10 Incompatible/important changes
11 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
13 - Instance policy can contain multiple instance specs, as described in
14 the “Constrained instance sizes” section of :doc:`Partitioned Ganeti
15 <design-partitioned>`. As a consequence, it's not possible to partially change
16 or override instance specs. Bounding specs (min and max) can be specified as a
17 whole using the new option ``--ipolicy-bounds-specs``, while standard
18 specs use the new option ``--ipolicy-std-specs``.
19 - The output of the info command of gnt-cluster, gnt-group, gnt-node,
20 gnt-instance is a valid YAML object.
21 - hail now honors network restrictions when allocating nodes. This led to an
22 update of the IAllocator protocol. See the IAllocator documentation for
28 - The :doc:`Remote API <rapi>` daemon now supports a command line flag
29 to always require authentication, ``--require-authentication``. It can
30 be specified in ``$sysconfdir/default/ganeti``.
31 - A new cluster attribute 'enabled_disk_templates' is introduced. It will
32 be used to manage the disk templates to be used by instances in the cluster.
33 Initially, it will be set to a list that includes plain, drbd, if they were
34 enabled by specifying a volume group name, and file and sharedfile, if those
35 were enabled at configure time. Additionally, it will include all disk
36 templates that are currently used by instances. The order of disk templates
37 will be based on Ganeti's history of supporting them. In the future, the
38 first entry of the list will be used as a default disk template on instance
40 - ``cfgupgrade`` now supports a ``--downgrade`` option to bring the
41 configuration back to the previous stable version.
42 - Disk templates in group ipolicy can be restored to the default value.
43 - Initial support for diskless instances and virtual clusters in QA.
44 - More QA and unit tests for instance policies.
45 - Every opcode now contains a reason trail (visible through ``gnt-job info``)
46 describing why the opcode itself was executed.
47 - The monitoring daemon is now available. It allows users to query the cluster
48 for obtaining information about the status of the system. The daemon is only
49 responsible for providing the information over the network: the actual data
50 gathering is performed by data collectors (currently, only the DRBD status
51 collector is available).
52 - In order to help developers work on Ganeti, a new script
53 (``devel/build_chroot``) is provided, for building a chroot that contains all
54 the required development libraries and tools for compiling Ganeti on a Debian
56 - A new tool, ``harep``, for performing self-repair and recreation of instances
57 in Ganeti has been added.
58 - Split queries are enabled for tags, network, exports, cluster info, groups,
60 - New command ``show-ispecs-cmd`` for ``gnt-cluster`` and ``gnt-group``.
61 It prints the command line to set the current policies, to ease
63 - Add the ``vnet_hdr`` HV parameter for KVM, to control whether the tap
64 devices for KVM virtio-net interfaces will get created with VNET_HDR
65 (IFF_VNET_HDR) support. If set to false, it disables offloading on the
66 virtio-net interfaces, which prevents host kernel tainting and log
67 flooding, when dealing with broken or malicious virtio-net drivers.
68 It's set to true by default.
72 The following new dependencies have been added:
75 - The ``curl`` library is not optional anymore for compiling the Haskell code.
76 - ``snap-server`` library (if monitoring is enabled).
79 - The minimum Python version needed to run Ganeti is now 2.6.
80 - ``yaml`` library (only for running the QA).
85 - Fix upgrading/downgrading from 2.7
86 - Increase maximum RAPI message size
87 - Documentation updates
88 - Split ``confd`` between ``luxid`` and ``confd``
89 - Merge 2.7 series up to the 2.7.1 release
90 - Allow the ``modify_etc_hosts`` option to be changed
91 - Add better debugging for ``luxid`` queries
92 - Expose bulk parameter for GetJobs in RAPI client
93 - Expose missing ``network`` fields in RAPI
94 - Add some ``cluster verify`` tests
96 - Fix a malfunction in ``hspace``'s tiered allocation
97 - Fix query compatibility between haskell and python implementations
98 - Add the ``vnet_hdr`` HV parameter for KVM
104 *(Released Mon, 24 Jun 2013)*
106 This was the first beta release of the 2.8 series. All important changes
107 are listed in the latest 2.8 entry.
113 *(Released Thu, 25 Jul 2013)*
115 - Add logrotate functionality in daemon-util
116 - Add logrotate example file
117 - Add missing fields to network queries over rapi
118 - Fix network object timestamps
119 - Add support for querying network timestamps
120 - Fix a typo in the example crontab
121 - Fix a documentation typo
127 *(Released Thu, 04 Jul 2013)*
129 Incompatible/important changes
130 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
132 - Instance policies for disk size were documented to be on a per-disk
133 basis, but hail applied them to the sum of all disks. This has been
135 - ``hbal`` will now exit with status 0 if, during job execution over
136 LUXI, early exit has been requested and all jobs are successful;
137 before, exit status 1 was used, which cannot be differentiated from
139 - Compatibility with newer versions of rbd has been fixed
140 - ``gnt-instance batch-create`` has been changed to use the bulk create
141 opcode from Ganeti. This lead to incompatible changes in the format of
142 the JSON file. It's now not a custom dict anymore but a dict
143 compatible with the ``OpInstanceCreate`` opcode.
144 - Parent directories for file storage need to be listed in
145 ``$sysconfdir/ganeti/file-storage-paths`` now. ``cfgupgrade`` will
146 write the file automatically based on old configuration values, but it
147 can not distribute it across all nodes and the file contents should be
148 verified. Use ``gnt-cluster copyfile
149 $sysconfdir/ganeti/file-storage-paths`` once the cluster has been
150 upgraded. The reason for requiring this list of paths now is that
151 before it would have been possible to inject new paths via RPC,
152 allowing files to be created in arbitrary locations. The RPC protocol
153 is protected using SSL/X.509 certificates, but as a design principle
154 Ganeti does not permit arbitrary paths to be passed.
155 - The parsing of the variants file for OSes (see
156 :manpage:`ganeti-os-interface(7)`) has been slightly changed: now empty
157 lines and comment lines (starting with ``#``) are ignored for better
159 - The ``setup-ssh`` tool added in Ganeti 2.2 has been replaced and is no
160 longer available. ``gnt-node add`` now invokes a new tool on the
161 destination node, named ``prepare-node-join``, to configure the SSH
162 daemon. Paramiko is no longer necessary to configure nodes' SSH
163 daemons via ``gnt-node add``.
164 - Draining (``gnt-cluster queue drain``) and un-draining the job queue
165 (``gnt-cluster queue undrain``) now affects all nodes in a cluster and
166 the flag is not reset after a master failover.
167 - Python 2.4 has *not* been tested with this release. Using 2.6 or above
168 is recommended. 2.6 will be mandatory from the 2.8 series.
174 - New network management functionality to support automatic allocation
175 of IP addresses and managing of network parameters. See
176 :manpage:`gnt-network(8)` for more details.
177 - New external storage backend, to allow managing arbitrary storage
178 systems external to the cluster. See
179 :manpage:`ganeti-extstorage-interface(7)`.
180 - New ``exclusive-storage`` node parameter added, restricted to
181 nodegroup level. When it's set to true, physical disks are assigned in
182 an exclusive fashion to instances, as documented in :doc:`Partitioned
183 Ganeti <design-partitioned>`. Currently, only instances using the
184 ``plain`` disk template are supported.
185 - The KVM hypervisor has been updated with many new hypervisor
186 parameters, including a generic one for passing arbitrary command line
187 values. See a complete list in :manpage:`gnt-instance(8)`. It is now
188 compatible up to qemu 1.4.
189 - A new tool, called ``mon-collector``, is the stand-alone executor of
190 the data collectors for a monitoring system. As of this version, it
191 just includes the DRBD data collector, that can be executed by calling
192 ``mon-collector`` using the ``drbd`` parameter. See
193 :manpage:`mon-collector(7)`.
194 - A new user option, :pyeval:`rapi.RAPI_ACCESS_READ`, has been added
195 for RAPI users. It allows granting permissions to query for
196 information to a specific user without giving
197 :pyeval:`rapi.RAPI_ACCESS_WRITE` permissions.
198 - A new tool named ``node-cleanup`` has been added. It cleans remains of
199 a cluster from a machine by stopping all daemons, removing
200 certificates and ssconf files. Unless the ``--no-backup`` option is
201 given, copies of the certificates are made.
202 - Instance creations now support the use of opportunistic locking,
203 potentially speeding up the (parallel) creation of multiple instances.
204 This feature is currently only available via the :doc:`RAPI
205 <rapi>` interface and when an instance allocator is used. If the
206 ``opportunistic_locking`` parameter is set the opcode will try to
207 acquire as many locks as possible, but will not wait for any locks
208 held by other opcodes. If not enough resources can be found to
209 allocate the instance, the temporary error code
210 :pyeval:`errors.ECODE_TEMP_NORES` is returned. The operation can be
211 retried thereafter, with or without opportunistic locking.
212 - New experimental linux-ha resource scripts.
213 - Restricted-commands support: ganeti can now be asked (via command line
214 or rapi) to perform commands on a node. These are passed via ganeti
215 RPC rather than ssh. This functionality is restricted to commands
216 specified on the ``$sysconfdir/ganeti/restricted-commands`` for security
217 reasons. The file is not copied automatically.
223 - Diskless instances are now externally mirrored (Issue 237). This for
224 now has only been tested in conjunction with explicit target nodes for
226 - Queries not needing locks or RPC access to the node can now be
227 performed by the confd daemon, making them independent from jobs, and
228 thus faster to execute. This is selectable at configure time.
229 - The functionality for allocating multiple instances at once has been
230 overhauled and is now also available through :doc:`RAPI <rapi>`.
232 There are no significant changes from version 2.7.0~rc3.
238 *(Released Tue, 25 Jun 2013)*
240 - Fix permissions on the confd query socket (Issue 477)
241 - Fix permissions on the job archive dir (Issue 498)
242 - Fix handling of an internal exception in replace-disks (Issue 472)
243 - Fix gnt-node info handling of shortened names (Issue 497)
244 - Fix gnt-instance grow-disk when wiping is enabled
245 - Documentation improvements, and support for newer pandoc
246 - Fix hspace honoring ipolicy for disks (Issue 484)
247 - Improve handling of the ``kvm_extra`` HV parameter
253 *(Released Fri, 24 May 2013)*
255 - ``devel/upload`` now works when ``/var/run`` on the target nodes is a
257 - Disks added through ``gnt-instance modify`` or created through
258 ``gnt-instance recreate-disks`` are wiped, if the
259 ``prealloc_wipe_disks`` flag is set.
260 - If wiping newly created disks fails, the disks are removed. Also,
261 partial failures in creating disks through ``gnt-instance modify``
262 triggers a cleanup of the partially-created disks.
263 - Removing the master IP address doesn't fail if the address has been
265 - Fix ownership of the OS log dir
266 - Workaround missing SO_PEERCRED constant (Issue 191)
272 *(Released Fri, 3 May 2013)*
274 This was the first release candidate of the 2.7 series. Since beta3:
276 - Fix kvm compatibility with qemu 1.4 (Issue 389)
277 - Documentation updates (admin guide, upgrade notes, install
278 instructions) (Issue 372)
279 - Fix gnt-group list nodes and instances count (Issue 436)
280 - Fix compilation without non-mandatory libraries (Issue 441)
281 - Fix xen-hvm hypervisor forcing nics to type 'ioemu' (Issue 247)
282 - Make confd logging more verbose at INFO level (Issue 435)
283 - Improve "networks" documentation in :manpage:`gnt-instance(8)`
284 - Fix failure path for instance storage type conversion (Issue 229)
285 - Update htools text backend documentation
286 - Improve the renew-crypto section of :manpage:`gnt-cluster(8)`
287 - Disable inter-cluster instance move for file-based instances, because
288 it is dependant on instance export, which is not supported for
289 file-based instances. (Issue 414)
290 - Fix gnt-job crashes on non-ascii characters (Issue 427)
291 - Fix volume group checks on non-vm-capable nodes (Issue 432)
297 *(Released Mon, 22 Apr 2013)*
299 This was the third beta release of the 2.7 series. Since beta2:
301 - Fix hail to verify disk instance policies on a per-disk basis (Issue 418).
