus integrated fixes from the 2.5 branch:
835 - fixed ``kvm-ifup`` to use ``/bin/bash``
836 - fixed parallel build failures
837 - KVM live migration when using a custom keymap
843 *(Released Tue, 24 Jul 2012)*
845 A small bugfix release, with no new features:
847 - fixed bash-isms in kvm-ifup, for compatibility with systems which use a
848 different default shell (e.g. Debian, Ubuntu)
849 - fixed KVM startup and live migration with a custom keymap (fixes Issue
850 243 and Debian bug #650664)
851 - fixed compatibility with KVM versions that don't support multiple boot
852 devices (fixes Issue 230 and Debian bug #624256)
854 Additionally, a few fixes were done to the build system (fixed parallel
855 build failures) and to the unittests (fixed race condition in test for
856 FileID functions, and the default enable/disable mode for QA test is now
863 *(Released Fri, 11 May 2012)*
865 A small bugfix release.
867 The main issues solved are on the topic of compatibility with newer LVM
870 - fixed parsing of ``lv_attr`` field
871 - adapted to new ``vgreduce --removemissing`` behaviour where sometimes
872 the ``--force`` flag is needed
874 Also on the topic of compatibility, ``tools/lvmstrap`` has been changed
875 to accept kernel 3.x too (was hardcoded to 2.6.*).
877 A regression present in 2.5.0 that broke handling (in the gnt-* scripts)
878 of hook results and that also made display of other errors suboptimal
879 was fixed; the code behaves now like 2.4 and earlier.
881 Another change in 2.5, the cleanup of the OS scripts environment, is too
882 aggressive: it removed even the ``PATH`` variable, which requires the OS
883 scripts to *always* need to export it. Since this is a bit too strict,
884 we now export a minimal PATH, the same that we export for hooks.
886 The fix for issue 201 (Preserve bridge MTU in KVM ifup script) was
887 integrated into this release.
889 Finally, a few other miscellaneous changes were done (no new features,
890 just small improvements):
892 - Fix ``gnt-group --help`` display
893 - Fix hardcoded Xen kernel path
894 - Fix grow-disk handling of invalid units
895 - Update synopsis for ``gnt-cluster repair-disk-sizes``
896 - Accept both PUT and POST in noded (makes future upgrade to 2.6 easier)
902 *(Released Thu, 12 Apr 2012)*
904 Incompatible/important changes and bugfixes
905 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
907 - The default of the ``/2/instances/[instance_name]/rename`` RAPI
908 resource's ``ip_check`` parameter changed from ``True`` to ``False``
909 to match the underlying LUXI interface.
910 - The ``/2/nodes/[node_name]/evacuate`` RAPI resource was changed to use
911 body parameters, see :doc:`RAPI documentation <rapi>`. The server does
912 not maintain backwards-compatibility as the underlying operation
913 changed in an incompatible way. The RAPI client can talk to old
914 servers, but it needs to be told so as the return value changed.
915 - When creating file-based instances via RAPI, the ``file_driver``
916 parameter no longer defaults to ``loop`` and must be specified.
917 - The deprecated ``bridge`` NIC parameter is no longer supported. Use
919 - Support for the undocumented and deprecated RAPI instance creation
920 request format version 0 has been dropped. Use version 1, supported
921 since Ganeti 2.1.3 and :doc:`documented <rapi>`, instead.
922 - Pyparsing 1.4.6 or above is required, see :doc:`installation
923 documentation <install>`.
924 - The "cluster-verify" hooks are now executed per group by the
925 ``OP_CLUSTER_VERIFY_GROUP`` opcode. This maintains the same behavior
926 if you just run ``gnt-cluster verify``, which generates one opcode per
928 - The environment as passed to the OS scripts is cleared, and thus no
929 environment variables defined in the node daemon's environment will be
930 inherited by the scripts.
931 - The :doc:`iallocator <iallocator>` mode ``multi-evacuate`` has been
933 - :doc:`New iallocator modes <design-multi-reloc>` have been added to
934 support operations involving multiple node groups.
935 - Offline nodes are ignored when failing over an instance.
936 - Support for KVM version 1.0, which changed the version reporting format
938 - TCP/IP ports used by DRBD disks are returned to a pool upon instance
940 - ``Makefile`` is now compatible with Automake 1.11.2
941 - Includes all bugfixes made in the 2.4 series
946 - The ganeti-htools project has been merged into the ganeti-core source
947 tree and will be built as part of Ganeti (see :doc:`install-quick`).
948 - Implemented support for :doc:`shared storage <design-shared-storage>`.
949 - Add support for disks larger than 2 TB in ``lvmstrap`` by supporting
950 GPT-style partition tables (requires `parted
951 <http://www.gnu.org/s/parted/>`_).
952 - Added support for floppy drive and 2nd CD-ROM drive in KVM hypervisor.
953 - Allowed adding tags on instance creation.
954 - Export instance tags to hooks (``INSTANCE_TAGS``, see :doc:`hooks`)
955 - Allow instances to be started in a paused state, enabling the user to
956 see the complete console output on boot using the console.
957 - Added new hypervisor flag to control default reboot behaviour
958 (``reboot_behavior``).
959 - Added support for KVM keymaps (hypervisor parameter ``keymap``).
960 - Improved out-of-band management support:
962 - Added ``gnt-node health`` command reporting the health status of
964 - Added ``gnt-node power`` command to manage power status of nodes.
965 - Added command for emergency power-off (EPO), ``gnt-cluster epo``.
967 - Instance migration can fall back to failover if instance is not
969 - Filters can be used when listing nodes, instances, groups and locks;
970 see :manpage:`ganeti(7)` manpage.
971 - Added post-execution status as variables to :doc:`hooks <hooks>`
973 - Instance tags are exported/imported together with the instance.
974 - When given an explicit job ID, ``gnt-job info`` will work for archived
976 - Jobs can define dependencies on other jobs (not yet supported via
977 RAPI or command line, but used by internal commands and usable via
980 - Lock monitor (``gnt-debug locks``) shows jobs waiting for
983 - Instance failover is now available as a RAPI resource
984 (``/2/instances/[instance_name]/failover``).
985 - ``gnt-instance info`` defaults to static information if primary node
987 - Opcodes have a new ``comment`` attribute.
988 - Added basic SPICE support to KVM hypervisor.
989 - ``tools/ganeti-listrunner`` allows passing of arguments to executable.
991 Node group improvements
992 ~~~~~~~~~~~~~~~~~~~~~~~
994 - ``gnt-cluster verify`` has been modified to check groups separately,
995 thereby improving performance.
996 - Node group support has been added to ``gnt-cluster verify-disks``,
997 which now operates per node group.
998 - Watcher has been changed to work better with node groups.
1000 - One process and state file per node group.
1001 - Slow watcher in one group doesn't block other group's watcher.
1003 - Added new command, ``gnt-group evacuate``, to move all instances in a
1004 node group to other groups.
1005 - Added ``gnt-instance change-group`` to move an instance to another
1007 - ``gnt-cluster command`` and ``gnt-cluster copyfile`` now support
1008 per-group operations.
1009 - Node groups can be tagged.
1010 - Some operations switch from an exclusive to a shared lock as soon as
1012 - Instance's primary and secondary nodes' groups are now available as
1013 query fields (``pnode.group``, ``pnode.group.uuid``, ``snodes.group``
1014 and ``snodes.group.uuid``).
1019 - Numerous updates to documentation and manpages.
1021 - :doc:`RAPI <rapi>` documentation now has detailed parameter
1023 - Some opcode/job results are now also documented, see :doc:`RAPI
1026 - A lockset's internal lock is now also visible in lock monitor.
1027 - Log messages from job queue workers now contain information about the
1028 opcode they're processing.
1029 - ``gnt-instance console`` no longer requires the instance lock.
1030 - A short delay when waiting for job changes reduces the number of LUXI
1031 requests significantly.
1032 - DRBD metadata volumes are overwritten with zeros during disk creation.
1033 - Out-of-band commands no longer acquire the cluster lock in exclusive
1035 - ``devel/upload`` now uses correct permissions for directories.
