8 *(Released Thu, 29 Aug 2013)*
10 Incompatible/important changes
11 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
13 - hroller now also plans for capacity to move non-redundant instances off
14 any node to be rebooted; the old behavior of completely ignoring any
15 non-redundant instances can be restored by adding the --ignore-non-redundant
17 - The cluster option '--no-lvm-storage' was removed in favor of the new option
18 '--enabled-disk-templates'.
19 - On instance creation, disk templates no longer need to be specified
20 with '-t'. The default disk template will be taken from the list of
21 enabled disk templates.
22 - The monitoring daemon is now running as root, in order to be able to collect
23 information only available to root (such as the state of Xen instances).
24 - The ConfD client is now IPv6 compatible.
25 - File and shared file storage is no longer dis/enabled at configure time,
26 but using the option '--enabled-disk-templates' at cluster initialization and
28 - The default directories for file and shared file storage are not anymore
29 specified at configure time, but taken from the cluster's configuration.
30 They can be set at cluster initialization and modification with
31 '--file-storage-dir' and '--shared-file-storage-dir'.
32 - Cluster verification now includes stricter checks regarding the
33 default file and shared file storage directories. It now checks that
34 the directories are explicitely allowed in the 'file-storage-paths' file and
35 that the directories exist on all nodes.
36 - The list of allowed disk templates in the instance policy and the list
37 of cluster-wide enabled disk templates is now checked for consistency
38 on cluster or group modification. On cluster initialization, the ipolicy
39 disk templates are ensured to be a subset of the cluster-wide enabled
45 - DRBD 8.4 support. Depending on the installed DRBD version, Ganeti now uses
46 the correct command syntax. It is possible to use different DRBD versions
47 on different nodes as long as they are compatible to each other. This
48 enables rolling upgrades of DRBD with no downtime. As permanent operation
49 of different DRBD versions within a node group is discouraged,
50 ``gnt-cluster verify`` will emit a warning if it detects such a situation.
51 - New "inst-status-xen" data collector for the monitoring daemon, providing
52 information about the state of the xen instances on the nodes.
53 - New "lv" data collector for the monitoring daemon, collecting data about the
54 logical volumes on the nodes, and pairing them with the name of the instances
56 - New "diskstats" data collector, collecting the data from /proc/diskstats and
57 presenting them over the monitoring daemon interface.
58 - The ConfD client is now IPv6 compatible.
62 The following new dependencies have been added.
66 - ``python-mock`` (http://www.voidspace.org.uk/python/mock/) is now a required
67 for the unit tests (and only used for testing).
73 *(Released Tue, 17 Sep 2013)*
75 Incompatible/important changes
76 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
78 - Instance policy can contain multiple instance specs, as described in
79 the “Constrained instance sizes” section of :doc:`Partitioned Ganeti
80 <design-partitioned>`. As a consequence, it's not possible to partially change
81 or override instance specs. Bounding specs (min and max) can be specified as a
82 whole using the new option ``--ipolicy-bounds-specs``, while standard
83 specs use the new option ``--ipolicy-std-specs``.
84 - The output of the info command of gnt-cluster, gnt-group, gnt-node,
85 gnt-instance is a valid YAML object.
86 - hail now honors network restrictions when allocating nodes. This led to an
87 update of the IAllocator protocol. See the IAllocator documentation for
89 - confd now only answers static configuration request over the network. luxid
90 was extracted, listens on the local LUXI socket and responds to live queries.
91 This allows finer grained permissions if using separate users.
96 - The :doc:`Remote API <rapi>` daemon now supports a command line flag
97 to always require authentication, ``--require-authentication``. It can
98 be specified in ``$sysconfdir/default/ganeti``.
99 - A new cluster attribute 'enabled_disk_templates' is introduced. It will
100 be used to manage the disk templates to be used by instances in the cluster.
101 Initially, it will be set to a list that includes plain, drbd, if they were
102 enabled by specifying a volume group name, and file and sharedfile, if those
103 were enabled at configure time. Additionally, it will include all disk
104 templates that are currently used by instances. The order of disk templates
105 will be based on Ganeti's history of supporting them. In the future, the
106 first entry of the list will be used as a default disk template on instance
108 - ``cfgupgrade`` now supports a ``--downgrade`` option to bring the
109 configuration back to the previous stable version.
110 - Disk templates in group ipolicy can be restored to the default value.
111 - Initial support for diskless instances and virtual clusters in QA.
112 - More QA and unit tests for instance policies.
113 - Every opcode now contains a reason trail (visible through ``gnt-job info``)
114 describing why the opcode itself was executed.
115 - The monitoring daemon is now available. It allows users to query the cluster
116 for obtaining information about the status of the system. The daemon is only
117 responsible for providing the information over the network: the actual data
118 gathering is performed by data collectors (currently, only the DRBD status
119 collector is available).
120 - In order to help developers work on Ganeti, a new script
121 (``devel/build_chroot``) is provided, for building a chroot that contains all
122 the required development libraries and tools for compiling Ganeti on a Debian
124 - A new tool, ``harep``, for performing self-repair and recreation of instances
125 in Ganeti has been added.
126 - Split queries are enabled for tags, network, exports, cluster info, groups,
128 - New command ``show-ispecs-cmd`` for ``gnt-cluster`` and ``gnt-group``.
129 It prints the command line to set the current policies, to ease
131 - Add the ``vnet_hdr`` HV parameter for KVM, to control whether the tap
132 devices for KVM virtio-net interfaces will get created with VNET_HDR
133 (IFF_VNET_HDR) support. If set to false, it disables offloading on the
134 virtio-net interfaces, which prevents host kernel tainting and log
135 flooding, when dealing with broken or malicious virtio-net drivers.
136 It's set to true by default.
137 - Instance failover now supports a ``--cleanup`` parameter for fixing previous
139 - Support 'viridian' parameter in Xen HVM
140 - Support DSA SSH keys in bootstrap
141 - To simplify the work of packaging frameworks that want to add the needed users
142 and groups in a split-user setup themselves, at build time three files in
143 ``doc/users`` will be generated. The ``groups`` files contains, one per line,
144 the groups to be generated, the ``users`` file contains, one per line, the
145 users to be generated, optionally followed by their primary group, where
146 important. The ``groupmemberships`` file contains, one per line, additional
147 user-group membership relations that need to be established. The syntax of
148 these files will remain stable in all future versions.
153 The following new dependencies have been added:
156 - The ``curl`` library is not optional anymore for compiling the Haskell code.
157 - ``snap-server`` library (if monitoring is enabled).
160 - The minimum Python version needed to run Ganeti is now 2.6.
161 - ``yaml`` library (only for running the QA).
166 - To simplify the work of packaging frameworks that want to add the needed users
167 and groups in a split-user setup themselves, at build time three files in
168 ``doc/users`` will be generated. The ``groups`` files contains, one per line,
169 the groups to be generated, the ``users`` file contains, one per line, the
170 users to be generated, optionally followed by their primary group, where
171 important. The ``groupmemberships`` file contains, one per line, additional
172 user-group membership relations that need to be established. The syntax of
173 these files will remain stable in all future versions.
174 - Add a default to file-driver when unspecified over RAPI (Issue 571)
175 - Mark the DSA host pubkey as optional, and remove it during config downgrade
177 - Some documentation fixes
183 *(Released Tue, 27 Aug 2013)*
185 The second release candidate of the 2.8 series. Since 2.8.0. rc1:
187 - Support 'viridian' parameter in Xen HVM (Issue 233)
188 - Include VCS version in ``gnt-cluster version``
189 - Support DSA SSH keys in bootstrap (Issue 338)
190 - Fix batch creation of instances
191 - Use FQDN to check master node status (Issue 551)
192 - Make the DRBD collector more failure-resilient
198 *(Released Fri, 2 Aug 2013)*
200 The first release candidate of the 2.8 series. Since 2.8.0 beta1:
202 - Fix upgrading/downgrading from 2.7
203 - Increase maximum RAPI message size
204 - Documentation updates
205 - Split ``confd`` between ``luxid`` and ``confd``
206 - Merge 2.7 series up to the 2.7.1 release
207 - Allow the ``modify_etc_hosts`` option to be changed
208 - Add better debugging for ``luxid`` queries
209 - Expose bulk parameter for GetJobs in RAPI client
210 - Expose missing ``network`` fields in RAPI
211 - Add some ``cluster verify`` tests
212 - Some unittest fixes
213 - Fix a malfunction in ``hspace``'s tiered allocation
214 - Fix query compatibility between haskell and python implementations
215 - Add the ``vnet_hdr`` HV parameter for KVM
216 - Add ``--cleanup`` to instance failover
217 - Change the connected groups format in ``gnt-network info`` output; it
218 was previously displayed as a raw list by mistake. (Merged from 2.7)
224 *(Released Mon, 24 Jun 2013)*
226 This was the first beta release of the 2.8 series. All important changes
227 are listed in the latest 2.8 entry.
233 *(Released Thu, 25 Jul 2013)*
235 - Add logrotate functionality in daemon-util
236 - Add logrotate example file
237 - Add missing fields to network queries over rapi
238 - Fix network object timestamps
239 - Add support for querying network timestamps
240 - Fix a typo in the example crontab
241 - Fix a documentation typo
247 *(Released Thu, 04 Jul 2013)*
249 Incompatible/important changes
250 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
252 - Instance policies for disk size were documented to be on a per-disk
253 basis, but hail applied them to the sum of all disks. This has been
255 - ``hbal`` will now exit with status 0 if, during job execution over
256 LUXI, early exit has been requested and all jobs are successful;
257 before, exit status 1 was used, which cannot be differentiated from
259 - Compatibility with newer versions of rbd has been fixed
260 - ``gnt-instance batch-create`` has been changed to use the bulk create
261 opcode from Ganeti. This lead to incompatible changes in the format of
262 the JSON file. It's now not a custom dict anymore but a dict
263 compatible with the ``OpInstanceCreate`` opcode.
264 - Parent directories for file storage need to be listed in
265 ``$sysconfdir/ganeti/file-storage-paths`` now. ``cfgupgrade`` will
266 write the file automatically based on old configuration values, but it
267 can not distribute it across all nodes and the file contents should be
268 verified. Use ``gnt-cluster copyfile
269 $sysconfdir/ganeti/file-storage-paths`` once the cluster has been
270 upgraded. The reason for requiring this list of paths now is that
271 before it would have been possible to inject new paths via RPC,
272 allowing files to be created in arbitrary locations. The RPC protocol
273 is protected using SSL/X.509 certificates, but as a design principle
274 Ganeti does not permit arbitrary paths to be passed.
275 - The parsing of the variants file for OSes (see
276 :manpage:`ganeti-os-interface(7)`) has been slightly changed: now empty
277 lines and comment lines (starting with ``#``) are ignored for better
279 - The ``setup-ssh`` tool added in Ganeti 2.2 has been replaced and is no
280 longer available. ``gnt-node add`` now invokes a new tool on the
281 destination node, named ``prepare-node-join``, to configure the SSH
282 daemon. Paramiko is no longer necessary to configure nodes' SSH
283 daemons via ``gnt-node add``.
284 - Draining (``gnt-cluster queue drain``) and un-draining the job queue
285 (``gnt-cluster queue undrain``) now affects all nodes in a cluster and
286 the flag is not reset after a master failover.
287 - Python 2.4 has *not* been tested with this release. Using 2.6 or above
288 is recommended. 2.6 will be mandatory from the 2.8 series.
294 - New network management functionality to support automatic allocation
295 of IP addresses and managing of network parameters. See
296 :manpage:`gnt-network(8)` for more details.
297 - New external storage backend, to allow managing arbitrary storage
298 systems external to the cluster. See
299 :manpage:`ganeti-extstorage-interface(7)`.
300 - New ``exclusive-storage`` node parameter added, restricted to
301 nodegroup level. When it's set to true, physical disks are assigned in
302 an exclusive fashion to instances, as documented in :doc:`Partitioned
303 Ganeti <design-partitioned>`. Currently, only instances using the
304 ``plain`` disk template are supported.
305 - The KVM hypervisor has been updated with many new hypervisor
306 parameters, including a generic one for passing arbitrary command line
307 values. See a complete list in :manpage:`gnt-instance(8)`. It is now
308 compatible up to qemu 1.4.
309 - A new tool, called ``mon-collector``, is the stand-alone executor of
310 the data collectors for a monitoring system. As of this version, it
311 just includes the DRBD data collector, that can be executed by calling
312 ``mon-collector`` using the ``drbd`` parameter. See
313 :manpage:`mon-collector(7)`.
314 - A new user option, :pyeval:`rapi.RAPI_ACCESS_READ`, has been added
315 for RAPI users. It allows granting permissions to query for
316 information to a specific user without giving
317 :pyeval:`rapi.RAPI_ACCESS_WRITE` permissions.
