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 Mon, 30 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).
165 - Perform proper cleanup on termination of Haskell daemons
166 - Fix corner-case in handling of remaining retry time
172 *(Released Tue, 17 Sep 2013)*
174 - To simplify the work of packaging frameworks that want to add the needed users
175 and groups in a split-user setup themselves, at build time three files in
176 ``doc/users`` will be generated. The ``groups`` files contains, one per line,
177 the groups to be generated, the ``users`` file contains, one per line, the
178 users to be generated, optionally followed by their primary group, where
179 important. The ``groupmemberships`` file contains, one per line, additional
180 user-group membership relations that need to be established. The syntax of
181 these files will remain stable in all future versions.
182 - Add a default to file-driver when unspecified over RAPI (Issue 571)
183 - Mark the DSA host pubkey as optional, and remove it during config downgrade
185 - Some documentation fixes
191 *(Released Tue, 27 Aug 2013)*
193 The second release candidate of the 2.8 series. Since 2.8.0. rc1:
195 - Support 'viridian' parameter in Xen HVM (Issue 233)
196 - Include VCS version in ``gnt-cluster version``
197 - Support DSA SSH keys in bootstrap (Issue 338)
198 - Fix batch creation of instances
199 - Use FQDN to check master node status (Issue 551)
200 - Make the DRBD collector more failure-resilient
206 *(Released Fri, 2 Aug 2013)*
208 The first release candidate of the 2.8 series. Since 2.8.0 beta1:
210 - Fix upgrading/downgrading from 2.7
211 - Increase maximum RAPI message size
212 - Documentation updates
213 - Split ``confd`` between ``luxid`` and ``confd``
214 - Merge 2.7 series up to the 2.7.1 release
215 - Allow the ``modify_etc_hosts`` option to be changed
216 - Add better debugging for ``luxid`` queries
217 - Expose bulk parameter for GetJobs in RAPI client
218 - Expose missing ``network`` fields in RAPI
219 - Add some ``cluster verify`` tests
220 - Some unittest fixes
221 - Fix a malfunction in ``hspace``'s tiered allocation
222 - Fix query compatibility between haskell and python implementations
223 - Add the ``vnet_hdr`` HV parameter for KVM
224 - Add ``--cleanup`` to instance failover
225 - Change the connected groups format in ``gnt-network info`` output; it
226 was previously displayed as a raw list by mistake. (Merged from 2.7)
232 *(Released Mon, 24 Jun 2013)*
234 This was the first beta release of the 2.8 series. All important changes
235 are listed in the latest 2.8 entry.
241 *(Released Thu, 26 Sep 2013)*
243 - Change the connected groups format in ``gnt-network info`` output; it
244 was previously displayed as a raw list by mistake
245 - Check disk template in right dict when copying
246 - Support multi-instance allocs without iallocator
247 - Fix some errors in the documentation
248 - Fix formatting of tuple in an error message
254 *(Released Thu, 25 Jul 2013)*
256 - Add logrotate functionality in daemon-util
257 - Add logrotate example file
258 - Add missing fields to network queries over rapi
259 - Fix network object timestamps
260 - Add support for querying network timestamps
261 - Fix a typo in the example crontab
262 - Fix a documentation typo
268 *(Released Thu, 04 Jul 2013)*
270 Incompatible/important changes
271 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
273 - Instance policies for disk size were documented to be on a per-disk
274 basis, but hail applied them to the sum of all disks. This has been
276 - ``hbal`` will now exit with status 0 if, during job execution over
277 LUXI, early exit has been requested and all jobs are successful;
278 before, exit status 1 was used, which cannot be differentiated from
280 - Compatibility with newer versions of rbd has been fixed
281 - ``gnt-instance batch-create`` has been changed to use the bulk create
282 opcode from Ganeti. This lead to incompatible changes in the format of
283 the JSON file. It's now not a custom dict anymore but a dict
284 compatible with the ``OpInstanceCreate`` opcode.