1041 *(Released Fri, 23 Mar 2012)*
1043 This was the sixth release candidate of the 2.5 series.
1049 *(Released Mon, 9 Jan 2012)*
1051 This was the fifth release candidate of the 2.5 series.
1057 *(Released Thu, 27 Oct 2011)*
1059 This was the fourth release candidate of the 2.5 series.
1065 *(Released Wed, 26 Oct 2011)*
1067 This was the third release candidate of the 2.5 series.
1073 *(Released Tue, 18 Oct 2011)*
1075 This was the second release candidate of the 2.5 series.
1081 *(Released Tue, 4 Oct 2011)*
1083 This was the first release candidate of the 2.5 series.
1089 *(Released Wed, 31 Aug 2011)*
1091 This was the third beta release of the 2.5 series.
1097 *(Released Mon, 22 Aug 2011)*
1099 This was the second beta release of the 2.5 series.
1105 *(Released Fri, 12 Aug 2011)*
1107 This was the first beta release of the 2.5 series.
1113 *(Released Thu, 27 Oct 2011)*
1115 - Fixed bug when parsing command line parameter values ending in
1117 - Fixed assertion error after unclean master shutdown
1118 - Disable HTTP client pool for RPC, significantly reducing memory usage
1120 - Fixed queue archive creation with wrong permissions
1126 *(Released Tue, 23 Aug 2011)*
1130 - Fixed documentation for importing with ``--src-dir`` option
1131 - Fixed a bug in ``ensure-dirs`` with queue/archive permissions
1132 - Fixed a parsing issue with DRBD 8.3.11 in the Linux kernel
1138 *(Released Fri, 5 Aug 2011)*
1140 Many bug-fixes and a few small features:
1142 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
1143 issues when you tried to add an instance with two MAC addresses in one
1145 - KVM: fixed per-instance stored UID value
1146 - KVM: configure bridged NICs at migration start
1147 - KVM: Fix a bug where instance will not start with never KVM versions
1149 - Added OS search path to ``gnt-cluster info``
1150 - Fixed an issue with ``file_storage_dir`` where you were forced to
1151 provide an absolute path, but the documentation states it is a
1152 relative path, the documentation was right
1153 - Added a new parameter to instance stop/start called ``--no-remember``
1154 that will make the state change to not be remembered
1155 - Implemented ``no_remember`` at RAPI level
1156 - Improved the documentation
1157 - Node evacuation: don't call IAllocator if node is already empty
1158 - Fixed bug in DRBD8 replace disks on current nodes
1159 - Fixed bug in recreate-disks for DRBD instances
1160 - Moved assertion checking locks in ``gnt-instance replace-disks``
1161 causing it to abort with not owning the right locks for some situation
1162 - Job queue: Fixed potential race condition when cancelling queued jobs
1163 - Fixed off-by-one bug in job serial generation
1164 - ``gnt-node volumes``: Fix instance names
1165 - Fixed aliases in bash completion
1166 - Fixed a bug in reopening log files after being sent a SIGHUP
1167 - Added a flag to burnin to allow specifying VCPU count
1168 - Bugfixes to non-root Ganeti configuration
1174 *(Released Thu, 12 May 2011)*
1176 Many bug-fixes and a few new small features:
1178 - Fixed a bug related to log opening failures
1179 - Fixed a bug in instance listing with orphan instances
1180 - Fixed a bug which prevented resetting the cluster-level node parameter
1181 ``oob_program`` to the default
1182 - Many fixes related to the ``cluster-merge`` tool
1183 - Fixed a race condition in the lock monitor, which caused failures
1184 during (at least) creation of many instances in parallel
1185 - Improved output for gnt-job info
1186 - Removed the quiet flag on some ssh calls which prevented debugging
1188 - Improved the N+1 failure messages in cluster verify by actually
1189 showing the memory values (needed and available)
1190 - Increased lock attempt timeouts so that when executing long operations
1191 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
1192 too early and thus prevent the use of the 'fair' mechanism
1193 - Changed instance query data (``gnt-instance info``) to not acquire
1194 locks unless needed, thus allowing its use on locked instance if only
1195 static information is asked for
1196 - Improved behaviour with filesystems that do not support rename on an
1198 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
1199 kept locks on all nodes during the wipe, which is unneeded
1200 - Fixed ``gnt-watcher`` handling of errors during hooks execution
1201 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
1202 10GiB) which caused the wipe to fail right at the end in some cases
1203 - Fixed master IP activation when doing master failover with no-voting
1204 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
1205 the master node itself
1206 - Fixed potential data-loss in under disk full conditions, where Ganeti
1207 wouldn't check correctly the return code and would consider
1208 partially-written files 'correct'
1209 - Fixed bug related to multiple VGs and DRBD disk replacing
1210 - Added new disk parameter ``metavg`` that allows placement of the meta
1211 device for DRBD in a different volume group
1212 - Fixed error handling in the node daemon when the system libc doesn't
1213 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
1214 - Fixed lock release during replace-disks, which kept cluster-wide locks
1215 when doing disk replaces with an iallocator script
1216 - Added check for missing bridges in cluster verify
1217 - Handle EPIPE errors while writing to the terminal better, so that
1218 piping the output to e.g. ``less`` doesn't cause a backtrace
1219 - Fixed rare case where a ^C during Luxi calls could have been
1220 interpreted as server errors, instead of simply terminating
1221 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
1223 - Added a few more parameters to the KVM hypervisor, allowing a second
1224 CDROM, custom disk type for CDROMs and a floppy image
1225 - Removed redundant message in instance rename when the name is given
1227 - Added option to ``gnt-instance recreate-disks`` to allow creating the
1228 disks on new nodes, allowing recreation when the original instance
1229 nodes are completely gone
1230 - Added option when converting disk templates to DRBD to skip waiting
1231 for the resync, in order to make the instance available sooner
1232 - Added two new variables to the OS scripts environment (containing the
1234 - Made the root_path and optional parameter for the xen-pvm hypervisor,
1235 to allow use of ``pvgrub`` as bootloader
1236 - Changed the instance memory modifications to only check out-of-memory
1237 conditions on memory increases, and turned the secondary node warnings
1238 into errors (they can still be overridden via ``--force``)
1239 - Fixed the handling of a corner case when the Python installation gets
1240 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
1241 to execute a command that doesn't exist
1242 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
1243 node of the instance returned failures during instance shutdown; this
1244 adds the option ``--ignore-consistency`` to gnt-instance move
1246 And as usual, various improvements to the error messages, documentation
1253 *(Released Wed, 09 Mar 2011)*
1255 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
1256 the RAPI users file if run twice (even with ``--dry-run``).
1258 The release fixes that bug (nothing else changed).
1264 *(Released Mon, 07 Mar 2011)*
1266 Final 2.4.0 release. Just a few small fixes:
1268 - Fixed RAPI node evacuate
1269 - Fixed the kvm-ifup script
1270 - Fixed internal error handling for special job cases
1271 - Updated man page to specify the escaping feature for options
1277 *(Released Mon, 28 Feb 2011)*
1279 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
1280 that this feature wiped the disks of the wrong instance, leading to loss
1285 - Fixed title of query field containing instance name
1286 - Expanded the glossary in the documentation
1287 - Fixed one unittest (internal issue)
1293 *(Released Mon, 21 Feb 2011)*
1295 A number of bug fixes plus just a couple functionality changes.
1297 On the user-visible side, the ``gnt-* list`` command output has changed
1298 with respect to "special" field states. The current rc1 style of display
1299 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
1300 the default output mode special fields states are displayed as follows:
1302 - Offline resource: ``*``
1303 - Unavailable/not applicable: ``-``
1304 - Data missing (RPC failure): ``?``
1305 - Unknown field: ``??``
1307 Another user-visible change is the addition of ``--force-join`` to
1312 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
1313 - Fixed regression in RAPI/instance reinstall where all parameters were
1314 required (instead of optional)
1315 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
1316 - Fixed iallocator usage (offline nodes were not considered offline)
1317 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
1318 - Fixed hypervisor and OS parameter validation with respect to
1319 non-vm_capable nodes
1320 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
1322 - Fixed ``tools/listrunner`` with respect to agent-based usage
1328 *(Released Fri, 4 Feb 2011)*
1330 Many changes and fixes since the beta1 release. While there were some
1331 internal changes, the code has been mostly stabilised for the RC
1334 Note: the dumb allocator was removed in this release, as it was not kept
1335 up-to-date with the IAllocator protocol changes. It is recommended to
1336 use the ``hail`` command from the ganeti-htools package.