318 - A new tool named ``node-cleanup`` has been added. It cleans remains of
319 a cluster from a machine by stopping all daemons, removing
320 certificates and ssconf files. Unless the ``--no-backup`` option is
321 given, copies of the certificates are made.
322 - Instance creations now support the use of opportunistic locking,
323 potentially speeding up the (parallel) creation of multiple instances.
324 This feature is currently only available via the :doc:`RAPI
325 <rapi>` interface and when an instance allocator is used. If the
326 ``opportunistic_locking`` parameter is set the opcode will try to
327 acquire as many locks as possible, but will not wait for any locks
328 held by other opcodes. If not enough resources can be found to
329 allocate the instance, the temporary error code
330 :pyeval:`errors.ECODE_TEMP_NORES` is returned. The operation can be
331 retried thereafter, with or without opportunistic locking.
332 - New experimental linux-ha resource scripts.
333 - Restricted-commands support: ganeti can now be asked (via command line
334 or rapi) to perform commands on a node. These are passed via ganeti
335 RPC rather than ssh. This functionality is restricted to commands
336 specified on the ``$sysconfdir/ganeti/restricted-commands`` for security
337 reasons. The file is not copied automatically.
343 - Diskless instances are now externally mirrored (Issue 237). This for
344 now has only been tested in conjunction with explicit target nodes for
346 - Queries not needing locks or RPC access to the node can now be
347 performed by the confd daemon, making them independent from jobs, and
348 thus faster to execute. This is selectable at configure time.
349 - The functionality for allocating multiple instances at once has been
350 overhauled and is now also available through :doc:`RAPI <rapi>`.
352 There are no significant changes from version 2.7.0~rc3.
358 *(Released Tue, 25 Jun 2013)*
360 - Fix permissions on the confd query socket (Issue 477)
361 - Fix permissions on the job archive dir (Issue 498)
362 - Fix handling of an internal exception in replace-disks (Issue 472)
363 - Fix gnt-node info handling of shortened names (Issue 497)
364 - Fix gnt-instance grow-disk when wiping is enabled
365 - Documentation improvements, and support for newer pandoc
366 - Fix hspace honoring ipolicy for disks (Issue 484)
367 - Improve handling of the ``kvm_extra`` HV parameter
373 *(Released Fri, 24 May 2013)*
375 - ``devel/upload`` now works when ``/var/run`` on the target nodes is a
377 - Disks added through ``gnt-instance modify`` or created through
378 ``gnt-instance recreate-disks`` are wiped, if the
379 ``prealloc_wipe_disks`` flag is set.
380 - If wiping newly created disks fails, the disks are removed. Also,
381 partial failures in creating disks through ``gnt-instance modify``
382 triggers a cleanup of the partially-created disks.
383 - Removing the master IP address doesn't fail if the address has been
385 - Fix ownership of the OS log dir
386 - Workaround missing SO_PEERCRED constant (Issue 191)
392 *(Released Fri, 3 May 2013)*
394 This was the first release candidate of the 2.7 series. Since beta3:
396 - Fix kvm compatibility with qemu 1.4 (Issue 389)
397 - Documentation updates (admin guide, upgrade notes, install
398 instructions) (Issue 372)
399 - Fix gnt-group list nodes and instances count (Issue 436)
400 - Fix compilation without non-mandatory libraries (Issue 441)
401 - Fix xen-hvm hypervisor forcing nics to type 'ioemu' (Issue 247)
402 - Make confd logging more verbose at INFO level (Issue 435)
403 - Improve "networks" documentation in :manpage:`gnt-instance(8)`
404 - Fix failure path for instance storage type conversion (Issue 229)
405 - Update htools text backend documentation
406 - Improve the renew-crypto section of :manpage:`gnt-cluster(8)`
407 - Disable inter-cluster instance move for file-based instances, because
408 it is dependant on instance export, which is not supported for
409 file-based instances. (Issue 414)
410 - Fix gnt-job crashes on non-ascii characters (Issue 427)
411 - Fix volume group checks on non-vm-capable nodes (Issue 432)
417 *(Released Mon, 22 Apr 2013)*
419 This was the third beta release of the 2.7 series. Since beta2:
421 - Fix hail to verify disk instance policies on a per-disk basis (Issue 418).
422 - Fix data loss on wrong usage of ``gnt-instance move``
423 - Properly export errors in confd-based job queries
424 - Add ``users-setup`` tool
425 - Fix iallocator protocol to report 0 as a disk size for diskless
426 instances. This avoids hail breaking when a diskless instance is
428 - Fix job queue directory permission problem that made confd job queries
429 fail. This requires running an ``ensure-dirs --full-run`` on upgrade
430 for access to archived jobs (Issue 406).
431 - Limit the sizes of networks supported by ``gnt-network`` to something
432 between a ``/16`` and a ``/30`` to prevent memory bloat and crashes.
433 - Fix bugs in instance disk template conversion
434 - Fix GHC 7 compatibility
435 - Fix ``burnin`` install path (Issue 426).
436 - Allow very small disk grows (Issue 347).
437 - Fix a ``ganeti-noded`` memory bloat introduced in 2.5, by making sure
438 that noded doesn't import masterd code (Issue 419).
439 - Make sure the default metavg at cluster init is the same as the vg, if
440 unspecified (Issue 358).
441 - Fix cleanup of partially created disks (part of Issue 416)
447 *(Released Tue, 2 Apr 2013)*
449 This was the second beta release of the 2.7 series. Since beta1:
451 - Networks no longer have a "type" slot, since this information was
452 unused in Ganeti: instead of it tags should be used.
453 - The rapi client now has a ``target_node`` option to MigrateInstance.
454 - Fix early exit return code for hbal (Issue 386).
455 - Fix ``gnt-instance migrate/failover -n`` (Issue 396).
456 - Fix ``rbd showmapped`` output parsing (Issue 312).
457 - Networks are now referenced indexed by UUID, rather than name. This
458 will require running cfgupgrade, from 2.7.0beta1, if networks are in
460 - The OS environment now includes network information.
461 - Deleting of a network is now disallowed if any instance nic is using
462 it, to prevent dangling references.
463 - External storage is now documented in man pages.
464 - The exclusive_storage flag can now only be set at nodegroup level.
465 - Hbal can now submit an explicit priority with its jobs.
466 - Many network related locking fixes.
467 - Bump up the required pylint version to 0.25.1.
468 - Fix the ``no_remember`` option in RAPI client.
469 - Many ipolicy related tests, qa, and fixes.
470 - Many documentation improvements and fixes.
471 - Fix building with ``--disable-file-storage``.
472 - Fix ``-q`` option in htools, which was broken if passed more than
474 - Some haskell/python interaction improvements and fixes.
475 - Fix iallocator in case of missing LVM storage.
476 - Fix confd config load in case of ``--no-lvm-storage``.
477 - The confd/query functionality is now mentioned in the security
484 *(Released Wed, 6 Feb 2013)*
486 This was the first beta release of the 2.7 series. All important changes
487 are listed in the latest 2.7 entry.
493 *(Released Fri, 21 Dec 2012)*
495 Important behaviour change: hbal won't rebalance anymore instances which
496 have the ``auto_balance`` attribute set to false. This was the intention
497 all along, but until now it only skipped those from the N+1 memory
498 reservation (DRBD-specific).
500 A significant number of bug fixes in this release:
502 - Fixed disk adoption interaction with ipolicy checks.
503 - Fixed networking issues when instances are started, stopped or
504 migrated, by forcing the tap device's MAC prefix to "fe" (issue 217).
505 - Fixed the warning in cluster verify for shared storage instances not
507 - Fixed removal of storage directory on shared file storage (issue 262).
508 - Fixed validation of LVM volume group name in OpClusterSetParams
509 (``gnt-cluster modify``) (issue 285).
510 - Fixed runtime memory increases (``gnt-instance modify -m``).
511 - Fixed live migration under Xen's ``xl`` mode.
512 - Fixed ``gnt-instance console`` with ``xl``.
513 - Fixed building with newer Haskell compiler/libraries.
514 - Fixed PID file writing in Haskell daemons (confd); this prevents
515 restart issues if confd was launched manually (outside of
516 ``daemon-util``) while another copy of it was running
517 - Fixed a type error when doing live migrations with KVM (issue 297) and
518 the error messages for failing migrations have been improved.
519 - Fixed opcode validation for the out-of-band commands (``gnt-node
521 - Fixed a type error when unsetting OS hypervisor parameters (issue
522 311); now it's possible to unset all OS-specific hypervisor
524 - Fixed the ``dry-run`` mode for many operations: verification of
525 results was over-zealous but didn't take into account the ``dry-run``
526 operation, resulting in "wrong" failures.
527 - Fixed bash completion in ``gnt-job list`` when the job queue has
528 hundreds of entries; especially with older ``bash`` versions, this
529 results in significant CPU usage.
531 And lastly, a few other improvements have been made:
533 - Added option to force master-failover without voting (issue 282).
534 - Clarified error message on lock conflict (issue 287).
535 - Logging of newly submitted jobs has been improved (issue 290).
536 - Hostname checks have been made uniform between instance rename and
538 - The ``--submit`` option is now supported by ``gnt-debug delay``.
539 - Shutting down the master daemon by sending SIGTERM now stops it from
540 processing jobs waiting for locks; instead, those jobs will be started
541 once again after the master daemon is started the next time (issue
543 - Support for Xen's ``xl`` program has been improved (besides the fixes
545 - Reduced logging noise in the Haskell confd daemon (only show one log
546 entry for each config reload, instead of two).
547 - Several man page updates and typo fixes.
553 *(Released Fri, 12 Oct 2012)*
555 A small bugfix release. Among the bugs fixed:
557 - Fixed double use of ``PRIORITY_OPT`` in ``gnt-node migrate``, that
558 made the command unusable.
559 - Commands that issue many jobs don't fail anymore just because some jobs
560 take so long that other jobs are archived.
561 - Failures during ``gnt-instance reinstall`` are reflected by the exit
563 - Issue 190 fixed. Check for DRBD in cluster verify is enabled only when
565 - When ``always_failover`` is set, ``--allow-failover`` is not required
566 in migrate commands anymore.
567 - ``bash_completion`` works even if extglob is disabled.
568 - Fixed bug with locks that made failover for RDB-based instances fail.
569 - Fixed bug in non-mirrored instance allocation that made Ganeti choose
570 a random node instead of one based on the allocator metric.
571 - Support for newer versions of pylint and pep8.
572 - Hail doesn't fail anymore when trying to add an instance of type
573 ``file``, ``sharedfile`` or ``rbd``.
574 - Added new Makefile target to rebuild the whole distribution, so that
575 all files are included.
581 *(Released Fri, 27 Jul 2012)*
584 .. attention:: The ``LUXI`` protocol has been made more consistent
585 regarding its handling of command arguments. This, however, leads to
586 incompatibility issues with previous versions. Please ensure that you
587 restart Ganeti daemons soon after the upgrade, otherwise most
588 ``LUXI`` calls (job submission, setting/resetting the drain flag,
589 pausing/resuming the watcher, cancelling and archiving jobs, querying
590 the cluster configuration) will fail.
599 The current ``admin_up`` field, which used to denote whether an instance
600 should be running or not, has been removed. Instead, ``admin_state`` is
601 introduced, with 3 possible values -- ``up``, ``down`` and ``offline``.
603 The rational behind this is that an instance being “down” can have
606 - it could be down during a reboot
607 - it could be temporarily be down for a reinstall
608 - or it could be down because it is deprecated and kept just for its
611 The previous Boolean state was making it difficult to do capacity
612 calculations: should Ganeti reserve memory for a down instance? Now, the
613 tri-state field makes it clear:
615 - in ``up`` and ``down`` state, all resources are reserved for the
616 instance, and it can be at any time brought up if it is down
617 - in ``offline`` state, only disk space is reserved for it, but not
620 The field can have an extra use: since the transition between ``up`` and
621 ``down`` and vice-versus is done via ``gnt-instance start/stop``, but
622 transition between ``offline`` and ``down`` is done via ``gnt-instance
623 modify``, it is possible to given different rights to users. For
624 example, owners of an instance could be allowed to start/stop it, but
625 not transition it out of the offline state.
627 Instance policies and specs
628 +++++++++++++++++++++++++++
630 In previous Ganeti versions, an instance creation request was not
631 limited on the minimum size and on the maximum size just by the cluster
632 resources. As such, any policy could be implemented only in third-party
633 clients (RAPI clients, or shell wrappers over ``gnt-*``
634 tools). Furthermore, calculating cluster capacity via ``hspace`` again
635 required external input with regards to instance sizes.