285 - Parent directories for file storage need to be listed in
286 ``$sysconfdir/ganeti/file-storage-paths`` now. ``cfgupgrade`` will
287 write the file automatically based on old configuration values, but it
288 can not distribute it across all nodes and the file contents should be
289 verified. Use ``gnt-cluster copyfile
290 $sysconfdir/ganeti/file-storage-paths`` once the cluster has been
291 upgraded. The reason for requiring this list of paths now is that
292 before it would have been possible to inject new paths via RPC,
293 allowing files to be created in arbitrary locations. The RPC protocol
294 is protected using SSL/X.509 certificates, but as a design principle
295 Ganeti does not permit arbitrary paths to be passed.
296 - The parsing of the variants file for OSes (see
297 :manpage:`ganeti-os-interface(7)`) has been slightly changed: now empty
298 lines and comment lines (starting with ``#``) are ignored for better
300 - The ``setup-ssh`` tool added in Ganeti 2.2 has been replaced and is no
301 longer available. ``gnt-node add`` now invokes a new tool on the
302 destination node, named ``prepare-node-join``, to configure the SSH
303 daemon. Paramiko is no longer necessary to configure nodes' SSH
304 daemons via ``gnt-node add``.
305 - Draining (``gnt-cluster queue drain``) and un-draining the job queue
306 (``gnt-cluster queue undrain``) now affects all nodes in a cluster and
307 the flag is not reset after a master failover.
308 - Python 2.4 has *not* been tested with this release. Using 2.6 or above
309 is recommended. 2.6 will be mandatory from the 2.8 series.
315 - New network management functionality to support automatic allocation
316 of IP addresses and managing of network parameters. See
317 :manpage:`gnt-network(8)` for more details.
318 - New external storage backend, to allow managing arbitrary storage
319 systems external to the cluster. See
320 :manpage:`ganeti-extstorage-interface(7)`.
321 - New ``exclusive-storage`` node parameter added, restricted to
322 nodegroup level. When it's set to true, physical disks are assigned in
323 an exclusive fashion to instances, as documented in :doc:`Partitioned
324 Ganeti <design-partitioned>`. Currently, only instances using the
325 ``plain`` disk template are supported.
326 - The KVM hypervisor has been updated with many new hypervisor
327 parameters, including a generic one for passing arbitrary command line
328 values. See a complete list in :manpage:`gnt-instance(8)`. It is now
329 compatible up to qemu 1.4.
330 - A new tool, called ``mon-collector``, is the stand-alone executor of
331 the data collectors for a monitoring system. As of this version, it
332 just includes the DRBD data collector, that can be executed by calling
333 ``mon-collector`` using the ``drbd`` parameter. See
334 :manpage:`mon-collector(7)`.
335 - A new user option, :pyeval:`rapi.RAPI_ACCESS_READ`, has been added
336 for RAPI users. It allows granting permissions to query for
337 information to a specific user without giving
338 :pyeval:`rapi.RAPI_ACCESS_WRITE` permissions.
339 - A new tool named ``node-cleanup`` has been added. It cleans remains of
340 a cluster from a machine by stopping all daemons, removing
341 certificates and ssconf files. Unless the ``--no-backup`` option is
342 given, copies of the certificates are made.
343 - Instance creations now support the use of opportunistic locking,
344 potentially speeding up the (parallel) creation of multiple instances.
345 This feature is currently only available via the :doc:`RAPI
346 <rapi>` interface and when an instance allocator is used. If the
347 ``opportunistic_locking`` parameter is set the opcode will try to
348 acquire as many locks as possible, but will not wait for any locks
349 held by other opcodes. If not enough resources can be found to
350 allocate the instance, the temporary error code
351 :pyeval:`errors.ECODE_TEMP_NORES` is returned. The operation can be
352 retried thereafter, with or without opportunistic locking.
353 - New experimental linux-ha resource scripts.
354 - Restricted-commands support: ganeti can now be asked (via command line
355 or rapi) to perform commands on a node. These are passed via ganeti
356 RPC rather than ssh. This functionality is restricted to commands
357 specified on the ``$sysconfdir/ganeti/restricted-commands`` for security
358 reasons. The file is not copied automatically.
364 - Diskless instances are now externally mirrored (Issue 237). This for
365 now has only been tested in conjunction with explicit target nodes for
367 - Queries not needing locks or RPC access to the node can now be
368 performed by the confd daemon, making them independent from jobs, and
369 thus faster to execute. This is selectable at configure time.