1338 Note: the 2.4 and up versions of Ganeti are not compatible with the
1339 0.2.x branch of ganeti-htools. You need to upgrade to
1340 ganeti-htools-0.3.0 (or later).
1342 Regressions fixed from 2.3
1343 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1345 - Fixed the ``gnt-cluster verify-disks`` command
1346 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
1348 - Fixed disk adoption breakage
1349 - Fixed wrong headers in instance listing for field aliases
1354 - Fixed corner case in KVM handling of NICs
1355 - Fixed many cases of wrong handling of non-vm_capable nodes
1356 - Fixed a bug where a missing instance symlink was not possible to
1357 recreate with any ``gnt-*`` command (now ``gnt-instance
1358 activate-disks`` does it)
1359 - Fixed the volume group name as reported by ``gnt-cluster
1361 - Increased timeouts for the import-export code, hopefully leading to
1362 fewer aborts due network or instance timeouts
1363 - Fixed bug in ``gnt-node list-storage``
1364 - Fixed bug where not all daemons were started on cluster
1365 initialisation, but only at the first watcher run
1366 - Fixed many bugs in the OOB implementation
1367 - Fixed watcher behaviour in presence of instances with offline
1369 - Fixed instance list output for instances running on the wrong node
1370 - a few fixes to the cluster-merge tool, but it still cannot merge
1371 multi-node groups (currently it is not recommended to use this tool)
1377 - Improved network configuration for the KVM hypervisor
1378 - Added e1000 as a supported NIC for Xen-HVM
1379 - Improved the lvmstrap tool to also be able to use partitions, as
1380 opposed to full disks
1381 - Improved speed of disk wiping (the cluster parameter
1382 ``prealloc_wipe_disks``, so that it has a low impact on the total time
1383 of instance creations
1384 - Added documentation for the OS parameters
1385 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
1386 hypervisor is not responding
1387 - Added display of blacklisted and hidden OS information in
1388 ``gnt-cluster info``
1389 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
1390 NIC and node parameters, which might create problems with currently
1391 invalid (but undetected) configuration files, but prevents validation
1392 failures when unrelated parameters are modified
1393 - Changed cluster initialisation to wait for the master daemon to become
1395 - Expanded the RAPI interface:
1397 - Added config redistribution resource
1398 - Added activation/deactivation of instance disks
1399 - Added export of console information
1401 - Implemented log file reopening on SIGHUP, which allows using
1402 logrotate(8) for the Ganeti log files
1403 - Added a basic OOB helper script as an example
1409 *(Released Fri, 14 Jan 2011)*
1414 - Fixed timezone issues when formatting timestamps
1415 - Added support for node groups, available via ``gnt-group`` and other
1417 - Added out-of-band framework and management, see :doc:`design
1418 document <design-oob>`
1419 - Removed support for roman numbers from ``gnt-node list`` and
1420 ``gnt-instance list``.
1421 - Allowed modification of master network interface via ``gnt-cluster
1422 modify --master-netdev``
1423 - Accept offline secondaries while shutting down instance disks
1424 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
1425 - Added support for multiple LVM volume groups
1426 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
1428 - Added commands to list available fields:
1430 - ``gnt-node list-fields``
1431 - ``gnt-group list-fields``
1432 - ``gnt-instance list-fields``
1434 - Updated documentation and man pages
1439 - Moved ``rapi_users`` file into separate directory, now named
1440 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
1442 - Added new tool for running commands on many machines,
1443 ``tools/ganeti-listrunner``
1444 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
1445 improving the ``gnt-instance console`` output
1446 - Allowed customisation of disk index separator at ``configure`` time
1447 - Export node group allocation policy to :doc:`iallocator <iallocator>`
1448 - Added support for non-partitioned md disks in ``lvmstrap``
1449 - Added script to gracefully power off KVM instances
1450 - Split ``utils`` module into smaller parts
1451 - Changed query operations to return more detailed information, e.g.
1452 whether an information is unavailable due to an offline node. To use
1453 this new functionality, the LUXI call ``Query`` must be used. Field
1454 information is now stored by the master daemon and can be retrieved
1455 using ``QueryFields``. Instances, nodes and groups can also be queried
1456 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
1457 exposed via RAPI). The following commands make use of this
1458 infrastructure change:
1460 - ``gnt-group list``
1461 - ``gnt-group list-fields``
1463 - ``gnt-node list-fields``
1464 - ``gnt-instance list``
1465 - ``gnt-instance list-fields``
1466 - ``gnt-debug locks``
1471 - New RAPI resources (see :doc:`rapi`):
1475 - ``/2/groups/[group_name]``
1476 - ``/2/groups/[group_name]/assign-nodes``
1477 - ``/2/groups/[group_name]/modify``
1478 - ``/2/groups/[group_name]/rename``
1479 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
1483 - Implemented ``no_install`` for instance creation
1484 - Implemented OS parameters for instance reinstallation, allowing
1485 use of special settings on reinstallation (e.g. for preserving data)
1490 - Added IPv6 support in import/export
1491 - Pause DRBD synchronization while wiping disks on instance creation
1492 - Updated unittests and QA scripts
1493 - Improved network parameters passed to KVM
1494 - Converted man pages from docbook to reStructuredText
1500 *(Released Mon, 20 Dec 2010)*
1502 Released version 2.3.1~rc1 without any changes.
1508 *(Released Wed, 1 Dec 2010)*
1510 - impexpd: Disable OpenSSL compression in socat if possible (backport
1511 from master, commit e90739d625b, see :doc:`installation guide
1512 <install-quick>` for details)