637 In order to improve these workflows and to allow for example better
638 per-node group differentiation, we introduced instance specs, which
641 - minimum instance disk size, disk count, memory size, cpu count
642 - maximum values for the above metrics
643 - and “standard” values (used in ``hspace`` to calculate the standard
646 The minimum/maximum values can be also customised at node-group level,
647 for example allowing more powerful hardware to support bigger instance
650 Beside the instance specs, there are a few other settings belonging to
651 the instance policy framework. It is possible now to customise, per
652 cluster and node-group:
654 - the list of allowed disk templates
655 - the maximum ratio of VCPUs per PCPUs (to control CPU oversubscription)
656 - the maximum ratio of instance to spindles (see below for more
657 information) for local storage
659 All these together should allow all tools that talk to Ganeti to know
660 what are the ranges of allowed values for instances and the
661 over-subscription that is allowed.
663 For the VCPU/PCPU ratio, we already have the VCPU configuration from the
664 instance configuration, and the physical CPU configuration from the
665 node. For the spindle ratios however, we didn't track before these
666 values, so new parameters have been added:
668 - a new node parameter ``spindle_count``, defaults to 1, customisable at
669 node group or node level
670 - at new backend parameter (for instances), ``spindle_use`` defaults to 1
672 Note that spindles in this context doesn't need to mean actual
673 mechanical hard-drives; it's just a relative number for both the node
674 I/O capacity and instance I/O consumption.
676 Instance migration behaviour
677 ++++++++++++++++++++++++++++
679 While live-migration is in general desirable over failover, it is
680 possible that for some workloads it is actually worse, due to the
681 variable time of the “suspend” phase during live migration.
683 To allow the tools to work consistently over such instances (without
684 having to hard-code instance names), a new backend parameter
685 ``always_failover`` has been added to control the migration/failover
686 behaviour. When set to True, all migration requests for an instance will
687 instead fall-back to failover.
689 Instance memory ballooning
690 ++++++++++++++++++++++++++
692 Initial support for memory ballooning has been added. The memory for an
693 instance is no longer fixed (backend parameter ``memory``), but instead
694 can vary between minimum and maximum values (backend parameters
695 ``minmem`` and ``maxmem``). Currently we only change an instance's
698 - live migrating or failing over and instance and the target node
699 doesn't have enough memory
700 - user requests changing the memory via ``gnt-instance modify
706 In order to control the use of specific CPUs by instance, support for
707 controlling CPU pinning has been added for the Xen, HVM and LXC
708 hypervisors. This is controlled by a new hypervisor parameter
709 ``cpu_mask``; details about possible values for this are in the
710 :manpage:`gnt-instance(8)`. Note that use of the most specific (precise
711 VCPU-to-CPU mapping) form will work well only when all nodes in your
712 cluster have the same amount of CPUs.
717 Another area in which Ganeti was not customisable were the parameters
718 used for storage configuration, e.g. how many stripes to use for LVM,
719 DRBD resync configuration, etc.
721 To improve this area, we've added disks parameters, which are
722 customisable at cluster and node group level, and which allow to
723 specify various parameters for disks (DRBD has the most parameters
724 currently), for example:
726 - DRBD resync algorithm and parameters (e.g. speed)
727 - the default VG for meta-data volumes for DRBD
728 - number of stripes for LVM (plain disk template)
731 These parameters can be modified via ``gnt-cluster modify -D …`` and
732 ``gnt-group modify -D …``, and are used at either instance creation (in
733 case of LVM stripes, for example) or at disk “activation” time
736 Rados block device support
737 ++++++++++++++++++++++++++
739 A Rados (http://ceph.com/wiki/Rbd) storage backend has been added,
740 denoted by the ``rbd`` disk template type. This is considered
741 experimental, feedback is welcome. For details on configuring it, see
742 the :doc:`install` document and the :manpage:`gnt-cluster(8)` man page.
747 The existing master IP functionality works well only in simple setups (a
748 single network shared by all nodes); however, if nodes belong to
749 different networks, then the ``/32`` setup and lack of routing
750 information is not enough.
752 To allow the master IP to function well in more complex cases, the
753 system was reworked as follows:
755 - a master IP netmask setting has been added
756 - the master IP activation/turn-down code was moved from the node daemon
758 - whether to run the Ganeti-supplied master IP script or a user-supplied
759 on is a ``gnt-cluster init`` setting
761 Details about the location of the standard and custom setup scripts are
762 in the man page :manpage:`gnt-cluster(8)`; for information about the
763 setup script protocol, look at the Ganeti-supplied script.
768 The `SPICE <http://www.linux-kvm.org/page/SPICE>`_ support has been
771 It is now possible to use TLS-protected connections, and when renewing
772 or changing the cluster certificates (via ``gnt-cluster renew-crypto``,
773 it is now possible to specify spice or spice CA certificates. Also, it
774 is possible to configure a password for SPICE sessions via the
775 hypervisor parameter ``spice_password_file``.
777 There are also new parameters to control the compression and streaming
778 options (e.g. ``spice_image_compression``, ``spice_streaming_video``,
779 etc.). For details, see the man page :manpage:`gnt-instance(8)` and look
780 for the spice parameters.
782 Lastly, it is now possible to see the SPICE connection information via
783 ``gnt-instance console``.
788 A new tool (``tools/ovfconverter``) has been added that supports
789 conversion between Ganeti and the `Open Virtualization Format
790 <http://en.wikipedia.org/wiki/Open_Virtualization_Format>`_ (both to and
793 This relies on the ``qemu-img`` tool to convert the disk formats, so the
794 actual compatibility with other virtualization solutions depends on it.
799 The configuration query daemon (``ganeti-confd``) is now optional, and
800 has been rewritten in Haskell; whether to use the daemon at all, use the
801 Python (default) or the Haskell version is selectable at configure time
802 via the ``--enable-confd`` parameter, which can take one of the
803 ``haskell``, ``python`` or ``no`` values. If not used, disabling the
804 daemon will result in a smaller footprint; for larger systems, we
805 welcome feedback on the Haskell version which might become the default
808 If you want to use ``gnt-node list-drbd`` you need to have the Haskell
809 daemon running. The Python version doesn't implement the new call.
812 User interface changes
813 ~~~~~~~~~~~~~~~~~~~~~~
815 We have replaced the ``--disks`` option of ``gnt-instance
816 replace-disks`` with a more flexible ``--disk`` option, which allows
817 adding and removing disks at arbitrary indices (Issue 188). Furthermore,
818 disk size and mode can be changed upon recreation (via ``gnt-instance
819 recreate-disks``, which accepts the same ``--disk`` option).
821 As many people are used to a ``show`` command, we have added that as an
822 alias to ``info`` on all ``gnt-*`` commands.
824 The ``gnt-instance grow-disk`` command has a new mode in which it can
825 accept the target size of the disk, instead of the delta; this can be
826 more safe since two runs in absolute mode will be idempotent, and
827 sometimes it's also easier to specify the desired size directly.
829 Also the handling of instances with regard to offline secondaries has
830 been improved. Instance operations should not fail because one of it's
831 secondary nodes is offline, even though it's safe to proceed.
833 A new command ``list-drbd`` has been added to the ``gnt-node`` script to
834 support debugging of DRBD issues on nodes. It provides a mapping of DRBD
835 minors to instance name.
840 RAPI coverage has improved, with (for example) new resources for
841 recreate-disks, node power-cycle, etc.
846 There is partial support for ``xl`` in the Xen hypervisor; feedback is
849 Python 2.7 is better supported, and after Ganeti 2.6 we will investigate
850 whether to still support Python 2.4 or move to Python 2.6 as minimum
853 Support for Fedora has been slightly improved; the provided example
854 init.d script should work better on it and the INSTALL file should
855 document the needed dependencies.
860 The deprecated ``QueryLocks`` LUXI request has been removed. Use
861 ``Query(what=QR_LOCK, ...)`` instead.
863 The LUXI requests :pyeval:`luxi.REQ_QUERY_JOBS`,
864 :pyeval:`luxi.REQ_QUERY_INSTANCES`, :pyeval:`luxi.REQ_QUERY_NODES`,
865 :pyeval:`luxi.REQ_QUERY_GROUPS`, :pyeval:`luxi.REQ_QUERY_EXPORTS` and
866 :pyeval:`luxi.REQ_QUERY_TAGS` are deprecated and will be removed in a
867 future version. :pyeval:`luxi.REQ_QUERY` should be used instead.
869 RAPI client: ``CertificateError`` now derives from
870 ``GanetiApiError``. This should make it more easy to handle Ganeti
873 Deprecation warnings due to PyCrypto/paramiko import in
874 ``tools/setup-ssh`` have been silenced, as usually they are safe; please
875 make sure to run an up-to-date paramiko version, if you use this tool.
877 The QA scripts now depend on Python 2.5 or above (the main code base
878 still works with Python 2.4).
880 The configuration file (``config.data``) is now written without
881 indentation for performance reasons; if you want to edit it, it can be
882 re-formatted via ``tools/fmtjson``.
884 A number of bugs has been fixed in the cluster merge tool.
886 ``x509`` certification verification (used in import-export) has been
887 changed to allow the same clock skew as permitted by the cluster
888 verification. This will remove some rare but hard to diagnose errors in
895 *(Released Thu, 19 Jul 2012)*
897 Very few changes from rc4 to the final release, only bugfixes:
899 - integrated fixes from release 2.5.2 (fix general boot flag for KVM
900 instance, fix CDROM booting for KVM instances)
901 - fixed node group modification of node parameters
902 - fixed issue in LUClusterVerifyGroup with multi-group clusters
903 - fixed generation of bash completion to ensure a stable ordering
910 *(Released Fri, 13 Jul 2012)*
912 Third release candidate for 2.6. The following changes were done from
915 - Fixed ``UpgradeConfig`` w.r.t. to disk parameters on disk objects.
916 - Fixed an inconsistency in the LUXI protocol with the provided
917 arguments (NOT backwards compatible)
918 - Fixed a bug with node groups ipolicy where ``min`` was greater than
919 the cluster ``std`` value
920 - Implemented a new ``gnt-node list-drbd`` call to list DRBD minors for
921 easier instance debugging on nodes (requires ``hconfd`` to work)
927 *(Released Tue, 03 Jul 2012)*
929 Second release candidate for 2.6. The following changes were done from
932 - Fixed ``gnt-cluster verify`` regarding ``master-ip-script`` on non
934 - Fixed a RAPI regression on missing beparams/memory
935 - Fixed redistribution of files on offline nodes
936 - Added possibility to run activate-disks even though secondaries are
937 offline. With this change it relaxes also the strictness on some other
938 commands which use activate disks internally:
939 * ``gnt-instance start|reboot|rename|backup|export``
940 - Made it possible to remove safely an instance if its secondaries are
942 - Made it possible to reinstall even though secondaries are offline
948 *(Released Mon, 25 Jun 2012)*
950 First release candidate for 2.6. The following changes were done from
953 - Fixed bugs with disk parameters and ``rbd`` templates as well as
955 - Made ``gnt-instance modify`` more consistent regarding new NIC/Disk
956 behaviour. It supports now the modify operation
957 - ``hcheck`` implemented to analyze cluster health and possibility of
958 improving health by rebalance
959 - ``hbal`` has been improved in dealing with split instances
965 *(Released Mon, 11 Jun 2012)*
967 Second beta release of 2.6. The following changes were done from beta2
970 - Fixed ``daemon-util`` with non-root user models
971 - Fixed creation of plain instances with ``--no-wait-for-sync``
972 - Fix wrong iv_names when running ``cfgupgrade``
973 - Export more information in RAPI group queries
974 - Fixed bug when changing instance network interfaces
975 - Extended burnin to do NIC changes
976 - query: Added ``<``, ``>``, ``<=``, ``>=`` comparison operators
977 - Changed default for DRBD barriers
978 - Fixed DRBD error reporting for syncer rate
979 - Verify the options on disk parameters
981 And of course various fixes to documentation and improved unittests and
988 *(Released Wed, 23 May 2012)*
990 First beta release of 2.6. The following changes were done from beta1 to
993 - integrated patch for distributions without ``start-stop-daemon``
994 - adapted example init.d script to work on Fedora
995 - fixed log handling in Haskell daemons
996 - adapted checks in the watcher for pycurl linked against libnss
997 - add partial support for ``xl`` instead of ``xm`` for Xen
998 - fixed a type issue in cluster verification
999 - fixed ssconf handling in the Haskell code (was breaking confd in IPv6
1002 Plus integrated fixes from the 2.5 branch:
1004 - fixed ``kvm-ifup`` to use ``/bin/bash``
1005 - fixed parallel build failures
1006 - KVM live migration when using a custom keymap
1012 *(Released Tue, 24 Jul 2012)*
1014 A small bugfix release, with no new features:
1016 - fixed bash-isms in kvm-ifup, for compatibility with systems which use a
1017 different default shell (e.g. Debian, Ubuntu)
1018 - fixed KVM startup and live migration with a custom keymap (fixes Issue
1019 243 and Debian bug #650664)
1020 - fixed compatibility with KVM versions that don't support multiple boot
1021 devices (fixes Issue 230 and Debian bug #624256)
1023 Additionally, a few fixes were done to the build system (fixed parallel
1024 build failures) and to the unittests (fixed race condition in test for
1025 FileID functions, and the default enable/disable mode for QA test is now
1032 *(Released Fri, 11 May 2012)*
1034 A small bugfix release.
1036 The main issues solved are on the topic of compatibility with newer LVM
1039 - fixed parsing of ``lv_attr`` field
1040 - adapted to new ``vgreduce --removemissing`` behaviour where sometimes
1041 the ``--force`` flag is needed
1043 Also on the topic of compatibility, ``tools/lvmstrap`` has been changed
1044 to accept kernel 3.x too (was hardcoded to 2.6.*).