370 - The functionality for allocating multiple instances at once has been
371 overhauled and is now also available through :doc:`RAPI <rapi>`.
373 There are no significant changes from version 2.7.0~rc3.
379 *(Released Tue, 25 Jun 2013)*
381 - Fix permissions on the confd query socket (Issue 477)
382 - Fix permissions on the job archive dir (Issue 498)
383 - Fix handling of an internal exception in replace-disks (Issue 472)
384 - Fix gnt-node info handling of shortened names (Issue 497)
385 - Fix gnt-instance grow-disk when wiping is enabled
386 - Documentation improvements, and support for newer pandoc
387 - Fix hspace honoring ipolicy for disks (Issue 484)
388 - Improve handling of the ``kvm_extra`` HV parameter
394 *(Released Fri, 24 May 2013)*
396 - ``devel/upload`` now works when ``/var/run`` on the target nodes is a
398 - Disks added through ``gnt-instance modify`` or created through
399 ``gnt-instance recreate-disks`` are wiped, if the
400 ``prealloc_wipe_disks`` flag is set.
401 - If wiping newly created disks fails, the disks are removed. Also,
402 partial failures in creating disks through ``gnt-instance modify``
403 triggers a cleanup of the partially-created disks.
404 - Removing the master IP address doesn't fail if the address has been
406 - Fix ownership of the OS log dir
407 - Workaround missing SO_PEERCRED constant (Issue 191)
413 *(Released Fri, 3 May 2013)*
415 This was the first release candidate of the 2.7 series. Since beta3:
417 - Fix kvm compatibility with qemu 1.4 (Issue 389)
418 - Documentation updates (admin guide, upgrade notes, install
419 instructions) (Issue 372)
420 - Fix gnt-group list nodes and instances count (Issue 436)
421 - Fix compilation without non-mandatory libraries (Issue 441)
422 - Fix xen-hvm hypervisor forcing nics to type 'ioemu' (Issue 247)
423 - Make confd logging more verbose at INFO level (Issue 435)
424 - Improve "networks" documentation in :manpage:`gnt-instance(8)`
425 - Fix failure path for instance storage type conversion (Issue 229)
426 - Update htools text backend documentation
427 - Improve the renew-crypto section of :manpage:`gnt-cluster(8)`
428 - Disable inter-cluster instance move for file-based instances, because
429 it is dependant on instance export, which is not supported for
430 file-based instances. (Issue 414)
431 - Fix gnt-job crashes on non-ascii characters (Issue 427)
432 - Fix volume group checks on non-vm-capable nodes (Issue 432)
438 *(Released Mon, 22 Apr 2013)*
440 This was the third beta release of the 2.7 series. Since beta2:
442 - Fix hail to verify disk instance policies on a per-disk basis (Issue 418).
443 - Fix data loss on wrong usage of ``gnt-instance move``
444 - Properly export errors in confd-based job queries
445 - Add ``users-setup`` tool
446 - Fix iallocator protocol to report 0 as a disk size for diskless
447 instances. This avoids hail breaking when a diskless instance is
449 - Fix job queue directory permission problem that made confd job queries
450 fail. This requires running an ``ensure-dirs --full-run`` on upgrade
451 for access to archived jobs (Issue 406).
452 - Limit the sizes of networks supported by ``gnt-network`` to something
453 between a ``/16`` and a ``/30`` to prevent memory bloat and crashes.
454 - Fix bugs in instance disk template conversion
455 - Fix GHC 7 compatibility
456 - Fix ``burnin`` install path (Issue 426).
457 - Allow very small disk grows (Issue 347).
458 - Fix a ``ganeti-noded`` memory bloat introduced in 2.5, by making sure
459 that noded doesn't import masterd code (Issue 419).
460 - Make sure the default metavg at cluster init is the same as the vg, if
461 unspecified (Issue 358).
462 - Fix cleanup of partially created disks (part of Issue 416)
468 *(Released Tue, 2 Apr 2013)*
470 This was the second beta release of the 2.7 series. Since beta1:
472 - Networks no longer have a "type" slot, since this information was
473 unused in Ganeti: instead of it tags should be used.
474 - The rapi client now has a ``target_node`` option to MigrateInstance.
475 - Fix early exit return code for hbal (Issue 386).