1513 - Changed unittest coverage report to exclude test scripts
1514 - Added script to check version format
1520 *(Released Wed, 1 Dec 2010)*
1522 Released version 2.3.0~rc1 without any changes.
1528 *(Released Fri, 19 Nov 2010)*
1530 A number of bugfixes and documentation updates:
1532 - Update ganeti-os-interface documentation
1533 - Fixed a bug related to duplicate MACs or similar items which should be
1535 - Fix breakage in OS state modify
1536 - Reinstall instance: disallow offline secondaries (fixes bug related to
1537 OS changing but reinstall failing)
1538 - plus all the other fixes between 2.2.1 and 2.2.2
1544 *(Released Tue, 2 Nov 2010)*
1546 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
1547 - Fixed master failover race with watcher
1548 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
1550 - Accept previously stopped instance for export with instance removal
1551 - Simplify and extend the environment variables for instance OS scripts
1552 - Added new node flags, ``master_capable`` and ``vm_capable``
1553 - Added optional instance disk wiping prior during allocation. This is a
1554 cluster-wide option and can be set/modified using
1555 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
1556 - Added IPv6 support, see :doc:`design document <design-2.3>` and
1557 :doc:`install-quick`
1558 - Added a new watcher option (``--ignore-pause``)
1559 - Added option to ignore offline node on instance start/stop
1560 (``--ignore-offline``)
1561 - Allow overriding OS parameters with ``gnt-instance reinstall``
1562 - Added ability to change node's secondary IP address using ``gnt-node
1564 - Implemented privilege separation for all daemons except
1565 ``ganeti-noded``, see ``configure`` options
1566 - Complain if an instance's disk is marked faulty in ``gnt-cluster
1568 - Implemented job priorities (see ``ganeti(7)`` manpage)
1569 - Ignore failures while shutting down instances during failover from
1571 - Exit daemon's bootstrap process only once daemon is ready
1572 - Export more information via ``LUInstanceQuery``/remote API
1573 - Improved documentation, QA and unittests
1574 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
1575 restart if the file was created or changed
1576 - Added LUXI protocol version sent with each request and response,
1577 allowing detection of server/client mismatches
1578 - Moved the Python scripts among gnt-* and ganeti-* into modules
1579 - Moved all code related to setting up SSH to an external script,
1581 - Infrastructure changes for node group support in future versions
1587 *(Released Fri, 19 Nov 2010)*
1589 A few small bugs fixed, and some improvements to the build system:
1591 - Fix documentation regarding conversion to drbd
1592 - Fix validation of parameters in cluster modify (``gnt-cluster modify
1594 - Fix error handling in node modify with multiple changes
1595 - Allow remote imports without checked names
1601 *(Released Tue, 19 Oct 2010)*
1603 - Disable SSL session ID cache in RPC client
1609 *(Released Thu, 14 Oct 2010)*
1611 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
1612 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
1614 - Fix problems with interaction between Curl and Python's HTTP server,
1615 resulting in increased speed in many RPC calls
1616 - Improve our release script to prevent breakage with older aclocal and
1623 *(Released Thu, 7 Oct 2010)*
1625 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
1626 value retrieved from master
1627 - Added support for blacklisted or hidden OS definitions
1628 - Added simple lock monitor (accessible via (``gnt-debug locks``)
1629 - Added support for -mem-path in KVM hypervisor abstraction layer
1630 - Allow overriding instance parameters in tool for inter-cluster
1631 instance moves (``tools/move-instance``)
1632 - Improved opcode summaries (e.g. in ``gnt-job list``)
1633 - Improve consistency of OS listing by sorting it
1634 - Documentation updates
1640 *(Released Fri, 8 Oct 2010)*
1642 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
1648 *(Released Mon, 4 Oct 2010)*
1650 - Fixed regression in ``gnt-instance rename``
1656 *(Released Wed, 22 Sep 2010)*
1658 - Fixed OS_VARIANT variable for OS scripts
1659 - Fixed cluster tag operations via RAPI
1660 - Made ``setup-ssh`` exit with non-zero code if an error occurred
1661 - Disabled RAPI CA checks in watcher
1667 *(Released Mon, 23 Aug 2010)*
1669 - Support DRBD versions of the format "a.b.c.d"
1671 - Re-introduce support for usage from multiple threads in RAPI client
1672 - Instance renames and modify via RAPI
1673 - Work around race condition between processing and archival in job
1675 - Mark opcodes following failed one as failed, too
1676 - Job field ``lock_status`` was removed due to difficulties making it
1677 work with the changed job queue in Ganeti 2.2; a better way to monitor
1678 locks is expected for a later 2.2.x release
1679 - Fixed dry-run behaviour with many commands
1680 - Support ``ssh-agent`` again when adding nodes
1681 - Many additional bugfixes
1687 *(Released Fri, 30 Jul 2010)*
1689 Important change: the internal RPC mechanism between Ganeti nodes has
1690 changed from using a home-grown http library (based on the Python base
1691 libraries) to use the PycURL library. This requires that PycURL is
1692 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
1693 against GnuTLS by default. cURL's support for GnuTLS had known issues
1694 before cURL 7.21.0 and we recommend using the latest cURL release or
1695 linking against OpenSSL. Most other distributions already link PycURL
1696 and cURL against OpenSSL. The command::
1698 python -c 'import pycurl; print pycurl.version'
1700 can be used to determine the libraries PycURL and cURL are linked
1703 Other significant changes:
1705 - Rewrote much of the internals of the job queue, in order to achieve
1706 better parallelism; this decouples job query operations from the job
1707 processing, and it should allow much nicer behaviour of the master
1708 daemon under load, and it also has uncovered some long-standing bugs
1709 related to the job serialisation (now fixed)
1710 - Added a default iallocator setting to the cluster parameters,
1711 eliminating the need to always pass nodes or an iallocator for
1712 operations that require selection of new node(s)
1713 - Added experimental support for the LXC virtualization method
1714 - Added support for OS parameters, which allows the installation of
1715 instances to pass parameter to OS scripts in order to customise the
1717 - Added a hypervisor parameter controlling the migration type (live or
1718 non-live), since hypervisors have various levels of reliability; this
1719 has renamed the 'live' parameter to 'mode'
1720 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
1721 logical volumes, meaning that cluster verify will ignore them and not
1722 flag their presence as errors
1723 - The watcher will now reset the error count for failed instances after
1724 8 hours, thus allowing self-healing if the problem that caused the
1725 instances to be down/fail to start has cleared in the meantime
1726 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
1727 check for, and warn, if the drbd module parameter ``usermode_helper``
1728 is not consistent with the cluster-wide setting; this is needed to
1729 make diagnose easier of failed drbd creations
1730 - Started adding base IPv6 support, but this is not yet
1731 enabled/available for use
1732 - Rename operations (cluster, instance) will now return the new name,
1733 which is especially useful if a short name was passed in
1734 - Added support for instance migration in RAPI
1735 - Added a tool to pre-configure nodes for the SSH setup, before joining
1736 them to the cluster; this will allow in the future a simplified model
1737 for node joining (but not yet fully enabled in 2.2); this needs the
1738 paramiko python library
1739 - Fixed handling of name-resolving errors
1740 - Fixed consistency of job results on the error path
1741 - Fixed master-failover race condition when executed multiple times in
1743 - Fixed many bugs related to the job queue (mostly introduced during the
1744 2.2 development cycle, so not all are impacting 2.1)
1745 - Fixed instance migration with missing disk symlinks
1746 - Fixed handling of unknown jobs in ``gnt-job archive``
1747 - And many other small fixes/improvements
1751 - Enhanced both the unittest and the QA coverage
1752 - Switched the opcode validation to a generic model, and extended the
1753 validation to all opcode parameters
1754 - Changed more parts of the code that write shell scripts to use the
1756 - Switched the master daemon to use the asyncore library for the Luxi
1763 *(Released Thu, 17 Jun 2010)*
1765 - Added tool (``move-instance``) and infrastructure to move instances
1766 between separate clusters (see :doc:`separate documentation
1767 <move-instance>` and :doc:`design document <design-2.2>`)
1768 - Added per-request RPC timeout
1769 - RAPI now requires a Content-Type header for requests with a body (e.g.
1770 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
1771 :rfc:`2616` (HTTP/1.1), section 7.2.1)
1772 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
1774 - Implemented initial support for running Ganeti daemons as separate
1775 users, see configure-time flags ``--with-user-prefix`` and
1776 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
1778 - Instances can be removed after export (``gnt-backup export
1779 --remove-instance``)
1780 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
1781 key (instead of 1024 bit)
1782 - Added new cluster configuration file for cluster domain secret
1783 - Import/export now use SSL instead of SSH
1784 - Added support for showing estimated time when exporting an instance,
1785 see the ``ganeti-os-interface(7)`` manpage and look for
1792 *(Released Tue, 16 Nov 2010)*
1794 Some more bugfixes. Unless critical bugs occur, this will be the last
1797 - Fix case of MAC special-values
1798 - Fix mac checker regex
1799 - backend: Fix typo causing "out of range" error
1800 - Add missing --units in gnt-instance list man page
1806 *(Released Tue, 24 Aug 2010)*
1809 - Don't ignore secondary node silently on non-mirrored disk templates
1811 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
1812 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
1813 - Properly document the usb_mouse parameter
1814 - Fix path in ganeti-rapi(8) (issue 116)
1815 - Adjust error message when the ganeti user's .ssh directory is
1817 - Add same-node-check when changing the disk template to drbd
1823 *(Released Fri, 16 Jul 2010)*
1826 - Add an option to only select some reboot types during qa/burnin.
1827 (on some hypervisors consequent reboots are not supported)
1828 - Fix infrequent race condition in master failover. Sometimes the old
1829 master ip address would be still detected as up for a short time
1830 after it was removed, causing failover to fail.
1831 - Decrease mlockall warnings when the ctypes module is missing. On
1832 Python 2.4 we support running even if no ctypes module is installed,
1833 but we were too verbose about this issue.