1046 A regression present in 2.5.0 that broke handling (in the gnt-* scripts)
1047 of hook results and that also made display of other errors suboptimal
1048 was fixed; the code behaves now like 2.4 and earlier.
1050 Another change in 2.5, the cleanup of the OS scripts environment, is too
1051 aggressive: it removed even the ``PATH`` variable, which requires the OS
1052 scripts to *always* need to export it. Since this is a bit too strict,
1053 we now export a minimal PATH, the same that we export for hooks.
1055 The fix for issue 201 (Preserve bridge MTU in KVM ifup script) was
1056 integrated into this release.
1058 Finally, a few other miscellaneous changes were done (no new features,
1059 just small improvements):
1061 - Fix ``gnt-group --help`` display
1062 - Fix hardcoded Xen kernel path
1063 - Fix grow-disk handling of invalid units
1064 - Update synopsis for ``gnt-cluster repair-disk-sizes``
1065 - Accept both PUT and POST in noded (makes future upgrade to 2.6 easier)
1071 *(Released Thu, 12 Apr 2012)*
1073 Incompatible/important changes and bugfixes
1074 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1076 - The default of the ``/2/instances/[instance_name]/rename`` RAPI
1077 resource's ``ip_check`` parameter changed from ``True`` to ``False``
1078 to match the underlying LUXI interface.
1079 - The ``/2/nodes/[node_name]/evacuate`` RAPI resource was changed to use
1080 body parameters, see :doc:`RAPI documentation <rapi>`. The server does
1081 not maintain backwards-compatibility as the underlying operation
1082 changed in an incompatible way. The RAPI client can talk to old
1083 servers, but it needs to be told so as the return value changed.
1084 - When creating file-based instances via RAPI, the ``file_driver``
1085 parameter no longer defaults to ``loop`` and must be specified.
1086 - The deprecated ``bridge`` NIC parameter is no longer supported. Use
1088 - Support for the undocumented and deprecated RAPI instance creation
1089 request format version 0 has been dropped. Use version 1, supported
1090 since Ganeti 2.1.3 and :doc:`documented <rapi>`, instead.
1091 - Pyparsing 1.4.6 or above is required, see :doc:`installation
1092 documentation <install>`.
1093 - The "cluster-verify" hooks are now executed per group by the
1094 ``OP_CLUSTER_VERIFY_GROUP`` opcode. This maintains the same behavior
1095 if you just run ``gnt-cluster verify``, which generates one opcode per
1097 - The environment as passed to the OS scripts is cleared, and thus no
1098 environment variables defined in the node daemon's environment will be
1099 inherited by the scripts.
1100 - The :doc:`iallocator <iallocator>` mode ``multi-evacuate`` has been
1102 - :doc:`New iallocator modes <design-multi-reloc>` have been added to
1103 support operations involving multiple node groups.
1104 - Offline nodes are ignored when failing over an instance.
1105 - Support for KVM version 1.0, which changed the version reporting format
1107 - TCP/IP ports used by DRBD disks are returned to a pool upon instance
1109 - ``Makefile`` is now compatible with Automake 1.11.2
1110 - Includes all bugfixes made in the 2.4 series
1115 - The ganeti-htools project has been merged into the ganeti-core source
1116 tree and will be built as part of Ganeti (see :doc:`install-quick`).
1117 - Implemented support for :doc:`shared storage <design-shared-storage>`.
1118 - Add support for disks larger than 2 TB in ``lvmstrap`` by supporting
1119 GPT-style partition tables (requires `parted
1120 <http://www.gnu.org/s/parted/>`_).
1121 - Added support for floppy drive and 2nd CD-ROM drive in KVM hypervisor.
1122 - Allowed adding tags on instance creation.
1123 - Export instance tags to hooks (``INSTANCE_TAGS``, see :doc:`hooks`)
1124 - Allow instances to be started in a paused state, enabling the user to
1125 see the complete console output on boot using the console.
1126 - Added new hypervisor flag to control default reboot behaviour
1127 (``reboot_behavior``).
1128 - Added support for KVM keymaps (hypervisor parameter ``keymap``).
1129 - Improved out-of-band management support:
1131 - Added ``gnt-node health`` command reporting the health status of
1133 - Added ``gnt-node power`` command to manage power status of nodes.
1134 - Added command for emergency power-off (EPO), ``gnt-cluster epo``.
1136 - Instance migration can fall back to failover if instance is not
1138 - Filters can be used when listing nodes, instances, groups and locks;
1139 see :manpage:`ganeti(7)` manpage.
1140 - Added post-execution status as variables to :doc:`hooks <hooks>`
1142 - Instance tags are exported/imported together with the instance.
1143 - When given an explicit job ID, ``gnt-job info`` will work for archived
1145 - Jobs can define dependencies on other jobs (not yet supported via
1146 RAPI or command line, but used by internal commands and usable via
1149 - Lock monitor (``gnt-debug locks``) shows jobs waiting for
1152 - Instance failover is now available as a RAPI resource
1153 (``/2/instances/[instance_name]/failover``).
1154 - ``gnt-instance info`` defaults to static information if primary node
1156 - Opcodes have a new ``comment`` attribute.
1157 - Added basic SPICE support to KVM hypervisor.
1158 - ``tools/ganeti-listrunner`` allows passing of arguments to executable.
1160 Node group improvements
1161 ~~~~~~~~~~~~~~~~~~~~~~~
1163 - ``gnt-cluster verify`` has been modified to check groups separately,
1164 thereby improving performance.
1165 - Node group support has been added to ``gnt-cluster verify-disks``,
1166 which now operates per node group.
1167 - Watcher has been changed to work better with node groups.
1169 - One process and state file per node group.
1170 - Slow watcher in one group doesn't block other group's watcher.
1172 - Added new command, ``gnt-group evacuate``, to move all instances in a
1173 node group to other groups.
1174 - Added ``gnt-instance change-group`` to move an instance to another
1176 - ``gnt-cluster command`` and ``gnt-cluster copyfile`` now support
1177 per-group operations.
1178 - Node groups can be tagged.
1179 - Some operations switch from an exclusive to a shared lock as soon as
1181 - Instance's primary and secondary nodes' groups are now available as
1182 query fields (``pnode.group``, ``pnode.group.uuid``, ``snodes.group``
1183 and ``snodes.group.uuid``).
1188 - Numerous updates to documentation and manpages.
1190 - :doc:`RAPI <rapi>` documentation now has detailed parameter
1192 - Some opcode/job results are now also documented, see :doc:`RAPI
1195 - A lockset's internal lock is now also visible in lock monitor.
1196 - Log messages from job queue workers now contain information about the
1197 opcode they're processing.
1198 - ``gnt-instance console`` no longer requires the instance lock.
1199 - A short delay when waiting for job changes reduces the number of LUXI
1200 requests significantly.
1201 - DRBD metadata volumes are overwritten with zeros during disk creation.
1202 - Out-of-band commands no longer acquire the cluster lock in exclusive
1204 - ``devel/upload`` now uses correct permissions for directories.
1210 *(Released Fri, 23 Mar 2012)*
1212 This was the sixth release candidate of the 2.5 series.
1218 *(Released Mon, 9 Jan 2012)*
1220 This was the fifth release candidate of the 2.5 series.
1226 *(Released Thu, 27 Oct 2011)*
1228 This was the fourth release candidate of the 2.5 series.
1234 *(Released Wed, 26 Oct 2011)*
1236 This was the third release candidate of the 2.5 series.
1242 *(Released Tue, 18 Oct 2011)*
1244 This was the second release candidate of the 2.5 series.
1250 *(Released Tue, 4 Oct 2011)*
1252 This was the first release candidate of the 2.5 series.
1258 *(Released Wed, 31 Aug 2011)*
1260 This was the third beta release of the 2.5 series.
1266 *(Released Mon, 22 Aug 2011)*
1268 This was the second beta release of the 2.5 series.
1274 *(Released Fri, 12 Aug 2011)*
1276 This was the first beta release of the 2.5 series.
1282 *(Released Thu, 27 Oct 2011)*
1284 - Fixed bug when parsing command line parameter values ending in
1286 - Fixed assertion error after unclean master shutdown
1287 - Disable HTTP client pool for RPC, significantly reducing memory usage
1289 - Fixed queue archive creation with wrong permissions
1295 *(Released Tue, 23 Aug 2011)*
1299 - Fixed documentation for importing with ``--src-dir`` option
1300 - Fixed a bug in ``ensure-dirs`` with queue/archive permissions
1301 - Fixed a parsing issue with DRBD 8.3.11 in the Linux kernel
1307 *(Released Fri, 5 Aug 2011)*
1309 Many bug-fixes and a few small features:
1311 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
1312 issues when you tried to add an instance with two MAC addresses in one
1314 - KVM: fixed per-instance stored UID value
1315 - KVM: configure bridged NICs at migration start
1316 - KVM: Fix a bug where instance will not start with never KVM versions
1318 - Added OS search path to ``gnt-cluster info``
1319 - Fixed an issue with ``file_storage_dir`` where you were forced to
1320 provide an absolute path, but the documentation states it is a
1321 relative path, the documentation was right
1322 - Added a new parameter to instance stop/start called ``--no-remember``
1323 that will make the state change to not be remembered
1324 - Implemented ``no_remember`` at RAPI level
1325 - Improved the documentation
1326 - Node evacuation: don't call IAllocator if node is already empty
1327 - Fixed bug in DRBD8 replace disks on current nodes
1328 - Fixed bug in recreate-disks for DRBD instances
1329 - Moved assertion checking locks in ``gnt-instance replace-disks``
1330 causing it to abort with not owning the right locks for some situation
1331 - Job queue: Fixed potential race condition when cancelling queued jobs
1332 - Fixed off-by-one bug in job serial generation
1333 - ``gnt-node volumes``: Fix instance names
1334 - Fixed aliases in bash completion
1335 - Fixed a bug in reopening log files after being sent a SIGHUP
1336 - Added a flag to burnin to allow specifying VCPU count
1337 - Bugfixes to non-root Ganeti configuration
1343 *(Released Thu, 12 May 2011)*
1345 Many bug-fixes and a few new small features:
1347 - Fixed a bug related to log opening failures
1348 - Fixed a bug in instance listing with orphan instances
1349 - Fixed a bug which prevented resetting the cluster-level node parameter
1350 ``oob_program`` to the default
1351 - Many fixes related to the ``cluster-merge`` tool
1352 - Fixed a race condition in the lock monitor, which caused failures
1353 during (at least) creation of many instances in parallel
1354 - Improved output for gnt-job info
1355 - Removed the quiet flag on some ssh calls which prevented debugging
1357 - Improved the N+1 failure messages in cluster verify by actually
1358 showing the memory values (needed and available)
1359 - Increased lock attempt timeouts so that when executing long operations
1360 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
1361 too early and thus prevent the use of the 'fair' mechanism
1362 - Changed instance query data (``gnt-instance info``) to not acquire
1363 locks unless needed, thus allowing its use on locked instance if only
1364 static information is asked for
1365 - Improved behaviour with filesystems that do not support rename on an
1367 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
1368 kept locks on all nodes during the wipe, which is unneeded
1369 - Fixed ``gnt-watcher`` handling of errors during hooks execution
1370 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
1371 10GiB) which caused the wipe to fail right at the end in some cases
1372 - Fixed master IP activation when doing master failover with no-voting
1373 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
1374 the master node itself
1375 - Fixed potential data-loss in under disk full conditions, where Ganeti
1376 wouldn't check correctly the return code and would consider
1377 partially-written files 'correct'
1378 - Fixed bug related to multiple VGs and DRBD disk replacing
1379 - Added new disk parameter ``metavg`` that allows placement of the meta
1380 device for DRBD in a different volume group
1381 - Fixed error handling in the node daemon when the system libc doesn't
1382 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
1383 - Fixed lock release during replace-disks, which kept cluster-wide locks
1384 when doing disk replaces with an iallocator script
1385 - Added check for missing bridges in cluster verify
1386 - Handle EPIPE errors while writing to the terminal better, so that
1387 piping the output to e.g. ``less`` doesn't cause a backtrace
1388 - Fixed rare case where a ^C during Luxi calls could have been
1389 interpreted as server errors, instead of simply terminating
1390 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
1392 - Added a few more parameters to the KVM hypervisor, allowing a second
1393 CDROM, custom disk type for CDROMs and a floppy image
1394 - Removed redundant message in instance rename when the name is given
1396 - Added option to ``gnt-instance recreate-disks`` to allow creating the
1397 disks on new nodes, allowing recreation when the original instance
1398 nodes are completely gone
1399 - Added option when converting disk templates to DRBD to skip waiting
1400 for the resync, in order to make the instance available sooner
1401 - Added two new variables to the OS scripts environment (containing the
1403 - Made the root_path and optional parameter for the xen-pvm hypervisor,
1404 to allow use of ``pvgrub`` as bootloader
1405 - Changed the instance memory modifications to only check out-of-memory
1406 conditions on memory increases, and turned the secondary node warnings
1407 into errors (they can still be overridden via ``--force``)
1408 - Fixed the handling of a corner case when the Python installation gets
1409 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
1410 to execute a command that doesn't exist
1411 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
1412 node of the instance returned failures during instance shutdown; this
1413 adds the option ``--ignore-consistency`` to gnt-instance move
1415 And as usual, various improvements to the error messages, documentation
1422 *(Released Wed, 09 Mar 2011)*
1424 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
1425 the RAPI users file if run twice (even with ``--dry-run``).