302 - Fix data loss on wrong usage of ``gnt-instance move``
303 - Properly export errors in confd-based job queries
304 - Add ``users-setup`` tool
305 - Fix iallocator protocol to report 0 as a disk size for diskless
306 instances. This avoids hail breaking when a diskless instance is
308 - Fix job queue directory permission problem that made confd job queries
309 fail. This requires running an ``ensure-dirs --full-run`` on upgrade
310 for access to archived jobs (Issue 406).
311 - Limit the sizes of networks supported by ``gnt-network`` to something
312 between a ``/16`` and a ``/30`` to prevent memory bloat and crashes.
313 - Fix bugs in instance disk template conversion
314 - Fix GHC 7 compatibility
315 - Fix ``burnin`` install path (Issue 426).
316 - Allow very small disk grows (Issue 347).
317 - Fix a ``ganeti-noded`` memory bloat introduced in 2.5, by making sure
318 that noded doesn't import masterd code (Issue 419).
319 - Make sure the default metavg at cluster init is the same as the vg, if
320 unspecified (Issue 358).
321 - Fix cleanup of partially created disks (part of Issue 416)
327 *(Released Tue, 2 Apr 2013)*
329 This was the second beta release of the 2.7 series. Since beta1:
331 - Networks no longer have a "type" slot, since this information was
332 unused in Ganeti: instead of it tags should be used.
333 - The rapi client now has a ``target_node`` option to MigrateInstance.
334 - Fix early exit return code for hbal (Issue 386).
335 - Fix ``gnt-instance migrate/failover -n`` (Issue 396).
336 - Fix ``rbd showmapped`` output parsing (Issue 312).
337 - Networks are now referenced indexed by UUID, rather than name. This
338 will require running cfgupgrade, from 2.7.0beta1, if networks are in
340 - The OS environment now includes network information.
341 - Deleting of a network is now disallowed if any instance nic is using
342 it, to prevent dangling references.
343 - External storage is now documented in man pages.
344 - The exclusive_storage flag can now only be set at nodegroup level.
345 - Hbal can now submit an explicit priority with its jobs.
346 - Many network related locking fixes.
347 - Bump up the required pylint version to 0.25.1.
348 - Fix the ``no_remember`` option in RAPI client.
349 - Many ipolicy related tests, qa, and fixes.
350 - Many documentation improvements and fixes.
351 - Fix building with ``--disable-file-storage``.
352 - Fix ``-q`` option in htools, which was broken if passed more than
354 - Some haskell/python interaction improvements and fixes.
355 - Fix iallocator in case of missing LVM storage.
356 - Fix confd config load in case of ``--no-lvm-storage``.
357 - The confd/query functionality is now mentioned in the security
364 *(Released Wed, 6 Feb 2013)*
366 This was the first beta release of the 2.7 series. All important changes
367 are listed in the latest 2.7 entry.
373 *(Released Fri, 21 Dec 2012)*
375 Important behaviour change: hbal won't rebalance anymore instances which
376 have the ``auto_balance`` attribute set to false. This was the intention
377 all along, but until now it only skipped those from the N+1 memory
378 reservation (DRBD-specific).
380 A significant number of bug fixes in this release:
382 - Fixed disk adoption interaction with ipolicy checks.
383 - Fixed networking issues when instances are started, stopped or
384 migrated, by forcing the tap device's MAC prefix to "fe" (issue 217).
385 - Fixed the warning in cluster verify for shared storage instances not
387 - Fixed removal of storage directory on shared file storage (issue 262).
388 - Fixed validation of LVM volume group name in OpClusterSetParams
389 (``gnt-cluster modify``) (issue 285).
390 - Fixed runtime memory increases (``gnt-instance modify -m``).
391 - Fixed live migration under Xen's ``xl`` mode.
392 - Fixed ``gnt-instance console`` with ``xl``.
393 - Fixed building with newer Haskell compiler/libraries.
394 - Fixed PID file writing in Haskell daemons (confd); this prevents
395 restart issues if confd was launched manually (outside of
396 ``daemon-util``) while another copy of it was running
397 - Fixed a type error when doing live migrations with KVM (issue 297) and
398 the error messages for failing migrations have been improved.
399 - Fixed opcode validation for the out-of-band commands (``gnt-node
401 - Fixed a type error when unsetting OS hypervisor parameters (issue
402 311); now it's possible to unset all OS-specific hypervisor
404 - Fixed the ``dry-run`` mode for many operations: verification of
405 results was over-zealous but didn't take into account the ``dry-run``
406 operation, resulting in "wrong" failures.
407 - Fixed bash completion in ``gnt-job list`` when the job queue has
408 hundreds of entries; especially with older ``bash`` versions, this
409 results in significant CPU usage.
411 And lastly, a few other improvements have been made:
413 - Added option to force master-failover without voting (issue 282).
414 - Clarified error message on lock conflict (issue 287).
415 - Logging of newly submitted jobs has been improved (issue 290).
416 - Hostname checks have been made uniform between instance rename and
418 - The ``--submit`` option is now supported by ``gnt-debug delay``.
419 - Shutting down the master daemon by sending SIGTERM now stops it from
420 processing jobs waiting for locks; instead, those jobs will be started
421 once again after the master daemon is started the next time (issue
423 - Support for Xen's ``xl`` program has been improved (besides the fixes
425 - Reduced logging noise in the Haskell confd daemon (only show one log
426 entry for each config reload, instead of two).
427 - Several man page updates and typo fixes.
433 *(Released Fri, 12 Oct 2012)*
435 A small bugfix release. Among the bugs fixed:
437 - Fixed double use of ``PRIORITY_OPT`` in ``gnt-node migrate``, that
438 made the command unusable.
439 - Commands that issue many jobs don't fail anymore just because some jobs
440 take so long that other jobs are archived.
441 - Failures during ``gnt-instance reinstall`` are reflected by the exit
443 - Issue 190 fixed. Check for DRBD in cluster verify is enabled only when
445 - When ``always_failover`` is set, ``--allow-failover`` is not required
446 in migrate commands anymore.
447 - ``bash_completion`` works even if extglob is disabled.
448 - Fixed bug with locks that made failover for RDB-based instances fail.
449 - Fixed bug in non-mirrored instance allocation that made Ganeti choose
450 a random node instead of one based on the allocator metric.
451 - Support for newer versions of pylint and pep8.
452 - Hail doesn't fail anymore when trying to add an instance of type
453 ``file``, ``sharedfile`` or ``rbd``.
454 - Added new Makefile target to rebuild the whole distribution, so that
455 all files are included.
461 *(Released Fri, 27 Jul 2012)*
464 .. attention:: The ``LUXI`` protocol has been made more consistent
465 regarding its handling of command arguments. This, however, leads to
466 incompatibility issues with previous versions. Please ensure that you
467 restart Ganeti daemons soon after the upgrade, otherwise most
468 ``LUXI`` calls (job submission, setting/resetting the drain flag,
469 pausing/resuming the watcher, cancelling and archiving jobs, querying
470 the cluster configuration) will fail.
479 The current ``admin_up`` field, which used to denote whether an instance
480 should be running or not, has been removed. Instead, ``admin_state`` is
481 introduced, with 3 possible values -- ``up``, ``down`` and ``offline``.
483 The rational behind this is that an instance being “down” can have
486 - it could be down during a reboot
487 - it could be temporarily be down for a reinstall
488 - or it could be down because it is deprecated and kept just for its
491 The previous Boolean state was making it difficult to do capacity
492 calculations: should Ganeti reserve memory for a down instance? Now, the
493 tri-state field makes it clear:
495 - in ``up`` and ``down`` state, all resources are reserved for the
496 instance, and it can be at any time brought up if it is down
497 - in ``offline`` state, only disk space is reserved for it, but not
500 The field can have an extra use: since the transition between ``up`` and
501 ``down`` and vice-versus is done via ``gnt-instance start/stop``, but
502 transition between ``offline`` and ``down`` is done via ``gnt-instance
503 modify``, it is possible to given different rights to users. For
504 example, owners of an instance could be allowed to start/stop it, but
505 not transition it out of the offline state.
507 Instance policies and specs
508 +++++++++++++++++++++++++++
510 In previous Ganeti versions, an instance creation request was not
511 limited on the minimum size and on the maximum size just by the cluster
512 resources. As such, any policy could be implemented only in third-party
513 clients (RAPI clients, or shell wrappers over ``gnt-*``
514 tools). Furthermore, calculating cluster capacity via ``hspace`` again
515 required external input with regards to instance sizes.
517 In order to improve these workflows and to allow for example better
518 per-node group differentiation, we introduced instance specs, which
521 - minimum instance disk size, disk count, memory size, cpu count
522 - maximum values for the above metrics
523 - and “standard” values (used in ``hspace`` to calculate the standard
526 The minimum/maximum values can be also customised at node-group level,
527 for example allowing more powerful hardware to support bigger instance
530 Beside the instance specs, there are a few other settings belonging to
531 the instance policy framework. It is possible now to customise, per
532 cluster and node-group:
534 - the list of allowed disk templates
535 - the maximum ratio of VCPUs per PCPUs (to control CPU oversubscription)
536 - the maximum ratio of instance to spindles (see below for more
537 information) for local storage
539 All these together should allow all tools that talk to Ganeti to know
540 what are the ranges of allowed values for instances and the
541 over-subscription that is allowed.
543 For the VCPU/PCPU ratio, we already have the VCPU configuration from the
544 instance configuration, and the physical CPU configuration from the
545 node. For the spindle ratios however, we didn't track before these
546 values, so new parameters have been added:
548 - a new node parameter ``spindle_count``, defaults to 1, customisable at
549 node group or node level
550 - at new backend parameter (for instances), ``spindle_use`` defaults to 1
552 Note that spindles in this context doesn't need to mean actual
553 mechanical hard-drives; it's just a relative number for both the node
554 I/O capacity and instance I/O consumption.
556 Instance migration behaviour
557 ++++++++++++++++++++++++++++
559 While live-migration is in general desirable over failover, it is
560 possible that for some workloads it is actually worse, due to the
561 variable time of the “suspend” phase during live migration.
563 To allow the tools to work consistently over such instances (without
564 having to hard-code instance names), a new backend parameter
565 ``always_failover`` has been added to control the migration/failover
566 behaviour. When set to True, all migration requests for an instance will
567 instead fall-back to failover.
569 Instance memory ballooning
570 ++++++++++++++++++++++++++
572 Initial support for memory ballooning has been added. The memory for an
573 instance is no longer fixed (backend parameter ``memory``), but instead
574 can vary between minimum and maximum values (backend parameters
575 ``minmem`` and ``maxmem``). Currently we only change an instance's
578 - live migrating or failing over and instance and the target node
579 doesn't have enough memory
580 - user requests changing the memory via ``gnt-instance modify
586 In order to control the use of specific CPUs by instance, support for
587 controlling CPU pinning has been added for the Xen, HVM and LXC
588 hypervisors. This is controlled by a new hypervisor parameter
589 ``cpu_mask``; details about possible values for this are in the
590 :manpage:`gnt-instance(8)`. Note that use of the most specific (precise
591 VCPU-to-CPU mapping) form will work well only when all nodes in your
592 cluster have the same amount of CPUs.
597 Another area in which Ganeti was not customisable were the parameters
598 used for storage configuration, e.g. how many stripes to use for LVM,
599 DRBD resync configuration, etc.
601 To improve this area, we've added disks parameters, which are
602 customisable at cluster and node group level, and which allow to
603 specify various parameters for disks (DRBD has the most parameters
604 currently), for example:
606 - DRBD resync algorithm and parameters (e.g. speed)
607 - the default VG for meta-data volumes for DRBD
608 - number of stripes for LVM (plain disk template)
611 These parameters can be modified via ``gnt-cluster modify -D …`` and
612 ``gnt-group modify -D …``, and are used at either instance creation (in
613 case of LVM stripes, for example) or at disk “activation” time
616 Rados block device support
617 ++++++++++++++++++++++++++
619 A Rados (http://ceph.com/wiki/Rbd) storage backend has been added,
620 denoted by the ``rbd`` disk template type. This is considered
621 experimental, feedback is welcome. For details on configuring it, see
622 the :doc:`install` document and the :manpage:`gnt-cluster(8)` man page.