1834 - Fix building on old distributions, on which man doesn't have a
1836 - Fix RAPI not to ignore the MAC address on instance creation
1837 - Implement the old instance creation format in the RAPI client.
1843 *(Released Thu, 01 Jul 2010)*
1845 A small bugfix release:
1846 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
1847 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
1848 a non-existing option
1849 - Fix instance create: the --force-variant option was ignored
1850 - Improve pylint 0.21 compatibility and warnings with Python 2.6
1851 - Fix modify node storage with non-FQDN arguments
1852 - Fix RAPI client to authenticate under Python 2.6 when used
1853 for more than 5 requests needing authentication
1854 - Fix gnt-instance modify -t (storage) giving a wrong error message
1855 when converting a non-shutdown drbd instance to plain
1861 *(Released Fri, 18 Jun 2010)*
1863 A small bugfix release:
1865 - Fix live migration of KVM instances started with older Ganeti
1866 versions which had fewer hypervisor parameters
1867 - Fix gnt-instance grow-disk on down instances
1868 - Fix an error-reporting bug during instance migration
1869 - Better checking of the ``--net`` and ``--disk`` values, to avoid
1870 silently ignoring broken ones
1871 - Fix an RPC error reporting bug affecting, for example, RAPI client
1873 - Fix bug triggered by different API version os-es on different nodes
1874 - Fix a bug in instance startup with custom hvparams: OS level
1875 parameters would fail to be applied.
1876 - Fix the RAPI client under Python 2.6 (but more work is needed to
1877 make it work completely well with OpenSSL)
1878 - Fix handling of errors when resolving names from DNS
1884 *(Released Thu, 3 Jun 2010)*
1886 A medium sized development cycle. Some new features, and some
1887 fixes/small improvements/cleanups.
1889 Significant features
1890 ~~~~~~~~~~~~~~~~~~~~
1892 The node deamon now tries to mlock itself into memory, unless the
1893 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
1894 its logs, and falls back to console logging. This allows emergency
1895 features such as ``gnt-node powercycle`` to work even in the event of a
1896 broken node disk (tested offlining the disk hosting the node's
1897 filesystem and dropping its memory caches; don't try this at home)
1899 KVM: add vhost-net acceleration support. It can be tested with a new
1900 enough version of the kernel and of qemu-kvm.
1902 KVM: Add instance chrooting feature. If you use privilege dropping for
1903 your VMs you can also now force them to chroot to an empty directory,
1904 before starting the emulated guest.
1906 KVM: Add maximum migration bandwith and maximum downtime tweaking
1907 support (requires a new-enough version of qemu-kvm).
1909 Cluster verify will now warn if the master node doesn't have the master
1910 ip configured on it.
1912 Add a new (incompatible) instance creation request format to RAPI which
1913 supports all parameters (previously only a subset was supported, and it
1914 wasn't possible to extend the old format to accomodate all the new
1915 features. The old format is still supported, and a client can check for
1916 this feature, before using it, by checking for its presence in the
1917 ``features`` RAPI resource.
1919 Now with ancient latin support. Try it passing the ``--roman`` option to
1920 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
1921 (requires the python-roman module to be installed, in order to work).
1926 As usual many internal code refactorings, documentation updates, and
1929 - Lots of improvements and cleanups to the experimental Remote API
1930 (RAPI) client library.
1931 - A new unit test suite for the core daemon libraries.
1932 - A fix to creating missing directories makes sure the umask is not
1933 applied anymore. This enforces the same directory permissions
1935 - Better handling terminating daemons with ctrl+c (used when running
1936 them in debugging mode).
1937 - Fix a race condition in live migrating a KVM instance, when stat()
1938 on the old proc status file returned EINVAL, which is an unexpected
1940 - Fixed manpage checking with newer man and utf-8 charachters. But now
1941 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
1947 *(Released Fri, 7 May 2010)*
1949 Fix a bug which prevented untagged KVM instances from starting.
1955 *(Released Fri, 7 May 2010)*
1957 Another release with a long development cycle, during which many
1958 different features were added.
1960 Significant features
1961 ~~~~~~~~~~~~~~~~~~~~
1963 The KVM hypervisor now can run the individual instances as non-root, to
1964 reduce the impact of a VM being hijacked due to bugs in the
1965 hypervisor. It is possible to run all instances as a single (non-root)
1966 user, to manually specify a user for each instance, or to dynamically
1967 allocate a user out of a cluster-wide pool to each instance, with the
1968 guarantee that no two instances will run under the same user ID on any
1971 An experimental RAPI client library, that can be used standalone
1972 (without the other Ganeti libraries), is provided in the source tree as
1973 ``lib/rapi/client.py``. Note this client might change its interface in
1974 the future, as we iterate on its capabilities.
1976 A new command, ``gnt-cluster renew-crypto`` has been added to easily
1977 replace the cluster's certificates and crypto keys. This might help in
1978 case they have been compromised, or have simply expired.
1980 A new disk option for instance creation has been added that allows one
1981 to "adopt" currently existing logical volumes, with data
1982 preservation. This should allow easier migration to Ganeti from
1983 unmanaged (or managed via other software) instances.
1985 Another disk improvement is the possibility to convert between redundant
1986 (DRBD) and plain (LVM) disk configuration for an instance. This should
1987 allow better scalability (starting with one node and growing the
1988 cluster, or shrinking a two-node cluster to one node).
1990 A new feature that could help with automated node failovers has been
1991 implemented: if a node sees itself as offline (by querying the master
1992 candidates), it will try to shutdown (hard) all instances and any active
1993 DRBD devices. This reduces the risk of duplicate instances if an
1994 external script automatically failovers the instances on such nodes. To
1995 enable this, the cluster parameter ``maintain_node_health`` should be
1996 enabled; in the future this option (per the name) will enable other
1997 automatic maintenance features.
1999 Instance export/import now will reuse the original instance
2000 specifications for all parameters; that means exporting an instance,
2001 deleting it and the importing it back should give an almost identical
2002 instance. Note that the default import behaviour has changed from
2003 before, where it created only one NIC; now it recreates the original
2006 Cluster verify has added a few new checks: SSL certificates validity,
2007 /etc/hosts consistency across the cluster, etc.
2012 As usual, many internal changes were done, documentation fixes,
2015 - Fixed cluster initialization with disabled cluster storage (regression
2016 introduced in 2.1.1)
2017 - File-based storage supports growing the disks
2018 - Fixed behaviour of node role changes
2019 - Fixed cluster verify for some corner cases, plus a general rewrite of
2020 cluster verify to allow future extension with more checks
2021 - Fixed log spamming by watcher and node daemon (regression introduced
2023 - Fixed possible validation issues when changing the list of enabled
2025 - Fixed cleanup of /etc/hosts during node removal
2026 - Fixed RAPI response for invalid methods
2027 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
2028 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
2029 from ide disks, etc.)
2030 - Allow OS changes without re-installation (to record a changed OS
2031 outside of Ganeti, or to allow OS renames)
2032 - Allow instance creation without OS installation (useful for example if
2033 the OS will be installed manually, or restored from a backup not in
2035 - Implemented option to make cluster ``copyfile`` use the replication
2037 - Added list of enabled hypervisors to ssconf (possibly useful for
2039 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
2041 - A partial form of node re-IP is possible via node readd, which now
2042 allows changed node primary IP
2043 - Command line utilities now show an informational message if the job is
2045 - The logs of the master daemon now show the PID/UID/GID of the
2052 *(Released Fri, 12 Mar 2010)*
2054 During the 2.1.0 long release candidate cycle, a lot of improvements and
2055 changes have accumulated with were released later as 2.1.1.
2060 The node evacuate command (``gnt-node evacuate``) was significantly
2061 rewritten, and as such the IAllocator protocol was changed - a new
2062 request type has been added. This unfortunate change during a stable
2063 series is designed to improve performance of node evacuations; on
2064 clusters with more than about five nodes and which are well-balanced,
2065 evacuation should proceed in parallel for all instances of the node
2066 being evacuated. As such, any existing IAllocator scripts need to be
2067 updated, otherwise the above command will fail due to the unknown
2068 request. The provided "dumb" allocator has not been updated; but the
2069 ganeti-htools package supports the new protocol since version 0.2.4.