1427 The release fixes that bug (nothing else changed).
1433 *(Released Mon, 07 Mar 2011)*
1435 Final 2.4.0 release. Just a few small fixes:
1437 - Fixed RAPI node evacuate
1438 - Fixed the kvm-ifup script
1439 - Fixed internal error handling for special job cases
1440 - Updated man page to specify the escaping feature for options
1446 *(Released Mon, 28 Feb 2011)*
1448 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
1449 that this feature wiped the disks of the wrong instance, leading to loss
1454 - Fixed title of query field containing instance name
1455 - Expanded the glossary in the documentation
1456 - Fixed one unittest (internal issue)
1462 *(Released Mon, 21 Feb 2011)*
1464 A number of bug fixes plus just a couple functionality changes.
1466 On the user-visible side, the ``gnt-* list`` command output has changed
1467 with respect to "special" field states. The current rc1 style of display
1468 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
1469 the default output mode special fields states are displayed as follows:
1471 - Offline resource: ``*``
1472 - Unavailable/not applicable: ``-``
1473 - Data missing (RPC failure): ``?``
1474 - Unknown field: ``??``
1476 Another user-visible change is the addition of ``--force-join`` to
1481 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
1482 - Fixed regression in RAPI/instance reinstall where all parameters were
1483 required (instead of optional)
1484 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
1485 - Fixed iallocator usage (offline nodes were not considered offline)
1486 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
1487 - Fixed hypervisor and OS parameter validation with respect to
1488 non-vm_capable nodes
1489 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
1491 - Fixed ``tools/listrunner`` with respect to agent-based usage
1497 *(Released Fri, 4 Feb 2011)*
1499 Many changes and fixes since the beta1 release. While there were some
1500 internal changes, the code has been mostly stabilised for the RC
1503 Note: the dumb allocator was removed in this release, as it was not kept
1504 up-to-date with the IAllocator protocol changes. It is recommended to
1505 use the ``hail`` command from the ganeti-htools package.
1507 Note: the 2.4 and up versions of Ganeti are not compatible with the
1508 0.2.x branch of ganeti-htools. You need to upgrade to
1509 ganeti-htools-0.3.0 (or later).
1511 Regressions fixed from 2.3
1512 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1514 - Fixed the ``gnt-cluster verify-disks`` command
1515 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
1517 - Fixed disk adoption breakage
1518 - Fixed wrong headers in instance listing for field aliases
1523 - Fixed corner case in KVM handling of NICs
1524 - Fixed many cases of wrong handling of non-vm_capable nodes
1525 - Fixed a bug where a missing instance symlink was not possible to
1526 recreate with any ``gnt-*`` command (now ``gnt-instance
1527 activate-disks`` does it)
1528 - Fixed the volume group name as reported by ``gnt-cluster
1530 - Increased timeouts for the import-export code, hopefully leading to
1531 fewer aborts due network or instance timeouts
1532 - Fixed bug in ``gnt-node list-storage``
1533 - Fixed bug where not all daemons were started on cluster
1534 initialisation, but only at the first watcher run
1535 - Fixed many bugs in the OOB implementation
1536 - Fixed watcher behaviour in presence of instances with offline
1538 - Fixed instance list output for instances running on the wrong node
1539 - a few fixes to the cluster-merge tool, but it still cannot merge
1540 multi-node groups (currently it is not recommended to use this tool)
1546 - Improved network configuration for the KVM hypervisor
1547 - Added e1000 as a supported NIC for Xen-HVM
1548 - Improved the lvmstrap tool to also be able to use partitions, as
1549 opposed to full disks
1550 - Improved speed of disk wiping (the cluster parameter
1551 ``prealloc_wipe_disks``, so that it has a low impact on the total time
1552 of instance creations
1553 - Added documentation for the OS parameters
1554 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
1555 hypervisor is not responding
1556 - Added display of blacklisted and hidden OS information in
1557 ``gnt-cluster info``
1558 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
1559 NIC and node parameters, which might create problems with currently
1560 invalid (but undetected) configuration files, but prevents validation
1561 failures when unrelated parameters are modified
1562 - Changed cluster initialisation to wait for the master daemon to become
1564 - Expanded the RAPI interface:
1566 - Added config redistribution resource
1567 - Added activation/deactivation of instance disks
1568 - Added export of console information
1570 - Implemented log file reopening on SIGHUP, which allows using
1571 logrotate(8) for the Ganeti log files
1572 - Added a basic OOB helper script as an example
1578 *(Released Fri, 14 Jan 2011)*
1583 - Fixed timezone issues when formatting timestamps
1584 - Added support for node groups, available via ``gnt-group`` and other
1586 - Added out-of-band framework and management, see :doc:`design
1587 document <design-oob>`
1588 - Removed support for roman numbers from ``gnt-node list`` and
1589 ``gnt-instance list``.
1590 - Allowed modification of master network interface via ``gnt-cluster
1591 modify --master-netdev``
1592 - Accept offline secondaries while shutting down instance disks
1593 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
1594 - Added support for multiple LVM volume groups
1595 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
1597 - Added commands to list available fields:
1599 - ``gnt-node list-fields``
1600 - ``gnt-group list-fields``
1601 - ``gnt-instance list-fields``
1603 - Updated documentation and man pages
1608 - Moved ``rapi_users`` file into separate directory, now named
1609 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
1611 - Added new tool for running commands on many machines,
1612 ``tools/ganeti-listrunner``
1613 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
1614 improving the ``gnt-instance console`` output
1615 - Allowed customisation of disk index separator at ``configure`` time
1616 - Export node group allocation policy to :doc:`iallocator <iallocator>`
1617 - Added support for non-partitioned md disks in ``lvmstrap``
1618 - Added script to gracefully power off KVM instances
1619 - Split ``utils`` module into smaller parts
1620 - Changed query operations to return more detailed information, e.g.
1621 whether an information is unavailable due to an offline node. To use
1622 this new functionality, the LUXI call ``Query`` must be used. Field
1623 information is now stored by the master daemon and can be retrieved
1624 using ``QueryFields``. Instances, nodes and groups can also be queried
1625 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
1626 exposed via RAPI). The following commands make use of this
1627 infrastructure change:
1629 - ``gnt-group list``
1630 - ``gnt-group list-fields``
1632 - ``gnt-node list-fields``
1633 - ``gnt-instance list``
1634 - ``gnt-instance list-fields``
1635 - ``gnt-debug locks``
1640 - New RAPI resources (see :doc:`rapi`):
1644 - ``/2/groups/[group_name]``
1645 - ``/2/groups/[group_name]/assign-nodes``
1646 - ``/2/groups/[group_name]/modify``
1647 - ``/2/groups/[group_name]/rename``
1648 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
1652 - Implemented ``no_install`` for instance creation
1653 - Implemented OS parameters for instance reinstallation, allowing
1654 use of special settings on reinstallation (e.g. for preserving data)
1659 - Added IPv6 support in import/export
1660 - Pause DRBD synchronization while wiping disks on instance creation
1661 - Updated unittests and QA scripts
1662 - Improved network parameters passed to KVM
1663 - Converted man pages from docbook to reStructuredText
1669 *(Released Mon, 20 Dec 2010)*
1671 Released version 2.3.1~rc1 without any changes.
1677 *(Released Wed, 1 Dec 2010)*
1679 - impexpd: Disable OpenSSL compression in socat if possible (backport
1680 from master, commit e90739d625b, see :doc:`installation guide
1681 <install-quick>` for details)