476 - Fix ``gnt-instance migrate/failover -n`` (Issue 396).
477 - Fix ``rbd showmapped`` output parsing (Issue 312).
478 - Networks are now referenced indexed by UUID, rather than name. This
479 will require running cfgupgrade, from 2.7.0beta1, if networks are in
481 - The OS environment now includes network information.
482 - Deleting of a network is now disallowed if any instance nic is using
483 it, to prevent dangling references.
484 - External storage is now documented in man pages.
485 - The exclusive_storage flag can now only be set at nodegroup level.
486 - Hbal can now submit an explicit priority with its jobs.
487 - Many network related locking fixes.
488 - Bump up the required pylint version to 0.25.1.
489 - Fix the ``no_remember`` option in RAPI client.
490 - Many ipolicy related tests, qa, and fixes.
491 - Many documentation improvements and fixes.
492 - Fix building with ``--disable-file-storage``.
493 - Fix ``-q`` option in htools, which was broken if passed more than
495 - Some haskell/python interaction improvements and fixes.
496 - Fix iallocator in case of missing LVM storage.
497 - Fix confd config load in case of ``--no-lvm-storage``.
498 - The confd/query functionality is now mentioned in the security
505 *(Released Wed, 6 Feb 2013)*
507 This was the first beta release of the 2.7 series. All important changes
508 are listed in the latest 2.7 entry.
514 *(Released Fri, 21 Dec 2012)*
516 Important behaviour change: hbal won't rebalance anymore instances which
517 have the ``auto_balance`` attribute set to false. This was the intention
518 all along, but until now it only skipped those from the N+1 memory
519 reservation (DRBD-specific).
521 A significant number of bug fixes in this release:
523 - Fixed disk adoption interaction with ipolicy checks.
524 - Fixed networking issues when instances are started, stopped or
525 migrated, by forcing the tap device's MAC prefix to "fe" (issue 217).
526 - Fixed the warning in cluster verify for shared storage instances not
528 - Fixed removal of storage directory on shared file storage (issue 262).
529 - Fixed validation of LVM volume group name in OpClusterSetParams
530 (``gnt-cluster modify``) (issue 285).
531 - Fixed runtime memory increases (``gnt-instance modify -m``).
532 - Fixed live migration under Xen's ``xl`` mode.
533 - Fixed ``gnt-instance console`` with ``xl``.
534 - Fixed building with newer Haskell compiler/libraries.
535 - Fixed PID file writing in Haskell daemons (confd); this prevents
536 restart issues if confd was launched manually (outside of
537 ``daemon-util``) while another copy of it was running
538 - Fixed a type error when doing live migrations with KVM (issue 297) and
539 the error messages for failing migrations have been improved.
540 - Fixed opcode validation for the out-of-band commands (``gnt-node
542 - Fixed a type error when unsetting OS hypervisor parameters (issue
543 311); now it's possible to unset all OS-specific hypervisor
545 - Fixed the ``dry-run`` mode for many operations: verification of
546 results was over-zealous but didn't take into account the ``dry-run``
547 operation, resulting in "wrong" failures.
548 - Fixed bash completion in ``gnt-job list`` when the job queue has
549 hundreds of entries; especially with older ``bash`` versions, this
550 results in significant CPU usage.
552 And lastly, a few other improvements have been made:
554 - Added option to force master-failover without voting (issue 282).
555 - Clarified error message on lock conflict (issue 287).
556 - Logging of newly submitted jobs has been improved (issue 290).
557 - Hostname checks have been made uniform between instance rename and
559 - The ``--submit`` option is now supported by ``gnt-debug delay``.
560 - Shutting down the master daemon by sending SIGTERM now stops it from
561 processing jobs waiting for locks; instead, those jobs will be started
562 once again after the master daemon is started the next time (issue
564 - Support for Xen's ``xl`` program has been improved (besides the fixes
566 - Reduced logging noise in the Haskell confd daemon (only show one log
567 entry for each config reload, instead of two).
568 - Several man page updates and typo fixes.
574 *(Released Fri, 12 Oct 2012)*
576 A small bugfix release. Among the bugs fixed:
578 - Fixed double use of ``PRIORITY_OPT`` in ``gnt-node migrate``, that
579 made the command unusable.