2071 Another important change is increased validation of node and instance
2072 names. This might create problems in special cases, if invalid host
2073 names are being used.
2075 Also, a new layer of hypervisor parameters has been added, that sits at
2076 OS level between the cluster defaults and the instance ones. This allows
2077 customisation of virtualization parameters depending on the installed
2078 OS. For example instances with OS 'X' may have a different KVM kernel
2079 (or any other parameter) than the cluster defaults. This is intended to
2080 help managing a multiple OSes on the same cluster, without manual
2081 modification of each instance's parameters.
2083 A tool for merging clusters, ``cluster-merge``, has been added in the
2084 tools sub-directory.
2089 - Improved the int/float conversions that should make the code more
2090 robust in face of errors from the node daemons
2091 - Fixed the remove node code in case of internal configuration errors
2092 - Fixed the node daemon behaviour in face of inconsistent queue
2093 directory (e.g. read-only file-system where we can't open the files
2095 - Fixed the behaviour of gnt-node modify for master candidate demotion;
2096 now it either aborts cleanly or, if given the new "auto_promote"
2097 parameter, will automatically promote other nodes as needed
2098 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
2099 completely prevent Ganeti from working
2100 - Fixed bug for instance export when not all disks were successfully
2102 - Fixed behaviour of node add when the new node is slow in starting up
2104 - Fixed handling of signals in the LUXI client, which should improve
2105 behaviour of command-line scripts
2106 - Added checks for invalid node/instance names in the configuration (now
2107 flagged during cluster verify)
2108 - Fixed watcher behaviour for disk activation errors
2109 - Fixed two potentially endless loops in http library, which led to the
2110 RAPI daemon hanging and consuming 100% CPU in some cases
2111 - Fixed bug in RAPI daemon related to hashed passwords
2112 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
2114 - Enhanced compatibility with non-Debian OSes, but not using absolute
2115 path in some commands and allowing customisation of the ssh
2116 configuration directory
2117 - Fixed possible future issue with new Python versions by abiding to the
2118 proper use of ``__slots__`` attribute on classes
2119 - Added checks that should prevent directory traversal attacks
2120 - Many documentation fixes based on feedback from users
2125 - Added an "early_release" more for instance replace disks and node
2126 evacuate, where we release locks earlier and thus allow higher
2127 parallelism within the cluster
2128 - Added watcher hooks, intended to allow the watcher to restart other
2129 daemons (e.g. from the ganeti-nbma project), but they can be used of
2130 course for any other purpose
2131 - Added a compile-time disable for DRBD barriers, to increase
2132 performance if the administrator trusts the power supply or the
2133 storage system to not lose writes
2134 - Added the option of using syslog for logging instead of, or in
2135 addition to, Ganeti's own log files
2136 - Removed boot restriction for paravirtual NICs for KVM, recent versions
2137 can indeed boot from a paravirtual NIC
2138 - Added a generic debug level for many operations; while this is not
2139 used widely yet, it allows one to pass the debug value all the way to
2141 - Enhanced the hooks environment for instance moves (failovers,
2142 migrations) where the primary/secondary nodes changed during the
2143 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
2144 - Enhanced data validations for many user-supplied values; one important
2145 item is the restrictions imposed on instance and node names, which
2146 might reject some (invalid) host names
2147 - Add a configure-time option to disable file-based storage, if it's not
2148 needed; this allows greater security separation between the master
2149 node and the other nodes from the point of view of the inter-node RPC
2151 - Added user notification in interactive tools if job is waiting in the
2152 job queue or trying to acquire locks
2153 - Added log messages when a job is waiting for locks
2154 - Added filtering by node tags in instance operations which admit
2155 multiple instances (start, stop, reboot, reinstall)
2156 - Added a new tool for cluster mergers, ``cluster-merge``
2157 - Parameters from command line which are of the form ``a=b,c=d`` can now
2158 use backslash escapes to pass in values which contain commas,
2159 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
2161 - For KVM, the instance name is the first parameter passed to KVM, so
2162 that it's more visible in the process list
2168 *(Released Tue, 2 Mar 2010)*
2170 Ganeti 2.1 brings many improvements with it. Major changes:
2172 - Added infrastructure to ease automated disk repairs
2173 - Added new daemon to export configuration data in a cheaper way than
2174 using the remote API
2175 - Instance NICs can now be routed instead of being associated with a
2177 - Improved job locking logic to reduce impact of jobs acquiring multiple
2178 locks waiting for other long-running jobs
2180 In-depth implementation details can be found in the Ganeti 2.1 design
2186 - Added chroot hypervisor
2187 - Added more options to xen-hvm hypervisor (``kernel_path`` and
2189 - Added more options to xen-pvm hypervisor (``use_bootloader``,
2190 ``bootloader_path`` and ``bootloader_args``)
2191 - Added the ``use_localtime`` option for the xen-hvm and kvm
2192 hypervisors, and the default value for this has changed to false (in
2193 2.0 xen-hvm always enabled it)
2194 - Added luxi call to submit multiple jobs in one go
2195 - Added cluster initialization option to not modify ``/etc/hosts``
2197 - Added network interface parameters
2198 - Added dry run mode to some LUs
2199 - Added RAPI resources:
2201 - ``/2/instances/[instance_name]/info``
2202 - ``/2/instances/[instance_name]/replace-disks``
2203 - ``/2/nodes/[node_name]/evacuate``
2204 - ``/2/nodes/[node_name]/migrate``
2205 - ``/2/nodes/[node_name]/role``
2206 - ``/2/nodes/[node_name]/storage``
2207 - ``/2/nodes/[node_name]/storage/modify``
2208 - ``/2/nodes/[node_name]/storage/repair``
2210 - Added OpCodes to evacuate or migrate all instances on a node
2211 - Added new command to list storage elements on nodes (``gnt-node
2212 list-storage``) and modify them (``gnt-node modify-storage``)
2213 - Added new ssconf files with master candidate IP address
2214 (``ssconf_master_candidates_ips``), node primary IP address
2215 (``ssconf_node_primary_ips``) and node secondary IP address
2216 (``ssconf_node_secondary_ips``)
2217 - Added ``ganeti-confd`` and a client library to query the Ganeti
2218 configuration via UDP
2219 - Added ability to run hooks after cluster initialization and before
2221 - Added automatic mode for disk replace (``gnt-instance replace-disks
2223 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
2224 after catastrophic data-loss
2225 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
2227 - Added ``gnt-instance move`` command to move instances
2228 - Added ``gnt-cluster watcher`` command to control watcher
2229 - Added ``gnt-node powercycle`` command to powercycle nodes
2230 - Added new job status field ``lock_status``
2231 - Added parseable error codes to cluster verification (``gnt-cluster
2232 verify --error-codes``) and made output less verbose (use
2233 ``--verbose`` to restore previous behaviour)
2234 - Added UUIDs to the main config entities (cluster, nodes, instances)
2235 - Added support for OS variants
2236 - Added support for hashed passwords in the Ganeti remote API users file
2238 - Added option to specify maximum timeout on instance shutdown
2239 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
2240 - Added new helper script to start and stop Ganeti daemons
2241 (``daemon-util``), with the intent to reduce the work necessary to
2242 adjust Ganeti for non-Debian distributions and to start/stop daemons
2244 - Added more unittests
2245 - Fixed critical bug in ganeti-masterd startup
2246 - Removed the configure-time ``kvm-migration-port`` parameter, this is
2247 now customisable at the cluster level for both the KVM and Xen
2248 hypervisors using the new ``migration_port`` parameter
2249 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
2250 reinstalling an instance
2251 - Allowed ``@`` in tag names
2252 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
2253 - Many documentation updates
2254 - Distribute hypervisor files on ``gnt-cluster redist-conf``
2255 - ``gnt-instance reinstall`` can now reinstall multiple instances
2256 - Updated many command line parameters
2257 - Introduced new OS API version 15
2258 - No longer support a default hypervisor
2259 - Treat virtual LVs as inexistent
2260 - Improved job locking logic to reduce lock contention
2261 - Match instance and node names case insensitively
2262 - Reimplemented bash completion script to be more complete
2269 *(Released Thu, 4 Feb 2010)*
2271 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
2272 - Fix a string formatting bug
2273 - Improve safety of the code in some error paths
2274 - Improve data validation in the master of values returned from nodes
2280 *(Released Thu, 17 Dec 2009)*
2282 - Fix security issue due to missing validation of iallocator names; this
2283 allows local and remote execution of arbitrary executables
2284 - Fix failure of gnt-node list during instance removal
2285 - Ship the RAPI documentation in the archive
2291 *(Released Wed, 30 Sep 2009)*
2293 - Fixed many wrong messages