1682 - Changed unittest coverage report to exclude test scripts
1683 - Added script to check version format
1689 *(Released Wed, 1 Dec 2010)*
1691 Released version 2.3.0~rc1 without any changes.
1697 *(Released Fri, 19 Nov 2010)*
1699 A number of bugfixes and documentation updates:
1701 - Update ganeti-os-interface documentation
1702 - Fixed a bug related to duplicate MACs or similar items which should be
1704 - Fix breakage in OS state modify
1705 - Reinstall instance: disallow offline secondaries (fixes bug related to
1706 OS changing but reinstall failing)
1707 - plus all the other fixes between 2.2.1 and 2.2.2
1713 *(Released Tue, 2 Nov 2010)*
1715 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
1716 - Fixed master failover race with watcher
1717 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
1719 - Accept previously stopped instance for export with instance removal
1720 - Simplify and extend the environment variables for instance OS scripts
1721 - Added new node flags, ``master_capable`` and ``vm_capable``
1722 - Added optional instance disk wiping prior during allocation. This is a
1723 cluster-wide option and can be set/modified using
1724 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
1725 - Added IPv6 support, see :doc:`design document <design-2.3>` and
1726 :doc:`install-quick`
1727 - Added a new watcher option (``--ignore-pause``)
1728 - Added option to ignore offline node on instance start/stop
1729 (``--ignore-offline``)
1730 - Allow overriding OS parameters with ``gnt-instance reinstall``
1731 - Added ability to change node's secondary IP address using ``gnt-node
1733 - Implemented privilege separation for all daemons except
1734 ``ganeti-noded``, see ``configure`` options
1735 - Complain if an instance's disk is marked faulty in ``gnt-cluster
1737 - Implemented job priorities (see ``ganeti(7)`` manpage)
1738 - Ignore failures while shutting down instances during failover from
1740 - Exit daemon's bootstrap process only once daemon is ready
1741 - Export more information via ``LUInstanceQuery``/remote API
1742 - Improved documentation, QA and unittests
1743 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
1744 restart if the file was created or changed
1745 - Added LUXI protocol version sent with each request and response,
1746 allowing detection of server/client mismatches
1747 - Moved the Python scripts among gnt-* and ganeti-* into modules
1748 - Moved all code related to setting up SSH to an external script,
1750 - Infrastructure changes for node group support in future versions
1756 *(Released Fri, 19 Nov 2010)*
1758 A few small bugs fixed, and some improvements to the build system:
1760 - Fix documentation regarding conversion to drbd
1761 - Fix validation of parameters in cluster modify (``gnt-cluster modify
1763 - Fix error handling in node modify with multiple changes
1764 - Allow remote imports without checked names
1770 *(Released Tue, 19 Oct 2010)*
1772 - Disable SSL session ID cache in RPC client
1778 *(Released Thu, 14 Oct 2010)*
1780 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
1781 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
1783 - Fix problems with interaction between Curl and Python's HTTP server,
1784 resulting in increased speed in many RPC calls
1785 - Improve our release script to prevent breakage with older aclocal and
1792 *(Released Thu, 7 Oct 2010)*
1794 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
1795 value retrieved from master
1796 - Added support for blacklisted or hidden OS definitions
1797 - Added simple lock monitor (accessible via (``gnt-debug locks``)
1798 - Added support for -mem-path in KVM hypervisor abstraction layer
1799 - Allow overriding instance parameters in tool for inter-cluster
1800 instance moves (``tools/move-instance``)
1801 - Improved opcode summaries (e.g. in ``gnt-job list``)
1802 - Improve consistency of OS listing by sorting it
1803 - Documentation updates
1809 *(Released Fri, 8 Oct 2010)*
1811 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
1817 *(Released Mon, 4 Oct 2010)*
1819 - Fixed regression in ``gnt-instance rename``
1825 *(Released Wed, 22 Sep 2010)*
1827 - Fixed OS_VARIANT variable for OS scripts
1828 - Fixed cluster tag operations via RAPI
1829 - Made ``setup-ssh`` exit with non-zero code if an error occurred
1830 - Disabled RAPI CA checks in watcher
1836 *(Released Mon, 23 Aug 2010)*
1838 - Support DRBD versions of the format "a.b.c.d"
1840 - Re-introduce support for usage from multiple threads in RAPI client
1841 - Instance renames and modify via RAPI
1842 - Work around race condition between processing and archival in job
1844 - Mark opcodes following failed one as failed, too
1845 - Job field ``lock_status`` was removed due to difficulties making it
1846 work with the changed job queue in Ganeti 2.2; a better way to monitor
1847 locks is expected for a later 2.2.x release
1848 - Fixed dry-run behaviour with many commands
1849 - Support ``ssh-agent`` again when adding nodes
1850 - Many additional bugfixes
1856 *(Released Fri, 30 Jul 2010)*
1858 Important change: the internal RPC mechanism between Ganeti nodes has
1859 changed from using a home-grown http library (based on the Python base
1860 libraries) to use the PycURL library. This requires that PycURL is
1861 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
1862 against GnuTLS by default. cURL's support for GnuTLS had known issues
1863 before cURL 7.21.0 and we recommend using the latest cURL release or
1864 linking against OpenSSL. Most other distributions already link PycURL
1865 and cURL against OpenSSL. The command::
1867 python -c 'import pycurl; print pycurl.version'
1869 can be used to determine the libraries PycURL and cURL are linked
1872 Other significant changes:
1874 - Rewrote much of the internals of the job queue, in order to achieve
1875 better parallelism; this decouples job query operations from the job
1876 processing, and it should allow much nicer behaviour of the master
1877 daemon under load, and it also has uncovered some long-standing bugs
1878 related to the job serialisation (now fixed)
1879 - Added a default iallocator setting to the cluster parameters,
1880 eliminating the need to always pass nodes or an iallocator for
1881 operations that require selection of new node(s)
1882 - Added experimental support for the LXC virtualization method
1883 - Added support for OS parameters, which allows the installation of
1884 instances to pass parameter to OS scripts in order to customise the
1886 - Added a hypervisor parameter controlling the migration type (live or
1887 non-live), since hypervisors have various levels of reliability; this
1888 has renamed the 'live' parameter to 'mode'
1889 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
1890 logical volumes, meaning that cluster verify will ignore them and not
1891 flag their presence as errors
1892 - The watcher will now reset the error count for failed instances after
1893 8 hours, thus allowing self-healing if the problem that caused the
1894 instances to be down/fail to start has cleared in the meantime
1895 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
1896 check for, and warn, if the drbd module parameter ``usermode_helper``
1897 is not consistent with the cluster-wide setting; this is needed to
1898 make diagnose easier of failed drbd creations
1899 - Started adding base IPv6 support, but this is not yet
1900 enabled/available for use
1901 - Rename operations (cluster, instance) will now return the new name,
1902 which is especially useful if a short name was passed in
1903 - Added support for instance migration in RAPI
1904 - Added a tool to pre-configure nodes for the SSH setup, before joining
1905 them to the cluster; this will allow in the future a simplified model
1906 for node joining (but not yet fully enabled in 2.2); this needs the
1907 paramiko python library
1908 - Fixed handling of name-resolving errors
1909 - Fixed consistency of job results on the error path
1910 - Fixed master-failover race condition when executed multiple times in
1912 - Fixed many bugs related to the job queue (mostly introduced during the
1913 2.2 development cycle, so not all are impacting 2.1)
1914 - Fixed instance migration with missing disk symlinks
1915 - Fixed handling of unknown jobs in ``gnt-job archive``
1916 - And many other small fixes/improvements
1920 - Enhanced both the unittest and the QA coverage
1921 - Switched the opcode validation to a generic model, and extended the
1922 validation to all opcode parameters
1923 - Changed more parts of the code that write shell scripts to use the
1925 - Switched the master daemon to use the asyncore library for the Luxi
1932 *(Released Thu, 17 Jun 2010)*
1934 - Added tool (``move-instance``) and infrastructure to move instances
1935 between separate clusters (see :doc:`separate documentation
1936 <move-instance>` and :doc:`design document <design-2.2>`)
1937 - Added per-request RPC timeout
1938 - RAPI now requires a Content-Type header for requests with a body (e.g.
1939 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
1940 :rfc:`2616` (HTTP/1.1), section 7.2.1)
1941 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
1943 - Implemented initial support for running Ganeti daemons as separate
1944 users, see configure-time flags ``--with-user-prefix`` and
1945 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
1947 - Instances can be removed after export (``gnt-backup export
1948 --remove-instance``)
1949 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
1950 key (instead of 1024 bit)
1951 - Added new cluster configuration file for cluster domain secret
1952 - Import/export now use SSL instead of SSH
1953 - Added support for showing estimated time when exporting an instance,
1954 see the ``ganeti-os-interface(7)`` manpage and look for
1961 *(Released Tue, 16 Nov 2010)*
1963 Some more bugfixes. Unless critical bugs occur, this will be the last
1966 - Fix case of MAC special-values
1967 - Fix mac checker regex
1968 - backend: Fix typo causing "out of range" error
1969 - Add missing --units in gnt-instance list man page
1975 *(Released Tue, 24 Aug 2010)*
1978 - Don't ignore secondary node silently on non-mirrored disk templates
1980 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
1981 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
1982 - Properly document the usb_mouse parameter
1983 - Fix path in ganeti-rapi(8) (issue 116)
1984 - Adjust error message when the ganeti user's .ssh directory is
1986 - Add same-node-check when changing the disk template to drbd
1992 *(Released Fri, 16 Jul 2010)*
1995 - Add an option to only select some reboot types during qa/burnin.
1996 (on some hypervisors consequent reboots are not supported)
1997 - Fix infrequent race condition in master failover. Sometimes the old
1998 master ip address would be still detected as up for a short time
1999 after it was removed, causing failover to fail.
2000 - Decrease mlockall warnings when the ctypes module is missing. On
2001 Python 2.4 we support running even if no ctypes module is installed,
2002 but we were too verbose about this issue.
2003 - Fix building on old distributions, on which man doesn't have a
2005 - Fix RAPI not to ignore the MAC address on instance creation
2006 - Implement the old instance creation format in the RAPI client.
2012 *(Released Thu, 01 Jul 2010)*
2014 A small bugfix release:
2015 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
2016 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
2017 a non-existing option
2018 - Fix instance create: the --force-variant option was ignored
2019 - Improve pylint 0.21 compatibility and warnings with Python 2.6
2020 - Fix modify node storage with non-FQDN arguments
2021 - Fix RAPI client to authenticate under Python 2.6 when used
2022 for more than 5 requests needing authentication
2023 - Fix gnt-instance modify -t (storage) giving a wrong error message
2024 when converting a non-shutdown drbd instance to plain
2030 *(Released Fri, 18 Jun 2010)*
2032 A small bugfix release:
2034 - Fix live migration of KVM instances started with older Ganeti
2035 versions which had fewer hypervisor parameters
2036 - Fix gnt-instance grow-disk on down instances
2037 - Fix an error-reporting bug during instance migration
2038 - Better checking of the ``--net`` and ``--disk`` values, to avoid
2039 silently ignoring broken ones
2040 - Fix an RPC error reporting bug affecting, for example, RAPI client
2042 - Fix bug triggered by different API version os-es on different nodes
2043 - Fix a bug in instance startup with custom hvparams: OS level
2044 parameters would fail to be applied.
2045 - Fix the RAPI client under Python 2.6 (but more work is needed to
2046 make it work completely well with OpenSSL)
2047 - Fix handling of errors when resolving names from DNS
2053 *(Released Thu, 3 Jun 2010)*
2055 A medium sized development cycle. Some new features, and some
2056 fixes/small improvements/cleanups.
2058 Significant features
2059 ~~~~~~~~~~~~~~~~~~~~
2061 The node deamon now tries to mlock itself into memory, unless the
2062 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
2063 its logs, and falls back to console logging. This allows emergency
2064 features such as ``gnt-node powercycle`` to work even in the event of a
2065 broken node disk (tested offlining the disk hosting the node's
2066 filesystem and dropping its memory caches; don't try this at home)
2068 KVM: add vhost-net acceleration support. It can be tested with a new
2069 enough version of the kernel and of qemu-kvm.
2071 KVM: Add instance chrooting feature. If you use privilege dropping for
2072 your VMs you can also now force them to chroot to an empty directory,
2073 before starting the emulated guest.
2075 KVM: Add maximum migration bandwith and maximum downtime tweaking
2076 support (requires a new-enough version of qemu-kvm).
2078 Cluster verify will now warn if the master node doesn't have the master
2079 ip configured on it.
2081 Add a new (incompatible) instance creation request format to RAPI which
2082 supports all parameters (previously only a subset was supported, and it
2083 wasn't possible to extend the old format to accomodate all the new
2084 features. The old format is still supported, and a client can check for
2085 this feature, before using it, by checking for its presence in the
2086 ``features`` RAPI resource.
2088 Now with ancient latin support. Try it passing the ``--roman`` option to
2089 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
2090 (requires the python-roman module to be installed, in order to work).
2095 As usual many internal code refactorings, documentation updates, and
2098 - Lots of improvements and cleanups to the experimental Remote API
2099 (RAPI) client library.
2100 - A new unit test suite for the core daemon libraries.
2101 - A fix to creating missing directories makes sure the umask is not
2102 applied anymore. This enforces the same directory permissions
2104 - Better handling terminating daemons with ctrl+c (used when running
2105 them in debugging mode).
2106 - Fix a race condition in live migrating a KVM instance, when stat()
2107 on the old proc status file returned EINVAL, which is an unexpected
2109 - Fixed manpage checking with newer man and utf-8 charachters. But now
2110 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
2116 *(Released Fri, 7 May 2010)*
2118 Fix a bug which prevented untagged KVM instances from starting.
2124 *(Released Fri, 7 May 2010)*
2126 Another release with a long development cycle, during which many
2127 different features were added.
2129 Significant features
2130 ~~~~~~~~~~~~~~~~~~~~
2132 The KVM hypervisor now can run the individual instances as non-root, to
2133 reduce the impact of a VM being hijacked due to bugs in the
2134 hypervisor. It is possible to run all instances as a single (non-root)
2135 user, to manually specify a user for each instance, or to dynamically
2136 allocate a user out of a cluster-wide pool to each instance, with the
2137 guarantee that no two instances will run under the same user ID on any
2140 An experimental RAPI client library, that can be used standalone
2141 (without the other Ganeti libraries), is provided in the source tree as
2142 ``lib/rapi/client.py``. Note this client might change its interface in
2143 the future, as we iterate on its capabilities.
2145 A new command, ``gnt-cluster renew-crypto`` has been added to easily
2146 replace the cluster's certificates and crypto keys. This might help in
2147 case they have been compromised, or have simply expired.
2149 A new disk option for instance creation has been added that allows one
2150 to "adopt" currently existing logical volumes, with data
2151 preservation. This should allow easier migration to Ganeti from
2152 unmanaged (or managed via other software) instances.
2154 Another disk improvement is the possibility to convert between redundant
2155 (DRBD) and plain (LVM) disk configuration for an instance. This should
2156 allow better scalability (starting with one node and growing the
2157 cluster, or shrinking a two-node cluster to one node).
2159 A new feature that could help with automated node failovers has been
2160 implemented: if a node sees itself as offline (by querying the master
2161 candidates), it will try to shutdown (hard) all instances and any active
2162 DRBD devices. This reduces the risk of duplicate instances if an
2163 external script automatically failovers the instances on such nodes. To
2164 enable this, the cluster parameter ``maintain_node_health`` should be
2165 enabled; in the future this option (per the name) will enable other
2166 automatic maintenance features.
2168 Instance export/import now will reuse the original instance
2169 specifications for all parameters; that means exporting an instance,
2170 deleting it and the importing it back should give an almost identical
2171 instance. Note that the default import behaviour has changed from
2172 before, where it created only one NIC; now it recreates the original
2175 Cluster verify has added a few new checks: SSL certificates validity,
2176 /etc/hosts consistency across the cluster, etc.