627 The existing master IP functionality works well only in simple setups (a
628 single network shared by all nodes); however, if nodes belong to
629 different networks, then the ``/32`` setup and lack of routing
630 information is not enough.
632 To allow the master IP to function well in more complex cases, the
633 system was reworked as follows:
635 - a master IP netmask setting has been added
636 - the master IP activation/turn-down code was moved from the node daemon
638 - whether to run the Ganeti-supplied master IP script or a user-supplied
639 on is a ``gnt-cluster init`` setting
641 Details about the location of the standard and custom setup scripts are
642 in the man page :manpage:`gnt-cluster(8)`; for information about the
643 setup script protocol, look at the Ganeti-supplied script.
648 The `SPICE <http://www.linux-kvm.org/page/SPICE>`_ support has been
651 It is now possible to use TLS-protected connections, and when renewing
652 or changing the cluster certificates (via ``gnt-cluster renew-crypto``,
653 it is now possible to specify spice or spice CA certificates. Also, it
654 is possible to configure a password for SPICE sessions via the
655 hypervisor parameter ``spice_password_file``.
657 There are also new parameters to control the compression and streaming
658 options (e.g. ``spice_image_compression``, ``spice_streaming_video``,
659 etc.). For details, see the man page :manpage:`gnt-instance(8)` and look
660 for the spice parameters.
662 Lastly, it is now possible to see the SPICE connection information via
663 ``gnt-instance console``.
668 A new tool (``tools/ovfconverter``) has been added that supports
669 conversion between Ganeti and the `Open Virtualization Format
670 <http://en.wikipedia.org/wiki/Open_Virtualization_Format>`_ (both to and
673 This relies on the ``qemu-img`` tool to convert the disk formats, so the
674 actual compatibility with other virtualization solutions depends on it.
679 The configuration query daemon (``ganeti-confd``) is now optional, and
680 has been rewritten in Haskell; whether to use the daemon at all, use the
681 Python (default) or the Haskell version is selectable at configure time
682 via the ``--enable-confd`` parameter, which can take one of the
683 ``haskell``, ``python`` or ``no`` values. If not used, disabling the
684 daemon will result in a smaller footprint; for larger systems, we
685 welcome feedback on the Haskell version which might become the default
688 If you want to use ``gnt-node list-drbd`` you need to have the Haskell
689 daemon running. The Python version doesn't implement the new call.
692 User interface changes
693 ~~~~~~~~~~~~~~~~~~~~~~
695 We have replaced the ``--disks`` option of ``gnt-instance
696 replace-disks`` with a more flexible ``--disk`` option, which allows
697 adding and removing disks at arbitrary indices (Issue 188). Furthermore,
698 disk size and mode can be changed upon recreation (via ``gnt-instance
699 recreate-disks``, which accepts the same ``--disk`` option).
701 As many people are used to a ``show`` command, we have added that as an
702 alias to ``info`` on all ``gnt-*`` commands.
704 The ``gnt-instance grow-disk`` command has a new mode in which it can
705 accept the target size of the disk, instead of the delta; this can be
706 more safe since two runs in absolute mode will be idempotent, and
707 sometimes it's also easier to specify the desired size directly.
709 Also the handling of instances with regard to offline secondaries has
710 been improved. Instance operations should not fail because one of it's
711 secondary nodes is offline, even though it's safe to proceed.
713 A new command ``list-drbd`` has been added to the ``gnt-node`` script to
714 support debugging of DRBD issues on nodes. It provides a mapping of DRBD
715 minors to instance name.
720 RAPI coverage has improved, with (for example) new resources for
721 recreate-disks, node power-cycle, etc.
726 There is partial support for ``xl`` in the Xen hypervisor; feedback is
729 Python 2.7 is better supported, and after Ganeti 2.6 we will investigate
730 whether to still support Python 2.4 or move to Python 2.6 as minimum
733 Support for Fedora has been slightly improved; the provided example
734 init.d script should work better on it and the INSTALL file should
735 document the needed dependencies.
740 The deprecated ``QueryLocks`` LUXI request has been removed. Use
741 ``Query(what=QR_LOCK, ...)`` instead.
743 The LUXI requests :pyeval:`luxi.REQ_QUERY_JOBS`,
744 :pyeval:`luxi.REQ_QUERY_INSTANCES`, :pyeval:`luxi.REQ_QUERY_NODES`,
745 :pyeval:`luxi.REQ_QUERY_GROUPS`, :pyeval:`luxi.REQ_QUERY_EXPORTS` and
746 :pyeval:`luxi.REQ_QUERY_TAGS` are deprecated and will be removed in a
747 future version. :pyeval:`luxi.REQ_QUERY` should be used instead.
749 RAPI client: ``CertificateError`` now derives from
750 ``GanetiApiError``. This should make it more easy to handle Ganeti
753 Deprecation warnings due to PyCrypto/paramiko import in
754 ``tools/setup-ssh`` have been silenced, as usually they are safe; please
755 make sure to run an up-to-date paramiko version, if you use this tool.
757 The QA scripts now depend on Python 2.5 or above (the main code base
758 still works with Python 2.4).
760 The configuration file (``config.data``) is now written without
761 indentation for performance reasons; if you want to edit it, it can be
762 re-formatted via ``tools/fmtjson``.
764 A number of bugs has been fixed in the cluster merge tool.
766 ``x509`` certification verification (used in import-export) has been
767 changed to allow the same clock skew as permitted by the cluster
768 verification. This will remove some rare but hard to diagnose errors in
775 *(Released Thu, 19 Jul 2012)*
777 Very few changes from rc4 to the final release, only bugfixes:
779 - integrated fixes from release 2.5.2 (fix general boot flag for KVM
780 instance, fix CDROM booting for KVM instances)
781 - fixed node group modification of node parameters
782 - fixed issue in LUClusterVerifyGroup with multi-group clusters
783 - fixed generation of bash completion to ensure a stable ordering
790 *(Released Fri, 13 Jul 2012)*
792 Third release candidate for 2.6. The following changes were done from
795 - Fixed ``UpgradeConfig`` w.r.t. to disk parameters on disk objects.
796 - Fixed an inconsistency in the LUXI protocol with the provided
797 arguments (NOT backwards compatible)
798 - Fixed a bug with node groups ipolicy where ``min`` was greater than
799 the cluster ``std`` value
800 - Implemented a new ``gnt-node list-drbd`` call to list DRBD minors for
801 easier instance debugging on nodes (requires ``hconfd`` to work)
807 *(Released Tue, 03 Jul 2012)*
809 Second release candidate for 2.6. The following changes were done from
812 - Fixed ``gnt-cluster verify`` regarding ``master-ip-script`` on non
814 - Fixed a RAPI regression on missing beparams/memory
815 - Fixed redistribution of files on offline nodes
816 - Added possibility to run activate-disks even though secondaries are
817 offline. With this change it relaxes also the strictness on some other
818 commands which use activate disks internally:
819 * ``gnt-instance start|reboot|rename|backup|export``
820 - Made it possible to remove safely an instance if its secondaries are
822 - Made it possible to reinstall even though secondaries are offline
828 *(Released Mon, 25 Jun 2012)*
830 First release candidate for 2.6. The following changes were done from
833 - Fixed bugs with disk parameters and ``rbd`` templates as well as
835 - Made ``gnt-instance modify`` more consistent regarding new NIC/Disk
836 behaviour. It supports now the modify operation
837 - ``hcheck`` implemented to analyze cluster health and possibility of
838 improving health by rebalance
839 - ``hbal`` has been improved in dealing with split instances
845 *(Released Mon, 11 Jun 2012)*
847 Second beta release of 2.6. The following changes were done from beta2
850 - Fixed ``daemon-util`` with non-root user models
851 - Fixed creation of plain instances with ``--no-wait-for-sync``
852 - Fix wrong iv_names when running ``cfgupgrade``
853 - Export more information in RAPI group queries
854 - Fixed bug when changing instance network interfaces
855 - Extended burnin to do NIC changes
856 - query: Added ``<``, ``>``, ``<=``, ``>=`` comparison operators
857 - Changed default for DRBD barriers
858 - Fixed DRBD error reporting for syncer rate
859 - Verify the options on disk parameters
861 And of course various fixes to documentation and improved unittests and
868 *(Released Wed, 23 May 2012)*
870 First beta release of 2.6. The following changes were done from beta1 to
873 - integrated patch for distributions without ``start-stop-daemon``
874 - adapted example init.d script to work on Fedora
875 - fixed log handling in Haskell daemons
876 - adapted checks in the watcher for pycurl linked against libnss
877 - add partial support for ``xl`` instead of ``xm`` for Xen
878 - fixed a type issue in cluster verification
879 - fixed ssconf handling in the Haskell code (was breaking confd in IPv6
882 Plus integrated fixes from the 2.5 branch:
884 - fixed ``kvm-ifup`` to use ``/bin/bash``
885 - fixed parallel build failures
886 - KVM live migration when using a custom keymap
892 *(Released Tue, 24 Jul 2012)*
894 A small bugfix release, with no new features:
896 - fixed bash-isms in kvm-ifup, for compatibility with systems which use a
897 different default shell (e.g. Debian, Ubuntu)
898 - fixed KVM startup and live migration with a custom keymap (fixes Issue
899 243 and Debian bug #650664)
900 - fixed compatibility with KVM versions that don't support multiple boot
901 devices (fixes Issue 230 and Debian bug #624256)
903 Additionally, a few fixes were done to the build system (fixed parallel
904 build failures) and to the unittests (fixed race condition in test for
905 FileID functions, and the default enable/disable mode for QA test is now
912 *(Released Fri, 11 May 2012)*
914 A small bugfix release.
916 The main issues solved are on the topic of compatibility with newer LVM
919 - fixed parsing of ``lv_attr`` field
920 - adapted to new ``vgreduce --removemissing`` behaviour where sometimes
921 the ``--force`` flag is needed
923 Also on the topic of compatibility, ``tools/lvmstrap`` has been changed
924 to accept kernel 3.x too (was hardcoded to 2.6.*).
926 A regression present in 2.5.0 that broke handling (in the gnt-* scripts)
927 of hook results and that also made display of other errors suboptimal
928 was fixed; the code behaves now like 2.4 and earlier.
930 Another change in 2.5, the cleanup of the OS scripts environment, is too
931 aggressive: it removed even the ``PATH`` variable, which requires the OS
932 scripts to *always* need to export it. Since this is a bit too strict,
933 we now export a minimal PATH, the same that we export for hooks.
935 The fix for issue 201 (Preserve bridge MTU in KVM ifup script) was
936 integrated into this release.
938 Finally, a few other miscellaneous changes were done (no new features,
939 just small improvements):
941 - Fix ``gnt-group --help`` display
942 - Fix hardcoded Xen kernel path
943 - Fix grow-disk handling of invalid units
944 - Update synopsis for ``gnt-cluster repair-disk-sizes``
945 - Accept both PUT and POST in noded (makes future upgrade to 2.6 easier)
951 *(Released Thu, 12 Apr 2012)*
953 Incompatible/important changes and bugfixes
954 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
956 - The default of the ``/2/instances/[instance_name]/rename`` RAPI
957 resource's ``ip_check`` parameter changed from ``True`` to ``False``
958 to match the underlying LUXI interface.
959 - The ``/2/nodes/[node_name]/evacuate`` RAPI resource was changed to use
960 body parameters, see :doc:`RAPI documentation <rapi>`. The server does
961 not maintain backwards-compatibility as the underlying operation
962 changed in an incompatible way. The RAPI client can talk to old
963 servers, but it needs to be told so as the return value changed.
964 - When creating file-based instances via RAPI, the ``file_driver``
965 parameter no longer defaults to ``loop`` and must be specified.
966 - The deprecated ``bridge`` NIC parameter is no longer supported. Use
968 - Support for the undocumented and deprecated RAPI instance creation
969 request format version 0 has been dropped. Use version 1, supported
970 since Ganeti 2.1.3 and :doc:`documented <rapi>`, instead.