580 - Commands that issue many jobs don't fail anymore just because some jobs
581 take so long that other jobs are archived.
582 - Failures during ``gnt-instance reinstall`` are reflected by the exit
584 - Issue 190 fixed. Check for DRBD in cluster verify is enabled only when
586 - When ``always_failover`` is set, ``--allow-failover`` is not required
587 in migrate commands anymore.
588 - ``bash_completion`` works even if extglob is disabled.
589 - Fixed bug with locks that made failover for RDB-based instances fail.
590 - Fixed bug in non-mirrored instance allocation that made Ganeti choose
591 a random node instead of one based on the allocator metric.
592 - Support for newer versions of pylint and pep8.
593 - Hail doesn't fail anymore when trying to add an instance of type
594 ``file``, ``sharedfile`` or ``rbd``.
595 - Added new Makefile target to rebuild the whole distribution, so that
596 all files are included.
602 *(Released Fri, 27 Jul 2012)*
605 .. attention:: The ``LUXI`` protocol has been made more consistent
606 regarding its handling of command arguments. This, however, leads to
607 incompatibility issues with previous versions. Please ensure that you
608 restart Ganeti daemons soon after the upgrade, otherwise most
609 ``LUXI`` calls (job submission, setting/resetting the drain flag,
610 pausing/resuming the watcher, cancelling and archiving jobs, querying
611 the cluster configuration) will fail.
620 The current ``admin_up`` field, which used to denote whether an instance
621 should be running or not, has been removed. Instead, ``admin_state`` is
622 introduced, with 3 possible values -- ``up``, ``down`` and ``offline``.
624 The rational behind this is that an instance being “down” can have
627 - it could be down during a reboot
628 - it could be temporarily be down for a reinstall
629 - or it could be down because it is deprecated and kept just for its
632 The previous Boolean state was making it difficult to do capacity
633 calculations: should Ganeti reserve memory for a down instance? Now, the
634 tri-state field makes it clear:
636 - in ``up`` and ``down`` state, all resources are reserved for the
637 instance, and it can be at any time brought up if it is down
638 - in ``offline`` state, only disk space is reserved for it, but not
641 The field can have an extra use: since the transition between ``up`` and
642 ``down`` and vice-versus is done via ``gnt-instance start/stop``, but
643 transition between ``offline`` and ``down`` is done via ``gnt-instance
644 modify``, it is possible to given different rights to users. For
645 example, owners of an instance could be allowed to start/stop it, but
646 not transition it out of the offline state.
648 Instance policies and specs
649 +++++++++++++++++++++++++++
651 In previous Ganeti versions, an instance creation request was not
652 limited on the minimum size and on the maximum size just by the cluster
653 resources. As such, any policy could be implemented only in third-party
654 clients (RAPI clients, or shell wrappers over ``gnt-*``
655 tools). Furthermore, calculating cluster capacity via ``hspace`` again
656 required external input with regards to instance sizes.
658 In order to improve these workflows and to allow for example better
659 per-node group differentiation, we introduced instance specs, which
662 - minimum instance disk size, disk count, memory size, cpu count
663 - maximum values for the above metrics
664 - and “standard” values (used in ``hspace`` to calculate the standard
667 The minimum/maximum values can be also customised at node-group level,
668 for example allowing more powerful hardware to support bigger instance
671 Beside the instance specs, there are a few other settings belonging to
672 the instance policy framework. It is possible now to customise, per
673 cluster and node-group:
675 - the list of allowed disk templates
676 - the maximum ratio of VCPUs per PCPUs (to control CPU oversubscription)
677 - the maximum ratio of instance to spindles (see below for more
678 information) for local storage
680 All these together should allow all tools that talk to Ganeti to know
681 what are the ranges of allowed values for instances and the
682 over-subscription that is allowed.
684 For the VCPU/PCPU ratio, we already have the VCPU configuration from the
685 instance configuration, and the physical CPU configuration from the
686 node. For the spindle ratios however, we didn't track before these
687 values, so new parameters have been added:
689 - a new node parameter ``spindle_count``, defaults to 1, customisable at
690 node group or node level
691 - at new backend parameter (for instances), ``spindle_use`` defaults to 1
693 Note that spindles in this context doesn't need to mean actual
694 mechanical hard-drives; it's just a relative number for both the node
695 I/O capacity and instance I/O consumption.
697 Instance migration behaviour
698 ++++++++++++++++++++++++++++
700 While live-migration is in general desirable over failover, it is
701 possible that for some workloads it is actually worse, due to the
702 variable time of the “suspend” phase during live migration.