2181 As usual, many internal changes were done, documentation fixes,
2184 - Fixed cluster initialization with disabled cluster storage (regression
2185 introduced in 2.1.1)
2186 - File-based storage supports growing the disks
2187 - Fixed behaviour of node role changes
2188 - Fixed cluster verify for some corner cases, plus a general rewrite of
2189 cluster verify to allow future extension with more checks
2190 - Fixed log spamming by watcher and node daemon (regression introduced
2192 - Fixed possible validation issues when changing the list of enabled
2194 - Fixed cleanup of /etc/hosts during node removal
2195 - Fixed RAPI response for invalid methods
2196 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
2197 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
2198 from ide disks, etc.)
2199 - Allow OS changes without re-installation (to record a changed OS
2200 outside of Ganeti, or to allow OS renames)
2201 - Allow instance creation without OS installation (useful for example if
2202 the OS will be installed manually, or restored from a backup not in
2204 - Implemented option to make cluster ``copyfile`` use the replication
2206 - Added list of enabled hypervisors to ssconf (possibly useful for
2208 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
2210 - A partial form of node re-IP is possible via node readd, which now
2211 allows changed node primary IP
2212 - Command line utilities now show an informational message if the job is
2214 - The logs of the master daemon now show the PID/UID/GID of the
2221 *(Released Fri, 12 Mar 2010)*
2223 During the 2.1.0 long release candidate cycle, a lot of improvements and
2224 changes have accumulated with were released later as 2.1.1.
2229 The node evacuate command (``gnt-node evacuate``) was significantly
2230 rewritten, and as such the IAllocator protocol was changed - a new
2231 request type has been added. This unfortunate change during a stable
2232 series is designed to improve performance of node evacuations; on
2233 clusters with more than about five nodes and which are well-balanced,
2234 evacuation should proceed in parallel for all instances of the node
2235 being evacuated. As such, any existing IAllocator scripts need to be
2236 updated, otherwise the above command will fail due to the unknown
2237 request. The provided "dumb" allocator has not been updated; but the
2238 ganeti-htools package supports the new protocol since version 0.2.4.
2240 Another important change is increased validation of node and instance
2241 names. This might create problems in special cases, if invalid host
2242 names are being used.
2244 Also, a new layer of hypervisor parameters has been added, that sits at
2245 OS level between the cluster defaults and the instance ones. This allows
2246 customisation of virtualization parameters depending on the installed
2247 OS. For example instances with OS 'X' may have a different KVM kernel
2248 (or any other parameter) than the cluster defaults. This is intended to
2249 help managing a multiple OSes on the same cluster, without manual
2250 modification of each instance's parameters.
2252 A tool for merging clusters, ``cluster-merge``, has been added in the
2253 tools sub-directory.
2258 - Improved the int/float conversions that should make the code more
2259 robust in face of errors from the node daemons
2260 - Fixed the remove node code in case of internal configuration errors
2261 - Fixed the node daemon behaviour in face of inconsistent queue
2262 directory (e.g. read-only file-system where we can't open the files
2264 - Fixed the behaviour of gnt-node modify for master candidate demotion;
2265 now it either aborts cleanly or, if given the new "auto_promote"
2266 parameter, will automatically promote other nodes as needed
2267 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
2268 completely prevent Ganeti from working
2269 - Fixed bug for instance export when not all disks were successfully
2271 - Fixed behaviour of node add when the new node is slow in starting up
2273 - Fixed handling of signals in the LUXI client, which should improve
2274 behaviour of command-line scripts
2275 - Added checks for invalid node/instance names in the configuration (now
2276 flagged during cluster verify)
2277 - Fixed watcher behaviour for disk activation errors
2278 - Fixed two potentially endless loops in http library, which led to the
2279 RAPI daemon hanging and consuming 100% CPU in some cases
2280 - Fixed bug in RAPI daemon related to hashed passwords
2281 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
2283 - Enhanced compatibility with non-Debian OSes, but not using absolute
2284 path in some commands and allowing customisation of the ssh
2285 configuration directory
2286 - Fixed possible future issue with new Python versions by abiding to the
2287 proper use of ``__slots__`` attribute on classes
2288 - Added checks that should prevent directory traversal attacks
2289 - Many documentation fixes based on feedback from users
2294 - Added an "early_release" more for instance replace disks and node
2295 evacuate, where we release locks earlier and thus allow higher
2296 parallelism within the cluster
2297 - Added watcher hooks, intended to allow the watcher to restart other
2298 daemons (e.g. from the ganeti-nbma project), but they can be used of
2299 course for any other purpose
2300 - Added a compile-time disable for DRBD barriers, to increase
2301 performance if the administrator trusts the power supply or the
2302 storage system to not lose writes
2303 - Added the option of using syslog for logging instead of, or in
2304 addition to, Ganeti's own log files
2305 - Removed boot restriction for paravirtual NICs for KVM, recent versions
2306 can indeed boot from a paravirtual NIC
2307 - Added a generic debug level for many operations; while this is not
2308 used widely yet, it allows one to pass the debug value all the way to
2310 - Enhanced the hooks environment for instance moves (failovers,
2311 migrations) where the primary/secondary nodes changed during the
2312 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
2313 - Enhanced data validations for many user-supplied values; one important
2314 item is the restrictions imposed on instance and node names, which
2315 might reject some (invalid) host names
2316 - Add a configure-time option to disable file-based storage, if it's not
2317 needed; this allows greater security separation between the master
2318 node and the other nodes from the point of view of the inter-node RPC
2320 - Added user notification in interactive tools if job is waiting in the
2321 job queue or trying to acquire locks
2322 - Added log messages when a job is waiting for locks
2323 - Added filtering by node tags in instance operations which admit
2324 multiple instances (start, stop, reboot, reinstall)
2325 - Added a new tool for cluster mergers, ``cluster-merge``
2326 - Parameters from command line which are of the form ``a=b,c=d`` can now
2327 use backslash escapes to pass in values which contain commas,
2328 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
2330 - For KVM, the instance name is the first parameter passed to KVM, so
2331 that it's more visible in the process list
2337 *(Released Tue, 2 Mar 2010)*
2339 Ganeti 2.1 brings many improvements with it. Major changes:
2341 - Added infrastructure to ease automated disk repairs
2342 - Added new daemon to export configuration data in a cheaper way than
2343 using the remote API
2344 - Instance NICs can now be routed instead of being associated with a
2346 - Improved job locking logic to reduce impact of jobs acquiring multiple
2347 locks waiting for other long-running jobs
2349 In-depth implementation details can be found in the Ganeti 2.1 design
2355 - Added chroot hypervisor
2356 - Added more options to xen-hvm hypervisor (``kernel_path`` and
2358 - Added more options to xen-pvm hypervisor (``use_bootloader``,
2359 ``bootloader_path`` and ``bootloader_args``)
2360 - Added the ``use_localtime`` option for the xen-hvm and kvm
2361 hypervisors, and the default value for this has changed to false (in
2362 2.0 xen-hvm always enabled it)
2363 - Added luxi call to submit multiple jobs in one go
2364 - Added cluster initialization option to not modify ``/etc/hosts``
2366 - Added network interface parameters
2367 - Added dry run mode to some LUs
2368 - Added RAPI resources:
2370 - ``/2/instances/[instance_name]/info``
2371 - ``/2/instances/[instance_name]/replace-disks``
2372 - ``/2/nodes/[node_name]/evacuate``
2373 - ``/2/nodes/[node_name]/migrate``
2374 - ``/2/nodes/[node_name]/role``
2375 - ``/2/nodes/[node_name]/storage``
2376 - ``/2/nodes/[node_name]/storage/modify``
2377 - ``/2/nodes/[node_name]/storage/repair``
2379 - Added OpCodes to evacuate or migrate all instances on a node
2380 - Added new command to list storage elements on nodes (``gnt-node
2381 list-storage``) and modify them (``gnt-node modify-storage``)
2382 - Added new ssconf files with master candidate IP address
2383 (``ssconf_master_candidates_ips``), node primary IP address
2384 (``ssconf_node_primary_ips``) and node secondary IP address
2385 (``ssconf_node_secondary_ips``)
2386 - Added ``ganeti-confd`` and a client library to query the Ganeti
2387 configuration via UDP
2388 - Added ability to run hooks after cluster initialization and before
2390 - Added automatic mode for disk replace (``gnt-instance replace-disks
2392 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
2393 after catastrophic data-loss
2394 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
2396 - Added ``gnt-instance move`` command to move instances
2397 - Added ``gnt-cluster watcher`` command to control watcher
2398 - Added ``gnt-node powercycle`` command to powercycle nodes
2399 - Added new job status field ``lock_status``
2400 - Added parseable error codes to cluster verification (``gnt-cluster
2401 verify --error-codes``) and made output less verbose (use
2402 ``--verbose`` to restore previous behaviour)
2403 - Added UUIDs to the main config entities (cluster, nodes, instances)
2404 - Added support for OS variants
2405 - Added support for hashed passwords in the Ganeti remote API users file
2407 - Added option to specify maximum timeout on instance shutdown
2408 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
2409 - Added new helper script to start and stop Ganeti daemons
2410 (``daemon-util``), with the intent to reduce the work necessary to
2411 adjust Ganeti for non-Debian distributions and to start/stop daemons
2413 - Added more unittests
2414 - Fixed critical bug in ganeti-masterd startup
2415 - Removed the configure-time ``kvm-migration-port`` parameter, this is
2416 now customisable at the cluster level for both the KVM and Xen
2417 hypervisors using the new ``migration_port`` parameter
2418 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
2419 reinstalling an instance
2420 - Allowed ``@`` in tag names
2421 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
2422 - Many documentation updates
2423 - Distribute hypervisor files on ``gnt-cluster redist-conf``
2424 - ``gnt-instance reinstall`` can now reinstall multiple instances
2425 - Updated many command line parameters
2426 - Introduced new OS API version 15
2427 - No longer support a default hypervisor
2428 - Treat virtual LVs as inexistent
2429 - Improved job locking logic to reduce lock contention
2430 - Match instance and node names case insensitively
2431 - Reimplemented bash completion script to be more complete
2438 *(Released Thu, 4 Feb 2010)*
2440 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
2441 - Fix a string formatting bug
2442 - Improve safety of the code in some error paths
2443 - Improve data validation in the master of values returned from nodes
2449 *(Released Thu, 17 Dec 2009)*
2451 - Fix security issue due to missing validation of iallocator names; this
2452 allows local and remote execution of arbitrary executables
2453 - Fix failure of gnt-node list during instance removal
2454 - Ship the RAPI documentation in the archive
2460 *(Released Wed, 30 Sep 2009)*
2462 - Fixed many wrong messages
2463 - Fixed a few bugs related to the locking library
2464 - Fixed MAC checking at instance creation time
2465 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
2466 - Fixed a few issues related to signal handling in both daemons and
2468 - Fixed the example startup script provided
2469 - Fixed insserv dependencies in the example startup script (patch from
2471 - Fixed handling of drained nodes in the iallocator framework
2472 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
2474 - Fixed error related to invalid job IDs in job polling
2475 - Fixed job/opcode persistence on unclean master shutdown
2476 - Fixed handling of partial job processing after unclean master
2478 - Fixed error reporting from LUs, previously all errors were converted
2479 into execution errors
2480 - Fixed error reporting from burnin
2481 - Decreased significantly the memory usage of the job queue
2482 - Optimised slightly multi-job submission
2483 - Optimised slightly opcode loading
2484 - Backported the multi-job submit framework from the development
2485 branch; multi-instance start and stop should be faster
2486 - Added script to clean archived jobs after 21 days; this will reduce
2487 the size of the queue directory
2488 - Added some extra checks in disk size tracking
2489 - Added an example ethers hook script
2490 - Added a cluster parameter that prevents Ganeti from modifying of
2492 - Added more node information to RAPI responses
2493 - Added a ``gnt-job watch`` command that allows following the ouput of a
2495 - Added a bind-address option to ganeti-rapi
2496 - Added more checks to the configuration verify
2497 - Enhanced the burnin script such that some operations can be retried
2499 - Converted instance reinstall to multi-instance model
2505 *(Released Fri, 7 Aug 2009)*
2507 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
2508 to allow using the pre-2.0.2 behaviour in activation, if any existing
2509 instances have mismatched disk sizes in the configuration
2510 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
2511 any configuration mismatches for disk sizes
2512 - Added ``gnt-master cluste-failover --no-voting`` to allow master
2513 failover to work on two-node clusters
2514 - Fixed the ``--net`` option of ``gnt-backup import``, which was
2516 - Fixed detection of OS script errors in ``gnt-backup export``
2517 - Fixed exit code of ``gnt-backup export``
2523 *(Released Fri, 17 Jul 2009)*
2525 - Added experimental support for stripped logical volumes; this should
2526 enhance performance but comes with a higher complexity in the block
2527 device handling; stripping is only enabled when passing
2528 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
2529 affected even in the non-stripped mode
2530 - Improved resiliency against transient failures at the end of DRBD
2531 resyncs, and in general of DRBD resync checks
2532 - Fixed a couple of issues with exports and snapshot errors
2533 - Fixed a couple of issues in instance listing