971 - Pyparsing 1.4.6 or above is required, see :doc:`installation
972 documentation <install>`.
973 - The "cluster-verify" hooks are now executed per group by the
974 ``OP_CLUSTER_VERIFY_GROUP`` opcode. This maintains the same behavior
975 if you just run ``gnt-cluster verify``, which generates one opcode per
977 - The environment as passed to the OS scripts is cleared, and thus no
978 environment variables defined in the node daemon's environment will be
979 inherited by the scripts.
980 - The :doc:`iallocator <iallocator>` mode ``multi-evacuate`` has been
982 - :doc:`New iallocator modes <design-multi-reloc>` have been added to
983 support operations involving multiple node groups.
984 - Offline nodes are ignored when failing over an instance.
985 - Support for KVM version 1.0, which changed the version reporting format
987 - TCP/IP ports used by DRBD disks are returned to a pool upon instance
989 - ``Makefile`` is now compatible with Automake 1.11.2
990 - Includes all bugfixes made in the 2.4 series
995 - The ganeti-htools project has been merged into the ganeti-core source
996 tree and will be built as part of Ganeti (see :doc:`install-quick`).
997 - Implemented support for :doc:`shared storage <design-shared-storage>`.
998 - Add support for disks larger than 2 TB in ``lvmstrap`` by supporting
999 GPT-style partition tables (requires `parted
1000 <http://www.gnu.org/s/parted/>`_).
1001 - Added support for floppy drive and 2nd CD-ROM drive in KVM hypervisor.
1002 - Allowed adding tags on instance creation.
1003 - Export instance tags to hooks (``INSTANCE_TAGS``, see :doc:`hooks`)
1004 - Allow instances to be started in a paused state, enabling the user to
1005 see the complete console output on boot using the console.
1006 - Added new hypervisor flag to control default reboot behaviour
1007 (``reboot_behavior``).
1008 - Added support for KVM keymaps (hypervisor parameter ``keymap``).
1009 - Improved out-of-band management support:
1011 - Added ``gnt-node health`` command reporting the health status of
1013 - Added ``gnt-node power`` command to manage power status of nodes.
1014 - Added command for emergency power-off (EPO), ``gnt-cluster epo``.
1016 - Instance migration can fall back to failover if instance is not
1018 - Filters can be used when listing nodes, instances, groups and locks;
1019 see :manpage:`ganeti(7)` manpage.
1020 - Added post-execution status as variables to :doc:`hooks <hooks>`
1022 - Instance tags are exported/imported together with the instance.
1023 - When given an explicit job ID, ``gnt-job info`` will work for archived
1025 - Jobs can define dependencies on other jobs (not yet supported via
1026 RAPI or command line, but used by internal commands and usable via
1029 - Lock monitor (``gnt-debug locks``) shows jobs waiting for
1032 - Instance failover is now available as a RAPI resource
1033 (``/2/instances/[instance_name]/failover``).
1034 - ``gnt-instance info`` defaults to static information if primary node
1036 - Opcodes have a new ``comment`` attribute.
1037 - Added basic SPICE support to KVM hypervisor.
1038 - ``tools/ganeti-listrunner`` allows passing of arguments to executable.
1040 Node group improvements
1041 ~~~~~~~~~~~~~~~~~~~~~~~
1043 - ``gnt-cluster verify`` has been modified to check groups separately,
1044 thereby improving performance.
1045 - Node group support has been added to ``gnt-cluster verify-disks``,
1046 which now operates per node group.
1047 - Watcher has been changed to work better with node groups.
1049 - One process and state file per node group.
1050 - Slow watcher in one group doesn't block other group's watcher.
1052 - Added new command, ``gnt-group evacuate``, to move all instances in a
1053 node group to other groups.
1054 - Added ``gnt-instance change-group`` to move an instance to another
1056 - ``gnt-cluster command`` and ``gnt-cluster copyfile`` now support
1057 per-group operations.
1058 - Node groups can be tagged.
1059 - Some operations switch from an exclusive to a shared lock as soon as
1061 - Instance's primary and secondary nodes' groups are now available as
1062 query fields (``pnode.group``, ``pnode.group.uuid``, ``snodes.group``
1063 and ``snodes.group.uuid``).
1068 - Numerous updates to documentation and manpages.
1070 - :doc:`RAPI <rapi>` documentation now has detailed parameter
1072 - Some opcode/job results are now also documented, see :doc:`RAPI
1075 - A lockset's internal lock is now also visible in lock monitor.
1076 - Log messages from job queue workers now contain information about the
1077 opcode they're processing.
1078 - ``gnt-instance console`` no longer requires the instance lock.
1079 - A short delay when waiting for job changes reduces the number of LUXI
1080 requests significantly.
1081 - DRBD metadata volumes are overwritten with zeros during disk creation.
1082 - Out-of-band commands no longer acquire the cluster lock in exclusive
1084 - ``devel/upload`` now uses correct permissions for directories.
1090 *(Released Fri, 23 Mar 2012)*
1092 This was the sixth release candidate of the 2.5 series.
1098 *(Released Mon, 9 Jan 2012)*
1100 This was the fifth release candidate of the 2.5 series.
1106 *(Released Thu, 27 Oct 2011)*
1108 This was the fourth release candidate of the 2.5 series.
1114 *(Released Wed, 26 Oct 2011)*
1116 This was the third release candidate of the 2.5 series.
1122 *(Released Tue, 18 Oct 2011)*
1124 This was the second release candidate of the 2.5 series.
1130 *(Released Tue, 4 Oct 2011)*
1132 This was the first release candidate of the 2.5 series.
1138 *(Released Wed, 31 Aug 2011)*
1140 This was the third beta release of the 2.5 series.
1146 *(Released Mon, 22 Aug 2011)*
1148 This was the second beta release of the 2.5 series.
1154 *(Released Fri, 12 Aug 2011)*
1156 This was the first beta release of the 2.5 series.
1162 *(Released Thu, 27 Oct 2011)*
1164 - Fixed bug when parsing command line parameter values ending in
1166 - Fixed assertion error after unclean master shutdown
1167 - Disable HTTP client pool for RPC, significantly reducing memory usage
1169 - Fixed queue archive creation with wrong permissions
1175 *(Released Tue, 23 Aug 2011)*
1179 - Fixed documentation for importing with ``--src-dir`` option
1180 - Fixed a bug in ``ensure-dirs`` with queue/archive permissions
1181 - Fixed a parsing issue with DRBD 8.3.11 in the Linux kernel
1187 *(Released Fri, 5 Aug 2011)*
1189 Many bug-fixes and a few small features:
1191 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
1192 issues when you tried to add an instance with two MAC addresses in one
1194 - KVM: fixed per-instance stored UID value
1195 - KVM: configure bridged NICs at migration start
1196 - KVM: Fix a bug where instance will not start with never KVM versions
1198 - Added OS search path to ``gnt-cluster info``
1199 - Fixed an issue with ``file_storage_dir`` where you were forced to
1200 provide an absolute path, but the documentation states it is a
1201 relative path, the documentation was right
1202 - Added a new parameter to instance stop/start called ``--no-remember``
1203 that will make the state change to not be remembered
1204 - Implemented ``no_remember`` at RAPI level
1205 - Improved the documentation
1206 - Node evacuation: don't call IAllocator if node is already empty
1207 - Fixed bug in DRBD8 replace disks on current nodes
1208 - Fixed bug in recreate-disks for DRBD instances
1209 - Moved assertion checking locks in ``gnt-instance replace-disks``
1210 causing it to abort with not owning the right locks for some situation
1211 - Job queue: Fixed potential race condition when cancelling queued jobs
1212 - Fixed off-by-one bug in job serial generation
1213 - ``gnt-node volumes``: Fix instance names
1214 - Fixed aliases in bash completion
1215 - Fixed a bug in reopening log files after being sent a SIGHUP
1216 - Added a flag to burnin to allow specifying VCPU count
1217 - Bugfixes to non-root Ganeti configuration
1223 *(Released Thu, 12 May 2011)*
1225 Many bug-fixes and a few new small features:
1227 - Fixed a bug related to log opening failures
1228 - Fixed a bug in instance listing with orphan instances
1229 - Fixed a bug which prevented resetting the cluster-level node parameter
1230 ``oob_program`` to the default
1231 - Many fixes related to the ``cluster-merge`` tool
1232 - Fixed a race condition in the lock monitor, which caused failures
1233 during (at least) creation of many instances in parallel
1234 - Improved output for gnt-job info
1235 - Removed the quiet flag on some ssh calls which prevented debugging
1237 - Improved the N+1 failure messages in cluster verify by actually
1238 showing the memory values (needed and available)
1239 - Increased lock attempt timeouts so that when executing long operations
1240 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
1241 too early and thus prevent the use of the 'fair' mechanism
1242 - Changed instance query data (``gnt-instance info``) to not acquire
1243 locks unless needed, thus allowing its use on locked instance if only
1244 static information is asked for
1245 - Improved behaviour with filesystems that do not support rename on an
1247 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
1248 kept locks on all nodes during the wipe, which is unneeded
1249 - Fixed ``gnt-watcher`` handling of errors during hooks execution
1250 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
1251 10GiB) which caused the wipe to fail right at the end in some cases
1252 - Fixed master IP activation when doing master failover with no-voting
1253 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
1254 the master node itself
1255 - Fixed potential data-loss in under disk full conditions, where Ganeti
1256 wouldn't check correctly the return code and would consider
1257 partially-written files 'correct'
1258 - Fixed bug related to multiple VGs and DRBD disk replacing
1259 - Added new disk parameter ``metavg`` that allows placement of the meta
1260 device for DRBD in a different volume group
1261 - Fixed error handling in the node daemon when the system libc doesn't
1262 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
1263 - Fixed lock release during replace-disks, which kept cluster-wide locks
1264 when doing disk replaces with an iallocator script
1265 - Added check for missing bridges in cluster verify
1266 - Handle EPIPE errors while writing to the terminal better, so that
1267 piping the output to e.g. ``less`` doesn't cause a backtrace
1268 - Fixed rare case where a ^C during Luxi calls could have been
1269 interpreted as server errors, instead of simply terminating
1270 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
1272 - Added a few more parameters to the KVM hypervisor, allowing a second
1273 CDROM, custom disk type for CDROMs and a floppy image
1274 - Removed redundant message in instance rename when the name is given
1276 - Added option to ``gnt-instance recreate-disks`` to allow creating the
1277 disks on new nodes, allowing recreation when the original instance
1278 nodes are completely gone
1279 - Added option when converting disk templates to DRBD to skip waiting
1280 for the resync, in order to make the instance available sooner
1281 - Added two new variables to the OS scripts environment (containing the
1283 - Made the root_path and optional parameter for the xen-pvm hypervisor,
1284 to allow use of ``pvgrub`` as bootloader
1285 - Changed the instance memory modifications to only check out-of-memory
1286 conditions on memory increases, and turned the secondary node warnings
1287 into errors (they can still be overridden via ``--force``)
1288 - Fixed the handling of a corner case when the Python installation gets
1289 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
1290 to execute a command that doesn't exist
1291 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
1292 node of the instance returned failures during instance shutdown; this
1293 adds the option ``--ignore-consistency`` to gnt-instance move
1295 And as usual, various improvements to the error messages, documentation
1302 *(Released Wed, 09 Mar 2011)*
1304 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
1305 the RAPI users file if run twice (even with ``--dry-run``).