2294 - Fixed a few bugs related to the locking library
2295 - Fixed MAC checking at instance creation time
2296 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
2297 - Fixed a few issues related to signal handling in both daemons and
2299 - Fixed the example startup script provided
2300 - Fixed insserv dependencies in the example startup script (patch from
2302 - Fixed handling of drained nodes in the iallocator framework
2303 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
2305 - Fixed error related to invalid job IDs in job polling
2306 - Fixed job/opcode persistence on unclean master shutdown
2307 - Fixed handling of partial job processing after unclean master
2309 - Fixed error reporting from LUs, previously all errors were converted
2310 into execution errors
2311 - Fixed error reporting from burnin
2312 - Decreased significantly the memory usage of the job queue
2313 - Optimised slightly multi-job submission
2314 - Optimised slightly opcode loading
2315 - Backported the multi-job submit framework from the development
2316 branch; multi-instance start and stop should be faster
2317 - Added script to clean archived jobs after 21 days; this will reduce
2318 the size of the queue directory
2319 - Added some extra checks in disk size tracking
2320 - Added an example ethers hook script
2321 - Added a cluster parameter that prevents Ganeti from modifying of
2323 - Added more node information to RAPI responses
2324 - Added a ``gnt-job watch`` command that allows following the ouput of a
2326 - Added a bind-address option to ganeti-rapi
2327 - Added more checks to the configuration verify
2328 - Enhanced the burnin script such that some operations can be retried
2330 - Converted instance reinstall to multi-instance model
2336 *(Released Fri, 7 Aug 2009)*
2338 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
2339 to allow using the pre-2.0.2 behaviour in activation, if any existing
2340 instances have mismatched disk sizes in the configuration
2341 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
2342 any configuration mismatches for disk sizes
2343 - Added ``gnt-master cluste-failover --no-voting`` to allow master
2344 failover to work on two-node clusters
2345 - Fixed the ``--net`` option of ``gnt-backup import``, which was
2347 - Fixed detection of OS script errors in ``gnt-backup export``
2348 - Fixed exit code of ``gnt-backup export``
2354 *(Released Fri, 17 Jul 2009)*
2356 - Added experimental support for stripped logical volumes; this should
2357 enhance performance but comes with a higher complexity in the block
2358 device handling; stripping is only enabled when passing
2359 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
2360 affected even in the non-stripped mode
2361 - Improved resiliency against transient failures at the end of DRBD
2362 resyncs, and in general of DRBD resync checks
2363 - Fixed a couple of issues with exports and snapshot errors
2364 - Fixed a couple of issues in instance listing
2365 - Added display of the disk size in ``gnt-instance info``
2366 - Fixed checking for valid OSes in instance creation
2367 - Fixed handling of the "vcpus" parameter in instance listing and in
2368 general of invalid parameters
2369 - Fixed http server library, and thus RAPI, to handle invalid
2370 username/password combinations correctly; this means that now they
2371 report unauthorized for queries too, not only for modifications,
2372 allowing earlier detect of configuration problems
2373 - Added a new "role" node list field, equivalent to the master/master
2374 candidate/drained/offline flags combinations
2375 - Fixed cluster modify and changes of candidate pool size
2376 - Fixed cluster verify error messages for wrong files on regular nodes
2377 - Fixed a couple of issues with node demotion from master candidate role
2378 - Fixed node readd issues
2379 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
2380 - Added a new ``--no-voting`` option for masterfailover to fix failover
2381 on two-nodes clusters when the former master node is unreachable
2382 - Added instance reinstall over RAPI
2388 *(Released Tue, 16 Jun 2009)*
2390 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
2391 allow re-adding the start in single-user option (regression from 1.2)
2392 - the watcher writes the instance status to a file, to allow monitoring
2393 to report the instance status (from the master) based on cached
2394 results of the watcher's queries; while this can get stale if the
2395 watcher is being locked due to other work on the cluster, this is
2396 still an improvement
2397 - the watcher now also restarts the node daemon and the rapi daemon if
2399 - fixed the watcher to handle full and drained queue cases
2400 - hooks export more instance data in the environment, which helps if
2401 hook scripts need to take action based on the instance's properties
2402 (no longer need to query back into ganeti)
2403 - instance failovers when the instance is stopped do not check for free
2404 RAM, so that failing over a stopped instance is possible in low memory
2406 - rapi uses queries for tags instead of jobs (for less job traffic), and
2407 for cluster tags it won't talk to masterd at all but read them from
2409 - a couple of error handling fixes in RAPI
2410 - drbd handling: improved the error handling of inconsistent disks after
2411 resync to reduce the frequency of "there are some degraded disks for
2412 this instance" messages
2413 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
2414 error handling path called a wrong function name)
2420 *(Released Wed, 27 May 2009)*
2422 - no changes from rc5
2428 *(Released Wed, 20 May 2009)*
2430 - fix a couple of bugs (validation, argument checks)
2431 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
2432 - some small improvements to RAPI and IAllocator
2433 - make watcher automatically start the master daemon if down
2439 *(Released Mon, 27 Apr 2009)*
2441 - change the OS list to not require locks; this helps with big clusters
2442 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
2443 volume group is broken
2444 - ``gnt-instance info``, without any arguments, doesn't run for all
2445 instances anymore; either pass ``--all`` or pass the desired
2446 instances; this helps against mistakes on big clusters where listing
2447 the information for all instances takes a long time
2448 - miscellaneous doc and man pages fixes
2454 *(Released Wed, 8 Apr 2009)*
2456 - Change the internal locking model of some ``gnt-node`` commands, in
2457 order to reduce contention (and blocking of master daemon) when
2458 batching many creation/reinstall jobs
2459 - Fixes to Xen soft reboot
2460 - No longer build documentation at build time, instead distribute it in
2461 the archive, in order to reduce the need for the whole docbook/rst
2468 *(Released Fri, 27 Mar 2009)*
2470 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
2471 - Fix watcher startup sequence, improves the behaviour of busy clusters
2472 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
2473 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
2474 bind address and other places
2475 - Some documentation fixes and updates
2481 *(Released Mon, 2 Mar 2009)*
2483 - More documentation updates, now all docs should be more-or-less
2485 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
2487 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
2489 - Fix an issue related to $libdir/run/ganeti and cluster creation
2495 *(Released Thu, 19 Feb 2009)*
2497 - Xen PVM and KVM have switched the default value for the instance root
2498 disk to the first partition on the first drive, instead of the whole
2499 drive; this means that the OS installation scripts must be changed
2501 - Man pages have been updated
2502 - RAPI has been switched by default to HTTPS, and the exported functions
2503 should all work correctly
2504 - RAPI v1 has been removed
2505 - Many improvements to the KVM hypervisor
2506 - Block device errors are now better reported
2507 - Many other bugfixes and small improvements
2513 *(Released Mon, 26 Jan 2009)*
2515 - Version 2 is a general rewrite of the code and therefore the
2516 differences are too many to list, see the design document for 2.0 in
2517 the ``doc/`` subdirectory for more details
2518 - In this beta version there is not yet a migration path from 1.2 (there
2519 will be one in the final 2.0 release)
2520 - A few significant changes are:
2522 - all commands are executed by a daemon (``ganeti-masterd``) and the
2523 various ``gnt-*`` commands are just front-ends to it
2524 - all the commands are entered into, and executed from a job queue,
2525 see the ``gnt-job(8)`` manpage
2526 - the RAPI daemon supports read-write operations, secured by basic
2527 HTTP authentication on top of HTTPS
2528 - DRBD version 0.7 support has been removed, DRBD 8 is the only
2529 supported version (when migrating from Ganeti 1.2 to 2.0, you need
2530 to migrate to DRBD 8 first while still running Ganeti 1.2)
2531 - DRBD devices are using statically allocated minor numbers, which
2532 will be assigned to existing instances during the migration process
2533 - there is support for both Xen PVM and Xen HVM instances running on
2535 - KVM virtualization is supported too
2536 - file-based storage has been implemented, which means that it is
2537 possible to run the cluster without LVM and DRBD storage, for
2538 example using a shared filesystem exported from shared storage (and
2539 still have live migration)
2545 *(Released Tue, 13 Jan 2009)*
2547 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
2548 - Reuse the old instance mac address by default on instance import, if
2549 the instance name is the same.