2534 - Added display of the disk size in ``gnt-instance info``
2535 - Fixed checking for valid OSes in instance creation
2536 - Fixed handling of the "vcpus" parameter in instance listing and in
2537 general of invalid parameters
2538 - Fixed http server library, and thus RAPI, to handle invalid
2539 username/password combinations correctly; this means that now they
2540 report unauthorized for queries too, not only for modifications,
2541 allowing earlier detect of configuration problems
2542 - Added a new "role" node list field, equivalent to the master/master
2543 candidate/drained/offline flags combinations
2544 - Fixed cluster modify and changes of candidate pool size
2545 - Fixed cluster verify error messages for wrong files on regular nodes
2546 - Fixed a couple of issues with node demotion from master candidate role
2547 - Fixed node readd issues
2548 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
2549 - Added a new ``--no-voting`` option for masterfailover to fix failover
2550 on two-nodes clusters when the former master node is unreachable
2551 - Added instance reinstall over RAPI
2557 *(Released Tue, 16 Jun 2009)*
2559 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
2560 allow re-adding the start in single-user option (regression from 1.2)
2561 - the watcher writes the instance status to a file, to allow monitoring
2562 to report the instance status (from the master) based on cached
2563 results of the watcher's queries; while this can get stale if the
2564 watcher is being locked due to other work on the cluster, this is
2565 still an improvement
2566 - the watcher now also restarts the node daemon and the rapi daemon if
2568 - fixed the watcher to handle full and drained queue cases
2569 - hooks export more instance data in the environment, which helps if
2570 hook scripts need to take action based on the instance's properties
2571 (no longer need to query back into ganeti)
2572 - instance failovers when the instance is stopped do not check for free
2573 RAM, so that failing over a stopped instance is possible in low memory
2575 - rapi uses queries for tags instead of jobs (for less job traffic), and
2576 for cluster tags it won't talk to masterd at all but read them from
2578 - a couple of error handling fixes in RAPI
2579 - drbd handling: improved the error handling of inconsistent disks after
2580 resync to reduce the frequency of "there are some degraded disks for
2581 this instance" messages
2582 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
2583 error handling path called a wrong function name)
2589 *(Released Wed, 27 May 2009)*
2591 - no changes from rc5
2597 *(Released Wed, 20 May 2009)*
2599 - fix a couple of bugs (validation, argument checks)
2600 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
2601 - some small improvements to RAPI and IAllocator
2602 - make watcher automatically start the master daemon if down
2608 *(Released Mon, 27 Apr 2009)*
2610 - change the OS list to not require locks; this helps with big clusters
2611 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
2612 volume group is broken
2613 - ``gnt-instance info``, without any arguments, doesn't run for all
2614 instances anymore; either pass ``--all`` or pass the desired
2615 instances; this helps against mistakes on big clusters where listing
2616 the information for all instances takes a long time
2617 - miscellaneous doc and man pages fixes
2623 *(Released Wed, 8 Apr 2009)*
2625 - Change the internal locking model of some ``gnt-node`` commands, in
2626 order to reduce contention (and blocking of master daemon) when
2627 batching many creation/reinstall jobs
2628 - Fixes to Xen soft reboot
2629 - No longer build documentation at build time, instead distribute it in
2630 the archive, in order to reduce the need for the whole docbook/rst
2637 *(Released Fri, 27 Mar 2009)*
2639 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
2640 - Fix watcher startup sequence, improves the behaviour of busy clusters
2641 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
2642 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
2643 bind address and other places
2644 - Some documentation fixes and updates
2650 *(Released Mon, 2 Mar 2009)*
2652 - More documentation updates, now all docs should be more-or-less
2654 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
2656 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
2658 - Fix an issue related to $libdir/run/ganeti and cluster creation
2664 *(Released Thu, 19 Feb 2009)*
2666 - Xen PVM and KVM have switched the default value for the instance root
2667 disk to the first partition on the first drive, instead of the whole
2668 drive; this means that the OS installation scripts must be changed
2670 - Man pages have been updated
2671 - RAPI has been switched by default to HTTPS, and the exported functions
2672 should all work correctly
2673 - RAPI v1 has been removed
2674 - Many improvements to the KVM hypervisor
2675 - Block device errors are now better reported
2676 - Many other bugfixes and small improvements
2682 *(Released Mon, 26 Jan 2009)*
2684 - Version 2 is a general rewrite of the code and therefore the
2685 differences are too many to list, see the design document for 2.0 in
2686 the ``doc/`` subdirectory for more details
2687 - In this beta version there is not yet a migration path from 1.2 (there
2688 will be one in the final 2.0 release)
2689 - A few significant changes are:
2691 - all commands are executed by a daemon (``ganeti-masterd``) and the
2692 various ``gnt-*`` commands are just front-ends to it
2693 - all the commands are entered into, and executed from a job queue,
2694 see the ``gnt-job(8)`` manpage
2695 - the RAPI daemon supports read-write operations, secured by basic
2696 HTTP authentication on top of HTTPS
2697 - DRBD version 0.7 support has been removed, DRBD 8 is the only
2698 supported version (when migrating from Ganeti 1.2 to 2.0, you need
2699 to migrate to DRBD 8 first while still running Ganeti 1.2)
2700 - DRBD devices are using statically allocated minor numbers, which
2701 will be assigned to existing instances during the migration process
2702 - there is support for both Xen PVM and Xen HVM instances running on
2704 - KVM virtualization is supported too
2705 - file-based storage has been implemented, which means that it is
2706 possible to run the cluster without LVM and DRBD storage, for
2707 example using a shared filesystem exported from shared storage (and
2708 still have live migration)
2714 *(Released Tue, 13 Jan 2009)*
2716 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
2717 - Reuse the old instance mac address by default on instance import, if
2718 the instance name is the same.
2719 - Handle situations in which the node info rpc returns incomplete
2721 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
2722 - Improved version of batcher:
2724 - state file support
2725 - instance mac address support
2726 - support for HVM clusters/instances
2728 - Add an option to show the number of cpu sockets and nodes in
2730 - Support OSes that handle more than one version of the OS api (but do
2731 not change the current API in any other way)
2732 - Fix ``gnt-node migrate``
2733 - ``gnt-debug`` man page
2734 - Fixes various more typos and small issues
2735 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
2741 *(Released Wed, 24 Sep 2008)*
2743 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
2744 type of disk exported to fully virtualized instances.
2745 - provide access to the serial console of HVM instances
2746 - instance auto_balance flag, set by default. If turned off it will
2747 avoid warnings on cluster verify if there is not enough memory to fail
2748 over an instance. in the future it will prevent automatically failing
2749 it over when we will support that.
2750 - batcher tool for instance creation, see ``tools/README.batcher``
2751 - ``gnt-instance reinstall --select-os`` to interactively select a new
2752 operating system when reinstalling an instance.
2753 - when changing the memory amount on instance modify a check has been
2754 added that the instance will be able to start. also warnings are
2755 emitted if the instance will not be able to fail over, if auto_balance
2757 - documentation fixes
2758 - sync fields between ``gnt-instance list/modify/add/import``
2759 - fix a race condition in drbd when the sync speed was set after giving
2760 the device a remote peer.
2766 *(Released Tue, 22 Jul 2008)*
2768 - note: the allowed size and number of tags per object were reduced
2769 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
2770 - fixed twisted 8.x compatibility
2771 - fixed ``gnt-instance replace-disks`` with iallocator
2772 - add TCP keepalives on twisted connections to detect restarted nodes
2773 - disk increase support, see ``gnt-instance grow-disk``
2774 - implement bulk node/instance query for RAPI
2775 - add tags in node/instance listing (optional)
2776 - experimental migration (and live migration) support, read the man page
2777 for ``gnt-instance migrate``
2778 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
2779 has some small improvements in handling its state file
2785 *(Released Fri, 13 Jun 2008)*
2787 - Experimental readonly, REST-based remote API implementation;
2788 automatically started on master node, TCP port 5080, if enabled by
2789 ``--enable-rapi`` parameter to configure script.
2790 - Instance allocator support. Add and import instance accept a
2791 ``--iallocator`` parameter, and call that instance allocator to decide
2792 which node to use for the instance. The iallocator document describes
2793 what's expected from an allocator script.
2794 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
2795 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
2796 node is lost there is still enough memory to fail over the instances
2798 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
2799 ``gnt-cluster verify``, to check for site-specific compliance. All the
2800 hooks will run, and their output, if any, will be displayed. Any
2801 failing hook will make the verification return an error value.
2802 - ``gnt-cluster verify`` now checks that its peers are reachable on the
2803 primary and secondary interfaces
2804 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
2805 that is still declared as part of the cluster and has failed.
2806 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
2807 specifying output fields, that just adds the chosen fields to the
2809 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
2810 export from the filesystem.
2811 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
2812 have been added. Using them you can enable/disable acpi and pae
2813 support, and specify a path for a cd image to be exported to the
2814 instance. These parameters as the name suggest only work on HVM
2816 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
2817 PAE support will be set to the previously hardcoded values, but the
2818 (previously hardcoded) path to the CDROM ISO image will be unset and
2819 if required, needs to be set manually with ``gnt-instance modify``
2821 - The address to which an instance's VNC console is bound is now
2822 selectable per-instance, rather than being cluster wide. Of course
2823 this only applies to instances controlled via VNC, so currently just
2824 applies to HVM clusters.
2830 *(Released Mon, 18 Feb 2008)*
2832 - more tweaks to the disk activation code (especially helpful for DRBD)
2833 - change the default ``gnt-instance list`` output format, now there is
2834 one combined status field (see the manpage for the exact values this
2836 - some more fixes for the mac export to hooks change
2837 - make Ganeti not break with DRBD 8.2.x (which changed the version
2838 format in ``/proc/drbd``) (issue 24)
2839 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
2840 template, allowing migration from DRBD0.7+MD to DRBD8
2846 *(Released Wed, 30 Jan 2008)*
2848 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
2850 - add command aliases infrastructure and a few aliases
2851 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
2852 man pages and the ``--help`` option of ``gnt-node
2853 list``/``gnt-instance list``
2854 - fix ``gnt-backup list`` with down nodes (issue 21)
2855 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
2856 - fix the dist archive and add a check for including svn/git files in
2858 - some developer-related changes: improve the burnin and the QA suite,
2859 add an upload script for testing during development
2865 *(Released Wed, 16 Jan 2008)*
2867 - experimental HVM support, read the install document, section
2868 "Initializing the cluster"
2869 - allow for the PVM hypervisor per-instance kernel and initrd paths
2870 - add a new command ``gnt-cluster verify-disks`` which uses a new
2871 algorithm to improve the reconnection of the DRBD pairs if the device
2872 on the secondary node has gone away
2873 - make logical volume code auto-activate LVs at disk activation time
2874 - slightly improve the speed of activating disks
2875 - allow specification of the MAC address at instance creation time, and
2876 changing it later via ``gnt-instance modify``
2877 - fix handling of external commands that generate lots of output on
2879 - update documentation with regard to minimum version of DRBD8 supported
2885 *(Released Tue, 4 Dec 2007)*
2887 - Log the ``xm create`` output to the node daemon log on failure (to
2888 help diagnosing the error)
2889 - In debug mode, log all external commands output if failed to the logs
2890 - Change parsing of lvm commands to ignore stderr
2896 *(Released Wed, 28 Nov 2007)*
2898 - Another round of updates to the DRBD 8 code to deal with more failures
2899 in the replace secondary node operation
2900 - Some more logging of failures in disk operations (lvm, drbd)
2901 - A few documentation updates
2908 *(Released Tue, 13 Nov 2007)*
2910 - Change configuration file format from Python's Pickle to JSON.
2911 Upgrading is possible using the cfgupgrade utility.
2912 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
2913 faster replace disks and is more stable (DRBD 8 has many improvements
2914 compared to DRBD 0.7)
2915 - Added command line tags support (see man pages for ``gnt-instance``,
2916 ``gnt-node``, ``gnt-cluster``)
2917 - Added instance rename support
2918 - Added multi-instance startup/shutdown
2919 - Added cluster rename support
2920 - Added ``gnt-node evacuate`` to simplify some node operations
2921 - Added instance reboot operation that can speedup reboot as compared to
2923 - Soften the requirement that hostnames are in FQDN format
2924 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
2926 - Removed dependency on debian's patched fping that uses the
2927 non-standard ``-S`` option
2928 - Now the OS definitions are searched for in multiple, configurable
2929 paths (easier for distros to package)
2930 - Some changes to the hooks infrastructure (especially the new
2931 post-configuration update hook)
2932 - Other small bugfixes
2934 .. vim: set textwidth=72 syntax=rst :