1307 The release fixes that bug (nothing else changed).
1313 *(Released Mon, 07 Mar 2011)*
1315 Final 2.4.0 release. Just a few small fixes:
1317 - Fixed RAPI node evacuate
1318 - Fixed the kvm-ifup script
1319 - Fixed internal error handling for special job cases
1320 - Updated man page to specify the escaping feature for options
1326 *(Released Mon, 28 Feb 2011)*
1328 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
1329 that this feature wiped the disks of the wrong instance, leading to loss
1334 - Fixed title of query field containing instance name
1335 - Expanded the glossary in the documentation
1336 - Fixed one unittest (internal issue)
1342 *(Released Mon, 21 Feb 2011)*
1344 A number of bug fixes plus just a couple functionality changes.
1346 On the user-visible side, the ``gnt-* list`` command output has changed
1347 with respect to "special" field states. The current rc1 style of display
1348 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
1349 the default output mode special fields states are displayed as follows:
1351 - Offline resource: ``*``
1352 - Unavailable/not applicable: ``-``
1353 - Data missing (RPC failure): ``?``
1354 - Unknown field: ``??``
1356 Another user-visible change is the addition of ``--force-join`` to
1361 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
1362 - Fixed regression in RAPI/instance reinstall where all parameters were
1363 required (instead of optional)
1364 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
1365 - Fixed iallocator usage (offline nodes were not considered offline)
1366 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
1367 - Fixed hypervisor and OS parameter validation with respect to
1368 non-vm_capable nodes
1369 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
1371 - Fixed ``tools/listrunner`` with respect to agent-based usage
1377 *(Released Fri, 4 Feb 2011)*
1379 Many changes and fixes since the beta1 release. While there were some
1380 internal changes, the code has been mostly stabilised for the RC
1383 Note: the dumb allocator was removed in this release, as it was not kept
1384 up-to-date with the IAllocator protocol changes. It is recommended to
1385 use the ``hail`` command from the ganeti-htools package.
1387 Note: the 2.4 and up versions of Ganeti are not compatible with the
1388 0.2.x branch of ganeti-htools. You need to upgrade to
1389 ganeti-htools-0.3.0 (or later).
1391 Regressions fixed from 2.3
1392 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1394 - Fixed the ``gnt-cluster verify-disks`` command
1395 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
1397 - Fixed disk adoption breakage
1398 - Fixed wrong headers in instance listing for field aliases
1403 - Fixed corner case in KVM handling of NICs
1404 - Fixed many cases of wrong handling of non-vm_capable nodes
1405 - Fixed a bug where a missing instance symlink was not possible to
1406 recreate with any ``gnt-*`` command (now ``gnt-instance
1407 activate-disks`` does it)
1408 - Fixed the volume group name as reported by ``gnt-cluster
1410 - Increased timeouts for the import-export code, hopefully leading to
1411 fewer aborts due network or instance timeouts
1412 - Fixed bug in ``gnt-node list-storage``
1413 - Fixed bug where not all daemons were started on cluster
1414 initialisation, but only at the first watcher run
1415 - Fixed many bugs in the OOB implementation
1416 - Fixed watcher behaviour in presence of instances with offline
1418 - Fixed instance list output for instances running on the wrong node
1419 - a few fixes to the cluster-merge tool, but it still cannot merge
1420 multi-node groups (currently it is not recommended to use this tool)
1426 - Improved network configuration for the KVM hypervisor
1427 - Added e1000 as a supported NIC for Xen-HVM
1428 - Improved the lvmstrap tool to also be able to use partitions, as
1429 opposed to full disks
1430 - Improved speed of disk wiping (the cluster parameter
1431 ``prealloc_wipe_disks``, so that it has a low impact on the total time
1432 of instance creations
1433 - Added documentation for the OS parameters
1434 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
1435 hypervisor is not responding
1436 - Added display of blacklisted and hidden OS information in
1437 ``gnt-cluster info``
1438 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
1439 NIC and node parameters, which might create problems with currently
1440 invalid (but undetected) configuration files, but prevents validation
1441 failures when unrelated parameters are modified
1442 - Changed cluster initialisation to wait for the master daemon to become
1444 - Expanded the RAPI interface:
1446 - Added config redistribution resource
1447 - Added activation/deactivation of instance disks
1448 - Added export of console information
1450 - Implemented log file reopening on SIGHUP, which allows using
1451 logrotate(8) for the Ganeti log files
1452 - Added a basic OOB helper script as an example
1458 *(Released Fri, 14 Jan 2011)*
1463 - Fixed timezone issues when formatting timestamps
1464 - Added support for node groups, available via ``gnt-group`` and other
1466 - Added out-of-band framework and management, see :doc:`design
1467 document <design-oob>`
1468 - Removed support for roman numbers from ``gnt-node list`` and
1469 ``gnt-instance list``.
1470 - Allowed modification of master network interface via ``gnt-cluster
1471 modify --master-netdev``
1472 - Accept offline secondaries while shutting down instance disks
1473 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
1474 - Added support for multiple LVM volume groups
1475 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
1477 - Added commands to list available fields:
1479 - ``gnt-node list-fields``
1480 - ``gnt-group list-fields``
1481 - ``gnt-instance list-fields``
1483 - Updated documentation and man pages
1488 - Moved ``rapi_users`` file into separate directory, now named
1489 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
1491 - Added new tool for running commands on many machines,
1492 ``tools/ganeti-listrunner``
1493 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
1494 improving the ``gnt-instance console`` output
1495 - Allowed customisation of disk index separator at ``configure`` time
1496 - Export node group allocation policy to :doc:`iallocator <iallocator>`
1497 - Added support for non-partitioned md disks in ``lvmstrap``
1498 - Added script to gracefully power off KVM instances
1499 - Split ``utils`` module into smaller parts
1500 - Changed query operations to return more detailed information, e.g.
1501 whether an information is unavailable due to an offline node. To use
1502 this new functionality, the LUXI call ``Query`` must be used. Field
1503 information is now stored by the master daemon and can be retrieved
1504 using ``QueryFields``. Instances, nodes and groups can also be queried
1505 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
1506 exposed via RAPI). The following commands make use of this
1507 infrastructure change:
1509 - ``gnt-group list``
1510 - ``gnt-group list-fields``
1512 - ``gnt-node list-fields``
1513 - ``gnt-instance list``
1514 - ``gnt-instance list-fields``
1515 - ``gnt-debug locks``
1520 - New RAPI resources (see :doc:`rapi`):
1524 - ``/2/groups/[group_name]``
1525 - ``/2/groups/[group_name]/assign-nodes``
1526 - ``/2/groups/[group_name]/modify``
1527 - ``/2/groups/[group_name]/rename``
1528 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
1532 - Implemented ``no_install`` for instance creation
1533 - Implemented OS parameters for instance reinstallation, allowing
1534 use of special settings on reinstallation (e.g. for preserving data)
1539 - Added IPv6 support in import/export
1540 - Pause DRBD synchronization while wiping disks on instance creation
1541 - Updated unittests and QA scripts
1542 - Improved network parameters passed to KVM
1543 - Converted man pages from docbook to reStructuredText
1549 *(Released Mon, 20 Dec 2010)*
1551 Released version 2.3.1~rc1 without any changes.
1557 *(Released Wed, 1 Dec 2010)*
1559 - impexpd: Disable OpenSSL compression in socat if possible (backport
1560 from master, commit e90739d625b, see :doc:`installation guide
1561 <install-quick>` for details)
1562 - Changed unittest coverage report to exclude test scripts
1563 - Added script to check version format
1569 *(Released Wed, 1 Dec 2010)*
1571 Released version 2.3.0~rc1 without any changes.
1577 *(Released Fri, 19 Nov 2010)*
1579 A number of bugfixes and documentation updates:
1581 - Update ganeti-os-interface documentation
1582 - Fixed a bug related to duplicate MACs or similar items which should be
1584 - Fix breakage in OS state modify
1585 - Reinstall instance: disallow offline secondaries (fixes bug related to
1586 OS changing but reinstall failing)
1587 - plus all the other fixes between 2.2.1 and 2.2.2
1593 *(Released Tue, 2 Nov 2010)*
1595 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
1596 - Fixed master failover race with watcher
1597 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
1599 - Accept previously stopped instance for export with instance removal
1600 - Simplify and extend the environment variables for instance OS scripts
1601 - Added new node flags, ``master_capable`` and ``vm_capable``
1602 - Added optional instance disk wiping prior during allocation. This is a
1603 cluster-wide option and can be set/modified using
1604 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
1605 - Added IPv6 support, see :doc:`design document <design-2.3>` and
1606 :doc:`install-quick`
1607 - Added a new watcher option (``--ignore-pause``)
1608 - Added option to ignore offline node on instance start/stop
1609 (``--ignore-offline``)
1610 - Allow overriding OS parameters with ``gnt-instance reinstall``
1611 - Added ability to change node's secondary IP address using ``gnt-node
1613 - Implemented privilege separation for all daemons except
1614 ``ganeti-noded``, see ``configure`` options
1615 - Complain if an instance's disk is marked faulty in ``gnt-cluster
1617 - Implemented job priorities (see ``ganeti(7)`` manpage)
1618 - Ignore failures while shutting down instances during failover from
1620 - Exit daemon's bootstrap process only once daemon is ready
1621 - Export more information via ``LUInstanceQuery``/remote API
1622 - Improved documentation, QA and unittests
1623 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
1624 restart if the file was created or changed
1625 - Added LUXI protocol version sent with each request and response,
1626 allowing detection of server/client mismatches
1627 - Moved the Python scripts among gnt-* and ganeti-* into modules
1628 - Moved all code related to setting up SSH to an external script,
1630 - Infrastructure changes for node group support in future versions
1636 *(Released Fri, 19 Nov 2010)*
1638 A few small bugs fixed, and some improvements to the build system:
1640 - Fix documentation regarding conversion to drbd
1641 - Fix validation of parameters in cluster modify (``gnt-cluster modify
1643 - Fix error handling in node modify with multiple changes
1644 - Allow remote imports without checked names
1650 *(Released Tue, 19 Oct 2010)*
1652 - Disable SSL session ID cache in RPC client
1658 *(Released Thu, 14 Oct 2010)*
1660 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
1661 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
1663 - Fix problems with interaction between Curl and Python's HTTP server,
1664 resulting in increased speed in many RPC calls
1665 - Improve our release script to prevent breakage with older aclocal and
1672 *(Released Thu, 7 Oct 2010)*
1674 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
1675 value retrieved from master
1676 - Added support for blacklisted or hidden OS definitions
1677 - Added simple lock monitor (accessible via (``gnt-debug locks``)
1678 - Added support for -mem-path in KVM hypervisor abstraction layer
1679 - Allow overriding instance parameters in tool for inter-cluster
1680 instance moves (``tools/move-instance``)
1681 - Improved opcode summaries (e.g. in ``gnt-job list``)
1682 - Improve consistency of OS listing by sorting it
1683 - Documentation updates
1689 *(Released Fri, 8 Oct 2010)*
1691 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
1697 *(Released Mon, 4 Oct 2010)*
1699 - Fixed regression in ``gnt-instance rename``
1705 *(Released Wed, 22 Sep 2010)*
1707 - Fixed OS_VARIANT variable for OS scripts
1708 - Fixed cluster tag operations via RAPI
1709 - Made ``setup-ssh`` exit with non-zero code if an error occurred
1710 - Disabled RAPI CA checks in watcher
1716 *(Released Mon, 23 Aug 2010)*
1718 - Support DRBD versions of the format "a.b.c.d"
1720 - Re-introduce support for usage from multiple threads in RAPI client
1721 - Instance renames and modify via RAPI
1722 - Work around race condition between processing and archival in job
1724 - Mark opcodes following failed one as failed, too
1725 - Job field ``lock_status`` was removed due to difficulties making it
1726 work with the changed job queue in Ganeti 2.2; a better way to monitor
1727 locks is expected for a later 2.2.x release
1728 - Fixed dry-run behaviour with many commands
1729 - Support ``ssh-agent`` again when adding nodes
1730 - Many additional bugfixes
1736 *(Released Fri, 30 Jul 2010)*
1738 Important change: the internal RPC mechanism between Ganeti nodes has
1739 changed from using a home-grown http library (based on the Python base
1740 libraries) to use the PycURL library. This requires that PycURL is
1741 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
1742 against GnuTLS by default. cURL's support for GnuTLS had known issues
1743 before cURL 7.21.0 and we recommend using the latest cURL release or
1744 linking against OpenSSL. Most other distributions already link PycURL
1745 and cURL against OpenSSL. The command::
1747 python -c 'import pycurl; print pycurl.version'
1749 can be used to determine the libraries PycURL and cURL are linked
1752 Other significant changes:
1754 - Rewrote much of the internals of the job queue, in order to achieve
1755 better parallelism; this decouples job query operations from the job
1756 processing, and it should allow much nicer behaviour of the master
1757 daemon under load, and it also has uncovered some long-standing bugs
1758 related to the job serialisation (now fixed)
1759 - Added a default iallocator setting to the cluster parameters,
1760 eliminating the need to always pass nodes or an iallocator for
1761 operations that require selection of new node(s)
1762 - Added experimental support for the LXC virtualization method
1763 - Added support for OS parameters, which allows the installation of
1764 instances to pass parameter to OS scripts in order to customise the
1766 - Added a hypervisor parameter controlling the migration type (live or
1767 non-live), since hypervisors have various levels of reliability; this
1768 has renamed the 'live' parameter to 'mode'
1769 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
1770 logical volumes, meaning that cluster verify will ignore them and not
1771 flag their presence as errors
1772 - The watcher will now reset the error count for failed instances after
1773 8 hours, thus allowing self-healing if the problem that caused the
1774 instances to be down/fail to start has cleared in the meantime
1775 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
1776 check for, and warn, if the drbd module parameter ``usermode_helper``
1777 is not consistent with the cluster-wide setting; this is needed to
1778 make diagnose easier of failed drbd creations
1779 - Started adding base IPv6 support, but this is not yet
1780 enabled/available for use
1781 - Rename operations (cluster, instance) will now return the new name,
1782 which is especially useful if a short name was passed in
1783 - Added support for instance migration in RAPI
1784 - Added a tool to pre-configure nodes for the SSH setup, before joining
1785 them to the cluster; this will allow in the future a simplified model
1786 for node joining (but not yet fully enabled in 2.2); this needs the
1787 paramiko python library
1788 - Fixed handling of name-resolving errors
1789 - Fixed consistency of job results on the error path
1790 - Fixed master-failover race condition when executed multiple times in
1792 - Fixed many bugs related to the job queue (mostly introduced during the
1793 2.2 development cycle, so not all are impacting 2.1)
1794 - Fixed instance migration with missing disk symlinks
1795 - Fixed handling of unknown jobs in ``gnt-job archive``
1796 - And many other small fixes/improvements
1800 - Enhanced both the unittest and the QA coverage
1801 - Switched the opcode validation to a generic model, and extended the
1802 validation to all opcode parameters
1803 - Changed more parts of the code that write shell scripts to use the
1805 - Switched the master daemon to use the asyncore library for the Luxi
1812 *(Released Thu, 17 Jun 2010)*
1814 - Added tool (``move-instance``) and infrastructure to move instances
1815 between separate clusters (see :doc:`separate documentation
1816 <move-instance>` and :doc:`design document <design-2.2>`)
1817 - Added per-request RPC timeout
1818 - RAPI now requires a Content-Type header for requests with a body (e.g.