2550 - Handle situations in which the node info rpc returns incomplete
2552 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
2553 - Improved version of batcher:
2555 - state file support
2556 - instance mac address support
2557 - support for HVM clusters/instances
2559 - Add an option to show the number of cpu sockets and nodes in
2561 - Support OSes that handle more than one version of the OS api (but do
2562 not change the current API in any other way)
2563 - Fix ``gnt-node migrate``
2564 - ``gnt-debug`` man page
2565 - Fixes various more typos and small issues
2566 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
2572 *(Released Wed, 24 Sep 2008)*
2574 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
2575 type of disk exported to fully virtualized instances.
2576 - provide access to the serial console of HVM instances
2577 - instance auto_balance flag, set by default. If turned off it will
2578 avoid warnings on cluster verify if there is not enough memory to fail
2579 over an instance. in the future it will prevent automatically failing
2580 it over when we will support that.
2581 - batcher tool for instance creation, see ``tools/README.batcher``
2582 - ``gnt-instance reinstall --select-os`` to interactively select a new
2583 operating system when reinstalling an instance.
2584 - when changing the memory amount on instance modify a check has been
2585 added that the instance will be able to start. also warnings are
2586 emitted if the instance will not be able to fail over, if auto_balance
2588 - documentation fixes
2589 - sync fields between ``gnt-instance list/modify/add/import``
2590 - fix a race condition in drbd when the sync speed was set after giving
2591 the device a remote peer.
2597 *(Released Tue, 22 Jul 2008)*
2599 - note: the allowed size and number of tags per object were reduced
2600 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
2601 - fixed twisted 8.x compatibility
2602 - fixed ``gnt-instance replace-disks`` with iallocator
2603 - add TCP keepalives on twisted connections to detect restarted nodes
2604 - disk increase support, see ``gnt-instance grow-disk``
2605 - implement bulk node/instance query for RAPI
2606 - add tags in node/instance listing (optional)
2607 - experimental migration (and live migration) support, read the man page
2608 for ``gnt-instance migrate``
2609 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
2610 has some small improvements in handling its state file
2616 *(Released Fri, 13 Jun 2008)*
2618 - Experimental readonly, REST-based remote API implementation;
2619 automatically started on master node, TCP port 5080, if enabled by
2620 ``--enable-rapi`` parameter to configure script.
2621 - Instance allocator support. Add and import instance accept a
2622 ``--iallocator`` parameter, and call that instance allocator to decide
2623 which node to use for the instance. The iallocator document describes
2624 what's expected from an allocator script.
2625 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
2626 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
2627 node is lost there is still enough memory to fail over the instances
2629 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
2630 ``gnt-cluster verify``, to check for site-specific compliance. All the
2631 hooks will run, and their output, if any, will be displayed. Any
2632 failing hook will make the verification return an error value.
2633 - ``gnt-cluster verify`` now checks that its peers are reachable on the
2634 primary and secondary interfaces
2635 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
2636 that is still declared as part of the cluster and has failed.
2637 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
2638 specifying output fields, that just adds the chosen fields to the
2640 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
2641 export from the filesystem.
2642 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
2643 have been added. Using them you can enable/disable acpi and pae
2644 support, and specify a path for a cd image to be exported to the
2645 instance. These parameters as the name suggest only work on HVM
2647 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
2648 PAE support will be set to the previously hardcoded values, but the
2649 (previously hardcoded) path to the CDROM ISO image will be unset and
2650 if required, needs to be set manually with ``gnt-instance modify``
2652 - The address to which an instance's VNC console is bound is now
2653 selectable per-instance, rather than being cluster wide. Of course
2654 this only applies to instances controlled via VNC, so currently just
2655 applies to HVM clusters.
2661 *(Released Mon, 18 Feb 2008)*
2663 - more tweaks to the disk activation code (especially helpful for DRBD)
2664 - change the default ``gnt-instance list`` output format, now there is
2665 one combined status field (see the manpage for the exact values this
2667 - some more fixes for the mac export to hooks change
2668 - make Ganeti not break with DRBD 8.2.x (which changed the version
2669 format in ``/proc/drbd``) (issue 24)
2670 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
2671 template, allowing migration from DRBD0.7+MD to DRBD8
2677 *(Released Wed, 30 Jan 2008)*
2679 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
2681 - add command aliases infrastructure and a few aliases
2682 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
2683 man pages and the ``--help`` option of ``gnt-node
2684 list``/``gnt-instance list``
2685 - fix ``gnt-backup list`` with down nodes (issue 21)
2686 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
2687 - fix the dist archive and add a check for including svn/git files in
2689 - some developer-related changes: improve the burnin and the QA suite,
2690 add an upload script for testing during development
2696 *(Released Wed, 16 Jan 2008)*
2698 - experimental HVM support, read the install document, section
2699 "Initializing the cluster"
2700 - allow for the PVM hypervisor per-instance kernel and initrd paths
2701 - add a new command ``gnt-cluster verify-disks`` which uses a new
2702 algorithm to improve the reconnection of the DRBD pairs if the device
2703 on the secondary node has gone away
2704 - make logical volume code auto-activate LVs at disk activation time
2705 - slightly improve the speed of activating disks
2706 - allow specification of the MAC address at instance creation time, and
2707 changing it later via ``gnt-instance modify``
2708 - fix handling of external commands that generate lots of output on
2710 - update documentation with regard to minimum version of DRBD8 supported
2716 *(Released Tue, 4 Dec 2007)*
2718 - Log the ``xm create`` output to the node daemon log on failure (to
2719 help diagnosing the error)
2720 - In debug mode, log all external commands output if failed to the logs
2721 - Change parsing of lvm commands to ignore stderr
2727 *(Released Wed, 28 Nov 2007)*
2729 - Another round of updates to the DRBD 8 code to deal with more failures
2730 in the replace secondary node operation
2731 - Some more logging of failures in disk operations (lvm, drbd)
2732 - A few documentation updates
2739 *(Released Tue, 13 Nov 2007)*
2741 - Change configuration file format from Python's Pickle to JSON.
2742 Upgrading is possible using the cfgupgrade utility.
2743 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
2744 faster replace disks and is more stable (DRBD 8 has many improvements
2745 compared to DRBD 0.7)
2746 - Added command line tags support (see man pages for ``gnt-instance``,
2747 ``gnt-node``, ``gnt-cluster``)
2748 - Added instance rename support
2749 - Added multi-instance startup/shutdown
2750 - Added cluster rename support
2751 - Added ``gnt-node evacuate`` to simplify some node operations
2752 - Added instance reboot operation that can speedup reboot as compared to
2754 - Soften the requirement that hostnames are in FQDN format
2755 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
2757 - Removed dependency on debian's patched fping that uses the
2758 non-standard ``-S`` option
2759 - Now the OS definitions are searched for in multiple, configurable
2760 paths (easier for distros to package)
2761 - Some changes to the hooks infrastructure (especially the new
2762 post-configuration update hook)
2763 - Other small bugfixes
2765 .. vim: set textwidth=72 syntax=rst :