1819 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
1820 :rfc:`2616` (HTTP/1.1), section 7.2.1)
1821 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
1823 - Implemented initial support for running Ganeti daemons as separate
1824 users, see configure-time flags ``--with-user-prefix`` and
1825 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
1827 - Instances can be removed after export (``gnt-backup export
1828 --remove-instance``)
1829 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
1830 key (instead of 1024 bit)
1831 - Added new cluster configuration file for cluster domain secret
1832 - Import/export now use SSL instead of SSH
1833 - Added support for showing estimated time when exporting an instance,
1834 see the ``ganeti-os-interface(7)`` manpage and look for
1841 *(Released Tue, 16 Nov 2010)*
1843 Some more bugfixes. Unless critical bugs occur, this will be the last
1846 - Fix case of MAC special-values
1847 - Fix mac checker regex
1848 - backend: Fix typo causing "out of range" error
1849 - Add missing --units in gnt-instance list man page
1855 *(Released Tue, 24 Aug 2010)*
1858 - Don't ignore secondary node silently on non-mirrored disk templates
1860 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
1861 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
1862 - Properly document the usb_mouse parameter
1863 - Fix path in ganeti-rapi(8) (issue 116)
1864 - Adjust error message when the ganeti user's .ssh directory is
1866 - Add same-node-check when changing the disk template to drbd
1872 *(Released Fri, 16 Jul 2010)*
1875 - Add an option to only select some reboot types during qa/burnin.
1876 (on some hypervisors consequent reboots are not supported)
1877 - Fix infrequent race condition in master failover. Sometimes the old
1878 master ip address would be still detected as up for a short time
1879 after it was removed, causing failover to fail.
1880 - Decrease mlockall warnings when the ctypes module is missing. On
1881 Python 2.4 we support running even if no ctypes module is installed,
1882 but we were too verbose about this issue.
1883 - Fix building on old distributions, on which man doesn't have a
1885 - Fix RAPI not to ignore the MAC address on instance creation
1886 - Implement the old instance creation format in the RAPI client.
1892 *(Released Thu, 01 Jul 2010)*
1894 A small bugfix release:
1895 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
1896 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
1897 a non-existing option
1898 - Fix instance create: the --force-variant option was ignored
1899 - Improve pylint 0.21 compatibility and warnings with Python 2.6
1900 - Fix modify node storage with non-FQDN arguments
1901 - Fix RAPI client to authenticate under Python 2.6 when used
1902 for more than 5 requests needing authentication
1903 - Fix gnt-instance modify -t (storage) giving a wrong error message
1904 when converting a non-shutdown drbd instance to plain
1910 *(Released Fri, 18 Jun 2010)*
1912 A small bugfix release:
1914 - Fix live migration of KVM instances started with older Ganeti
1915 versions which had fewer hypervisor parameters
1916 - Fix gnt-instance grow-disk on down instances
1917 - Fix an error-reporting bug during instance migration
1918 - Better checking of the ``--net`` and ``--disk`` values, to avoid
1919 silently ignoring broken ones
1920 - Fix an RPC error reporting bug affecting, for example, RAPI client
1922 - Fix bug triggered by different API version os-es on different nodes
1923 - Fix a bug in instance startup with custom hvparams: OS level
1924 parameters would fail to be applied.
1925 - Fix the RAPI client under Python 2.6 (but more work is needed to
1926 make it work completely well with OpenSSL)
1927 - Fix handling of errors when resolving names from DNS
1933 *(Released Thu, 3 Jun 2010)*
1935 A medium sized development cycle. Some new features, and some
1936 fixes/small improvements/cleanups.
1938 Significant features
1939 ~~~~~~~~~~~~~~~~~~~~
1941 The node deamon now tries to mlock itself into memory, unless the
1942 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
1943 its logs, and falls back to console logging. This allows emergency
1944 features such as ``gnt-node powercycle`` to work even in the event of a
1945 broken node disk (tested offlining the disk hosting the node's
1946 filesystem and dropping its memory caches; don't try this at home)
1948 KVM: add vhost-net acceleration support. It can be tested with a new
1949 enough version of the kernel and of qemu-kvm.
1951 KVM: Add instance chrooting feature. If you use privilege dropping for
1952 your VMs you can also now force them to chroot to an empty directory,
1953 before starting the emulated guest.
1955 KVM: Add maximum migration bandwith and maximum downtime tweaking
1956 support (requires a new-enough version of qemu-kvm).
1958 Cluster verify will now warn if the master node doesn't have the master
1959 ip configured on it.
1961 Add a new (incompatible) instance creation request format to RAPI which
1962 supports all parameters (previously only a subset was supported, and it
1963 wasn't possible to extend the old format to accomodate all the new
1964 features. The old format is still supported, and a client can check for
1965 this feature, before using it, by checking for its presence in the
1966 ``features`` RAPI resource.
1968 Now with ancient latin support. Try it passing the ``--roman`` option to
1969 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
1970 (requires the python-roman module to be installed, in order to work).
1975 As usual many internal code refactorings, documentation updates, and
1978 - Lots of improvements and cleanups to the experimental Remote API
1979 (RAPI) client library.
1980 - A new unit test suite for the core daemon libraries.
1981 - A fix to creating missing directories makes sure the umask is not
1982 applied anymore. This enforces the same directory permissions
1984 - Better handling terminating daemons with ctrl+c (used when running
1985 them in debugging mode).
1986 - Fix a race condition in live migrating a KVM instance, when stat()
1987 on the old proc status file returned EINVAL, which is an unexpected
1989 - Fixed manpage checking with newer man and utf-8 charachters. But now
1990 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
1996 *(Released Fri, 7 May 2010)*
1998 Fix a bug which prevented untagged KVM instances from starting.
2004 *(Released Fri, 7 May 2010)*
2006 Another release with a long development cycle, during which many
2007 different features were added.
2009 Significant features
2010 ~~~~~~~~~~~~~~~~~~~~
2012 The KVM hypervisor now can run the individual instances as non-root, to
2013 reduce the impact of a VM being hijacked due to bugs in the
2014 hypervisor. It is possible to run all instances as a single (non-root)
2015 user, to manually specify a user for each instance, or to dynamically
2016 allocate a user out of a cluster-wide pool to each instance, with the
2017 guarantee that no two instances will run under the same user ID on any
2020 An experimental RAPI client library, that can be used standalone
2021 (without the other Ganeti libraries), is provided in the source tree as
2022 ``lib/rapi/client.py``. Note this client might change its interface in
2023 the future, as we iterate on its capabilities.
2025 A new command, ``gnt-cluster renew-crypto`` has been added to easily
2026 replace the cluster's certificates and crypto keys. This might help in
2027 case they have been compromised, or have simply expired.
2029 A new disk option for instance creation has been added that allows one
2030 to "adopt" currently existing logical volumes, with data
2031 preservation. This should allow easier migration to Ganeti from
2032 unmanaged (or managed via other software) instances.
2034 Another disk improvement is the possibility to convert between redundant
2035 (DRBD) and plain (LVM) disk configuration for an instance. This should
2036 allow better scalability (starting with one node and growing the
2037 cluster, or shrinking a two-node cluster to one node).
2039 A new feature that could help with automated node failovers has been
2040 implemented: if a node sees itself as offline (by querying the master
2041 candidates), it will try to shutdown (hard) all instances and any active
2042 DRBD devices. This reduces the risk of duplicate instances if an
2043 external script automatically failovers the instances on such nodes. To
2044 enable this, the cluster parameter ``maintain_node_health`` should be
2045 enabled; in the future this option (per the name) will enable other
2046 automatic maintenance features.
2048 Instance export/import now will reuse the original instance
2049 specifications for all parameters; that means exporting an instance,
2050 deleting it and the importing it back should give an almost identical
2051 instance. Note that the default import behaviour has changed from
2052 before, where it created only one NIC; now it recreates the original
2055 Cluster verify has added a few new checks: SSL certificates validity,
2056 /etc/hosts consistency across the cluster, etc.
2061 As usual, many internal changes were done, documentation fixes,
2064 - Fixed cluster initialization with disabled cluster storage (regression
2065 introduced in 2.1.1)
2066 - File-based storage supports growing the disks
2067 - Fixed behaviour of node role changes
2068 - Fixed cluster verify for some corner cases, plus a general rewrite of
2069 cluster verify to allow future extension with more checks
2070 - Fixed log spamming by watcher and node daemon (regression introduced
2072 - Fixed possible validation issues when changing the list of enabled
2074 - Fixed cleanup of /etc/hosts during node removal
2075 - Fixed RAPI response for invalid methods
2076 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
2077 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
2078 from ide disks, etc.)
2079 - Allow OS changes without re-installation (to record a changed OS
2080 outside of Ganeti, or to allow OS renames)
2081 - Allow instance creation without OS installation (useful for example if
2082 the OS will be installed manually, or restored from a backup not in
2084 - Implemented option to make cluster ``copyfile`` use the replication
2086 - Added list of enabled hypervisors to ssconf (possibly useful for
2088 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
2090 - A partial form of node re-IP is possible via node readd, which now
2091 allows changed node primary IP
2092 - Command line utilities now show an informational message if the job is
2094 - The logs of the master daemon now show the PID/UID/GID of the
2101 *(Released Fri, 12 Mar 2010)*
2103 During the 2.1.0 long release candidate cycle, a lot of improvements and
2104 changes have accumulated with were released later as 2.1.1.
2109 The node evacuate command (``gnt-node evacuate``) was significantly
2110 rewritten, and as such the IAllocator protocol was changed - a new
2111 request type has been added. This unfortunate change during a stable
2112 series is designed to improve performance of node evacuations; on
2113 clusters with more than about five nodes and which are well-balanced,
2114 evacuation should proceed in parallel for all instances of the node
2115 being evacuated. As such, any existing IAllocator scripts need to be
2116 updated, otherwise the above command will fail due to the unknown
2117 request. The provided "dumb" allocator has not been updated; but the
2118 ganeti-htools package supports the new protocol since version 0.2.4.
2120 Another important change is increased validation of node and instance
2121 names. This might create problems in special cases, if invalid host
2122 names are being used.
2124 Also, a new layer of hypervisor parameters has been added, that sits at
2125 OS level between the cluster defaults and the instance ones. This allows
2126 customisation of virtualization parameters depending on the installed
2127 OS. For example instances with OS 'X' may have a different KVM kernel
2128 (or any other parameter) than the cluster defaults. This is intended to
2129 help managing a multiple OSes on the same cluster, without manual
2130 modification of each instance's parameters.
2132 A tool for merging clusters, ``cluster-merge``, has been added in the
2133 tools sub-directory.
2138 - Improved the int/float conversions that should make the code more
2139 robust in face of errors from the node daemons
2140 - Fixed the remove node code in case of internal configuration errors
2141 - Fixed the node daemon behaviour in face of inconsistent queue
2142 directory (e.g. read-only file-system where we can't open the files
2144 - Fixed the behaviour of gnt-node modify for master candidate demotion;
2145 now it either aborts cleanly or, if given the new "auto_promote"
2146 parameter, will automatically promote other nodes as needed
2147 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
2148 completely prevent Ganeti from working
2149 - Fixed bug for instance export when not all disks were successfully
2151 - Fixed behaviour of node add when the new node is slow in starting up
2153 - Fixed handling of signals in the LUXI client, which should improve
2154 behaviour of command-line scripts
2155 - Added checks for invalid node/instance names in the configuration (now
2156 flagged during cluster verify)
2157 - Fixed watcher behaviour for disk activation errors
2158 - Fixed two potentially endless loops in http library, which led to the
2159 RAPI daemon hanging and consuming 100% CPU in some cases
2160 - Fixed bug in RAPI daemon related to hashed passwords
2161 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
2163 - Enhanced compatibility with non-Debian OSes, but not using absolute
2164 path in some commands and allowing customisation of the ssh
2165 configuration directory
2166 - Fixed possible future issue with new Python versions by abiding to the
2167 proper use of ``__slots__`` attribute on classes
2168 - Added checks that should prevent directory traversal attacks
2169 - Many documentation fixes based on feedback from users
2174 - Added an "early_release" more for instance replace disks and node
2175 evacuate, where we release locks earlier and thus allow higher
2176 parallelism within the cluster
2177 - Added watcher hooks, intended to allow the watcher to restart other
2178 daemons (e.g. from the ganeti-nbma project), but they can be used of
2179 course for any other purpose
2180 - Added a compile-time disable for DRBD barriers, to increase
2181 performance if the administrator trusts the power supply or the
2182 storage system to not lose writes
2183 - Added the option of using syslog for logging instead of, or in
2184 addition to, Ganeti's own log files
2185 - Removed boot restriction for paravirtual NICs for KVM, recent versions
2186 can indeed boot from a paravirtual NIC
2187 - Added a generic debug level for many operations; while this is not
2188 used widely yet, it allows one to pass the debug value all the way to
2190 - Enhanced the hooks environment for instance moves (failovers,
2191 migrations) where the primary/secondary nodes changed during the
2192 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
2193 - Enhanced data validations for many user-supplied values; one important
2194 item is the restrictions imposed on instance and node names, which
2195 might reject some (invalid) host names
2196 - Add a configure-time option to disable file-based storage, if it's not
2197 needed; this allows greater security separation between the master
2198 node and the other nodes from the point of view of the inter-node RPC
2200 - Added user notification in interactive tools if job is waiting in the
2201 job queue or trying to acquire locks
2202 - Added log messages when a job is waiting for locks
2203 - Added filtering by node tags in instance operations which admit
2204 multiple instances (start, stop, reboot, reinstall)
2205 - Added a new tool for cluster mergers, ``cluster-merge``
2206 - Parameters from command line which are of the form ``a=b,c=d`` can now
2207 use backslash escapes to pass in values which contain commas,
2208 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
2210 - For KVM, the instance name is the first parameter passed to KVM, so
2211 that it's more visible in the process list
2217 *(Released Tue, 2 Mar 2010)*
2219 Ganeti 2.1 brings many improvements with it. Major changes:
2221 - Added infrastructure to ease automated disk repairs
2222 - Added new daemon to export configuration data in a cheaper way than
2223 using the remote API
2224 - Instance NICs can now be routed instead of being associated with a
2226 - Improved job locking logic to reduce impact of jobs acquiring multiple
2227 locks waiting for other long-running jobs
2229 In-depth implementation details can be found in the Ganeti 2.1 design
2235 - Added chroot hypervisor
2236 - Added more options to xen-hvm hypervisor (``kernel_path`` and
2238 - Added more options to xen-pvm hypervisor (``use_bootloader``,
2239 ``bootloader_path`` and ``bootloader_args``)
2240 - Added the ``use_localtime`` option for the xen-hvm and kvm
2241 hypervisors, and the default value for this has changed to false (in
2242 2.0 xen-hvm always enabled it)
2243 - Added luxi call to submit multiple jobs in one go
2244 - Added cluster initialization option to not modify ``/etc/hosts``
2246 - Added network interface parameters
2247 - Added dry run mode to some LUs
2248 - Added RAPI resources:
2250 - ``/2/instances/[instance_name]/info``
2251 - ``/2/instances/[instance_name]/replace-disks``
2252 - ``/2/nodes/[node_name]/evacuate``
2253 - ``/2/nodes/[node_name]/migrate``
2254 - ``/2/nodes/[node_name]/role``
2255 - ``/2/nodes/[node_name]/storage``
2256 - ``/2/nodes/[node_name]/storage/modify``
2257 - ``/2/nodes/[node_name]/storage/repair``
2259 - Added OpCodes to evacuate or migrate all instances on a node
2260 - Added new command to list storage elements on nodes (``gnt-node
2261 list-storage``) and modify them (``gnt-node modify-storage``)
2262 - Added new ssconf files with master candidate IP address
2263 (``ssconf_master_candidates_ips``), node primary IP address
2264 (``ssconf_node_primary_ips``) and node secondary IP address
2265 (``ssconf_node_secondary_ips``)
2266 - Added ``ganeti-confd`` and a client library to query the Ganeti
2267 configuration via UDP
2268 - Added ability to run hooks after cluster initialization and before
2270 - Added automatic mode for disk replace (``gnt-instance replace-disks
2272 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
2273 after catastrophic data-loss
2274 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
2276 - Added ``gnt-instance move`` command to move instances
2277 - Added ``gnt-cluster watcher`` command to control watcher
2278 - Added ``gnt-node powercycle`` command to powercycle nodes
2279 - Added new job status field ``lock_status``
2280 - Added parseable error codes to cluster verification (``gnt-cluster
2281 verify --error-codes``) and made output less verbose (use
2282 ``--verbose`` to restore previous behaviour)
2283 - Added UUIDs to the main config entities (cluster, nodes, instances)
2284 - Added support for OS variants
2285 - Added support for hashed passwords in the Ganeti remote API users file
2287 - Added option to specify maximum timeout on instance shutdown
2288 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
2289 - Added new helper script to start and stop Ganeti daemons
2290 (``daemon-util``), with the intent to reduce the work necessary to
2291 adjust Ganeti for non-Debian distributions and to start/stop daemons
2293 - Added more unittests
2294 - Fixed critical bug in ganeti-masterd startup
2295 - Removed the configure-time ``kvm-migration-port`` parameter, this is
2296 now customisable at the cluster level for both the KVM and Xen
2297 hypervisors using the new ``migration_port`` parameter
2298 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
2299 reinstalling an instance
2300 - Allowed ``@`` in tag names
2301 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
2302 - Many documentation updates
2303 - Distribute hypervisor files on ``gnt-cluster redist-conf``
2304 - ``gnt-instance reinstall`` can now reinstall multiple instances
2305 - Updated many command line parameters
2306 - Introduced new OS API version 15
2307 - No longer support a default hypervisor
2308 - Treat virtual LVs as inexistent
2309 - Improved job locking logic to reduce lock contention
2310 - Match instance and node names case insensitively
2311 - Reimplemented bash completion script to be more complete
2318 *(Released Thu, 4 Feb 2010)*
2320 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
2321 - Fix a string formatting bug
2322 - Improve safety of the code in some error paths
2323 - Improve data validation in the master of values returned from nodes
2329 *(Released Thu, 17 Dec 2009)*
2331 - Fix security issue due to missing validation of iallocator names; this
2332 allows local and remote execution of arbitrary executables
2333 - Fix failure of gnt-node list during instance removal
2334 - Ship the RAPI documentation in the archive
2340 *(Released Wed, 30 Sep 2009)*
2342 - Fixed many wrong messages
2343 - Fixed a few bugs related to the locking library
2344 - Fixed MAC checking at instance creation time
2345 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
2346 - Fixed a few issues related to signal handling in both daemons and
2348 - Fixed the example startup script provided
2349 - Fixed insserv dependencies in the example startup script (patch from
2351 - Fixed handling of drained nodes in the iallocator framework
2352 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
2354 - Fixed error related to invalid job IDs in job polling
2355 - Fixed job/opcode persistence on unclean master shutdown
2356 - Fixed handling of partial job processing after unclean master
2358 - Fixed error reporting from LUs, previously all errors were converted
2359 into execution errors
2360 - Fixed error reporting from burnin
2361 - Decreased significantly the memory usage of the job queue
2362 - Optimised slightly multi-job submission
2363 - Optimised slightly opcode loading
2364 - Backported the multi-job submit framework from the development
2365 branch; multi-instance start and stop should be faster
2366 - Added script to clean archived jobs after 21 days; this will reduce
2367 the size of the queue directory
2368 - Added some extra checks in disk size tracking
2369 - Added an example ethers hook script
2370 - Added a cluster parameter that prevents Ganeti from modifying of
2372 - Added more node information to RAPI responses
2373 - Added a ``gnt-job watch`` command that allows following the ouput of a
2375 - Added a bind-address option to ganeti-rapi
2376 - Added more checks to the configuration verify
2377 - Enhanced the burnin script such that some operations can be retried
2379 - Converted instance reinstall to multi-instance model
2385 *(Released Fri, 7 Aug 2009)*
2387 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
2388 to allow using the pre-2.0.2 behaviour in activation, if any existing
2389 instances have mismatched disk sizes in the configuration
2390 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
2391 any configuration mismatches for disk sizes
2392 - Added ``gnt-master cluste-failover --no-voting`` to allow master
2393 failover to work on two-node clusters
2394 - Fixed the ``--net`` option of ``gnt-backup import``, which was
2396 - Fixed detection of OS script errors in ``gnt-backup export``
2397 - Fixed exit code of ``gnt-backup export``
2403 *(Released Fri, 17 Jul 2009)*
2405 - Added experimental support for stripped logical volumes; this should
2406 enhance performance but comes with a higher complexity in the block
2407 device handling; stripping is only enabled when passing
2408 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
2409 affected even in the non-stripped mode
2410 - Improved resiliency against transient failures at the end of DRBD
2411 resyncs, and in general of DRBD resync checks
2412 - Fixed a couple of issues with exports and snapshot errors
2413 - Fixed a couple of issues in instance listing
2414 - Added display of the disk size in ``gnt-instance info``
2415 - Fixed checking for valid OSes in instance creation
2416 - Fixed handling of the "vcpus" parameter in instance listing and in
2417 general of invalid parameters
2418 - Fixed http server library, and thus RAPI, to handle invalid
2419 username/password combinations correctly; this means that now they
2420 report unauthorized for queries too, not only for modifications,
2421 allowing earlier detect of configuration problems
2422 - Added a new "role" node list field, equivalent to the master/master
2423 candidate/drained/offline flags combinations
2424 - Fixed cluster modify and changes of candidate pool size
2425 - Fixed cluster verify error messages for wrong files on regular nodes
2426 - Fixed a couple of issues with node demotion from master candidate role
2427 - Fixed node readd issues
2428 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
2429 - Added a new ``--no-voting`` option for masterfailover to fix failover
2430 on two-nodes clusters when the former master node is unreachable
2431 - Added instance reinstall over RAPI
2437 *(Released Tue, 16 Jun 2009)*
2439 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
2440 allow re-adding the start in single-user option (regression from 1.2)
2441 - the watcher writes the instance status to a file, to allow monitoring
2442 to report the instance status (from the master) based on cached
2443 results of the watcher's queries; while this can get stale if the
2444 watcher is being locked due to other work on the cluster, this is
2445 still an improvement
2446 - the watcher now also restarts the node daemon and the rapi daemon if
2448 - fixed the watcher to handle full and drained queue cases
2449 - hooks export more instance data in the environment, which helps if
2450 hook scripts need to take action based on the instance's properties
2451 (no longer need to query back into ganeti)
2452 - instance failovers when the instance is stopped do not check for free
2453 RAM, so that failing over a stopped instance is possible in low memory
2455 - rapi uses queries for tags instead of jobs (for less job traffic), and
2456 for cluster tags it won't talk to masterd at all but read them from
2458 - a couple of error handling fixes in RAPI
2459 - drbd handling: improved the error handling of inconsistent disks after
2460 resync to reduce the frequency of "there are some degraded disks for
2461 this instance" messages
2462 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
2463 error handling path called a wrong function name)
2469 *(Released Wed, 27 May 2009)*
2471 - no changes from rc5
2477 *(Released Wed, 20 May 2009)*
2479 - fix a couple of bugs (validation, argument checks)
2480 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
2481 - some small improvements to RAPI and IAllocator
2482 - make watcher automatically start the master daemon if down
2488 *(Released Mon, 27 Apr 2009)*
2490 - change the OS list to not require locks; this helps with big clusters
2491 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
2492 volume group is broken
2493 - ``gnt-instance info``, without any arguments, doesn't run for all
2494 instances anymore; either pass ``--all`` or pass the desired
2495 instances; this helps against mistakes on big clusters where listing
2496 the information for all instances takes a long time
2497 - miscellaneous doc and man pages fixes
2503 *(Released Wed, 8 Apr 2009)*
2505 - Change the internal locking model of some ``gnt-node`` commands, in
2506 order to reduce contention (and blocking of master daemon) when
2507 batching many creation/reinstall jobs
2508 - Fixes to Xen soft reboot
2509 - No longer build documentation at build time, instead distribute it in
2510 the archive, in order to reduce the need for the whole docbook/rst
2517 *(Released Fri, 27 Mar 2009)*
2519 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
2520 - Fix watcher startup sequence, improves the behaviour of busy clusters
2521 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
2522 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
2523 bind address and other places
2524 - Some documentation fixes and updates
2530 *(Released Mon, 2 Mar 2009)*
2532 - More documentation updates, now all docs should be more-or-less
2534 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
2536 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
2538 - Fix an issue related to $libdir/run/ganeti and cluster creation
2544 *(Released Thu, 19 Feb 2009)*
2546 - Xen PVM and KVM have switched the default value for the instance root
2547 disk to the first partition on the first drive, instead of the whole
2548 drive; this means that the OS installation scripts must be changed
2550 - Man pages have been updated
2551 - RAPI has been switched by default to HTTPS, and the exported functions
2552 should all work correctly
2553 - RAPI v1 has been removed
2554 - Many improvements to the KVM hypervisor
2555 - Block device errors are now better reported
2556 - Many other bugfixes and small improvements
2562 *(Released Mon, 26 Jan 2009)*
2564 - Version 2 is a general rewrite of the code and therefore the
2565 differences are too many to list, see the design document for 2.0 in
2566 the ``doc/`` subdirectory for more details
2567 - In this beta version there is not yet a migration path from 1.2 (there
2568 will be one in the final 2.0 release)
2569 - A few significant changes are:
2571 - all commands are executed by a daemon (``ganeti-masterd``) and the
2572 various ``gnt-*`` commands are just front-ends to it
2573 - all the commands are entered into, and executed from a job queue,
2574 see the ``gnt-job(8)`` manpage
2575 - the RAPI daemon supports read-write operations, secured by basic
2576 HTTP authentication on top of HTTPS
2577 - DRBD version 0.7 support has been removed, DRBD 8 is the only
2578 supported version (when migrating from Ganeti 1.2 to 2.0, you need
2579 to migrate to DRBD 8 first while still running Ganeti 1.2)
2580 - DRBD devices are using statically allocated minor numbers, which
2581 will be assigned to existing instances during the migration process
2582 - there is support for both Xen PVM and Xen HVM instances running on
2584 - KVM virtualization is supported too
2585 - file-based storage has been implemented, which means that it is
2586 possible to run the cluster without LVM and DRBD storage, for
2587 example using a shared filesystem exported from shared storage (and
2588 still have live migration)
2594 *(Released Tue, 13 Jan 2009)*
2596 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
2597 - Reuse the old instance mac address by default on instance import, if
2598 the instance name is the same.
2599 - Handle situations in which the node info rpc returns incomplete
2601 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
2602 - Improved version of batcher:
2604 - state file support
2605 - instance mac address support
2606 - support for HVM clusters/instances
2608 - Add an option to show the number of cpu sockets and nodes in
2610 - Support OSes that handle more than one version of the OS api (but do
2611 not change the current API in any other way)
2612 - Fix ``gnt-node migrate``
2613 - ``gnt-debug`` man page
2614 - Fixes various more typos and small issues
2615 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
2621 *(Released Wed, 24 Sep 2008)*
2623 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
2624 type of disk exported to fully virtualized instances.
2625 - provide access to the serial console of HVM instances
2626 - instance auto_balance flag, set by default. If turned off it will
2627 avoid warnings on cluster verify if there is not enough memory to fail
2628 over an instance. in the future it will prevent automatically failing
2629 it over when we will support that.
2630 - batcher tool for instance creation, see ``tools/README.batcher``
2631 - ``gnt-instance reinstall --select-os`` to interactively select a new
2632 operating system when reinstalling an instance.
2633 - when changing the memory amount on instance modify a check has been
2634 added that the instance will be able to start. also warnings are
2635 emitted if the instance will not be able to fail over, if auto_balance
2637 - documentation fixes
2638 - sync fields between ``gnt-instance list/modify/add/import``
2639 - fix a race condition in drbd when the sync speed was set after giving
2640 the device a remote peer.
2646 *(Released Tue, 22 Jul 2008)*
2648 - note: the allowed size and number of tags per object were reduced
2649 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
2650 - fixed twisted 8.x compatibility
2651 - fixed ``gnt-instance replace-disks`` with iallocator
2652 - add TCP keepalives on twisted connections to detect restarted nodes
2653 - disk increase support, see ``gnt-instance grow-disk``
2654 - implement bulk node/instance query for RAPI
2655 - add tags in node/instance listing (optional)
2656 - experimental migration (and live migration) support, read the man page
2657 for ``gnt-instance migrate``
2658 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
2659 has some small improvements in handling its state file
2665 *(Released Fri, 13 Jun 2008)*
2667 - Experimental readonly, REST-based remote API implementation;
2668 automatically started on master node, TCP port 5080, if enabled by
2669 ``--enable-rapi`` parameter to configure script.
2670 - Instance allocator support. Add and import instance accept a
2671 ``--iallocator`` parameter, and call that instance allocator to decide
2672 which node to use for the instance. The iallocator document describes
2673 what's expected from an allocator script.
2674 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
2675 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
2676 node is lost there is still enough memory to fail over the instances
2678 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
2679 ``gnt-cluster verify``, to check for site-specific compliance. All the
2680 hooks will run, and their output, if any, will be displayed. Any
2681 failing hook will make the verification return an error value.
2682 - ``gnt-cluster verify`` now checks that its peers are reachable on the
2683 primary and secondary interfaces
2684 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
2685 that is still declared as part of the cluster and has failed.
2686 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
2687 specifying output fields, that just adds the chosen fields to the
2689 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
2690 export from the filesystem.
2691 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
2692 have been added. Using them you can enable/disable acpi and pae
2693 support, and specify a path for a cd image to be exported to the
2694 instance. These parameters as the name suggest only work on HVM
2696 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
2697 PAE support will be set to the previously hardcoded values, but the
2698 (previously hardcoded) path to the CDROM ISO image will be unset and
2699 if required, needs to be set manually with ``gnt-instance modify``
2701 - The address to which an instance's VNC console is bound is now
2702 selectable per-instance, rather than being cluster wide. Of course
2703 this only applies to instances controlled via VNC, so currently just
2704 applies to HVM clusters.
2710 *(Released Mon, 18 Feb 2008)*
2712 - more tweaks to the disk activation code (especially helpful for DRBD)
2713 - change the default ``gnt-instance list`` output format, now there is
2714 one combined status field (see the manpage for the exact values this
2716 - some more fixes for the mac export to hooks change
2717 - make Ganeti not break with DRBD 8.2.x (which changed the version
2718 format in ``/proc/drbd``) (issue 24)
2719 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
2720 template, allowing migration from DRBD0.7+MD to DRBD8
2726 *(Released Wed, 30 Jan 2008)*
2728 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
2730 - add command aliases infrastructure and a few aliases
2731 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
2732 man pages and the ``--help`` option of ``gnt-node
2733 list``/``gnt-instance list``
2734 - fix ``gnt-backup list`` with down nodes (issue 21)
2735 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
2736 - fix the dist archive and add a check for including svn/git files in
2738 - some developer-related changes: improve the burnin and the QA suite,
2739 add an upload script for testing during development
2745 *(Released Wed, 16 Jan 2008)*
2747 - experimental HVM support, read the install document, section
2748 "Initializing the cluster"
2749 - allow for the PVM hypervisor per-instance kernel and initrd paths
2750 - add a new command ``gnt-cluster verify-disks`` which uses a new
2751 algorithm to improve the reconnection of the DRBD pairs if the device
2752 on the secondary node has gone away
2753 - make logical volume code auto-activate LVs at disk activation time
2754 - slightly improve the speed of activating disks
2755 - allow specification of the MAC address at instance creation time, and
2756 changing it later via ``gnt-instance modify``
2757 - fix handling of external commands that generate lots of output on
2759 - update documentation with regard to minimum version of DRBD8 supported
2765 *(Released Tue, 4 Dec 2007)*
2767 - Log the ``xm create`` output to the node daemon log on failure (to
2768 help diagnosing the error)
2769 - In debug mode, log all external commands output if failed to the logs
2770 - Change parsing of lvm commands to ignore stderr
2776 *(Released Wed, 28 Nov 2007)*
2778 - Another round of updates to the DRBD 8 code to deal with more failures
2779 in the replace secondary node operation
2780 - Some more logging of failures in disk operations (lvm, drbd)
2781 - A few documentation updates
2788 *(Released Tue, 13 Nov 2007)*
2790 - Change configuration file format from Python's Pickle to JSON.
2791 Upgrading is possible using the cfgupgrade utility.
2792 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
2793 faster replace disks and is more stable (DRBD 8 has many improvements
2794 compared to DRBD 0.7)
2795 - Added command line tags support (see man pages for ``gnt-instance``,
2796 ``gnt-node``, ``gnt-cluster``)
2797 - Added instance rename support
2798 - Added multi-instance startup/shutdown
2799 - Added cluster rename support
2800 - Added ``gnt-node evacuate`` to simplify some node operations
2801 - Added instance reboot operation that can speedup reboot as compared to
2803 - Soften the requirement that hostnames are in FQDN format
2804 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
2806 - Removed dependency on debian's patched fping that uses the
2807 non-standard ``-S`` option
2808 - Now the OS definitions are searched for in multiple, configurable
2809 paths (easier for distros to package)
2810 - Some changes to the hooks infrastructure (especially the new
2811 post-configuration update hook)
2812 - Other small bugfixes
2814 .. vim: set textwidth=72 syntax=rst :