704 To allow the tools to work consistently over such instances (without
705 having to hard-code instance names), a new backend parameter
706 ``always_failover`` has been added to control the migration/failover
707 behaviour. When set to True, all migration requests for an instance will
708 instead fall-back to failover.
710 Instance memory ballooning
711 ++++++++++++++++++++++++++
713 Initial support for memory ballooning has been added. The memory for an
714 instance is no longer fixed (backend parameter ``memory``), but instead
715 can vary between minimum and maximum values (backend parameters
716 ``minmem`` and ``maxmem``). Currently we only change an instance's
719 - live migrating or failing over and instance and the target node
720 doesn't have enough memory
721 - user requests changing the memory via ``gnt-instance modify
727 In order to control the use of specific CPUs by instance, support for
728 controlling CPU pinning has been added for the Xen, HVM and LXC
729 hypervisors. This is controlled by a new hypervisor parameter
730 ``cpu_mask``; details about possible values for this are in the
731 :manpage:`gnt-instance(8)`. Note that use of the most specific (precise
732 VCPU-to-CPU mapping) form will work well only when all nodes in your
733 cluster have the same amount of CPUs.
738 Another area in which Ganeti was not customisable were the parameters
739 used for storage configuration, e.g. how many stripes to use for LVM,
740 DRBD resync configuration, etc.
742 To improve this area, we've added disks parameters, which are
743 customisable at cluster and node group level, and which allow to
744 specify various parameters for disks (DRBD has the most parameters
745 currently), for example:
747 - DRBD resync algorithm and parameters (e.g. speed)
748 - the default VG for meta-data volumes for DRBD
749 - number of stripes for LVM (plain disk template)
752 These parameters can be modified via ``gnt-cluster modify -D …`` and
753 ``gnt-group modify -D …``, and are used at either instance creation (in
754 case of LVM stripes, for example) or at disk “activation” time
757 Rados block device support
758 ++++++++++++++++++++++++++
760 A Rados (http://ceph.com/wiki/Rbd) storage backend has been added,
761 denoted by the ``rbd`` disk template type. This is considered
762 experimental, feedback is welcome. For details on configuring it, see
763 the :doc:`install` document and the :manpage:`gnt-cluster(8)` man page.
768 The existing master IP functionality works well only in simple setups (a
769 single network shared by all nodes); however, if nodes belong to
770 different networks, then the ``/32`` setup and lack of routing
771 information is not enough.
773 To allow the master IP to function well in more complex cases, the
774 system was reworked as follows:
776 - a master IP netmask setting has been added
777 - the master IP activation/turn-down code was moved from the node daemon
779 - whether to run the Ganeti-supplied master IP script or a user-supplied
780 on is a ``gnt-cluster init`` setting
782 Details about the location of the standard and custom setup scripts are
783 in the man page :manpage:`gnt-cluster(8)`; for information about the
784 setup script protocol, look at the Ganeti-supplied script.
789 The `SPICE <http://www.linux-kvm.org/page/SPICE>`_ support has been
792 It is now possible to use TLS-protected connections, and when renewing
793 or changing the cluster certificates (via ``gnt-cluster renew-crypto``,
794 it is now possible to specify spice or spice CA certificates. Also, it
795 is possible to configure a password for SPICE sessions via the
796 hypervisor parameter ``spice_password_file``.
798 There are also new parameters to control the compression and streaming
799 options (e.g. ``spice_image_compression``, ``spice_streaming_video``,
800 etc.). For details, see the man page :manpage:`gnt-instance(8)` and look
801 for the spice parameters.
803 Lastly, it is now possible to see the SPICE connection information via
804 ``gnt-instance console``.
809 A new tool (``tools/ovfconverter``) has been added that supports
810 conversion between Ganeti and the `Open Virtualization Format
811 <http://en.wikipedia.org/wiki/Open_Virtualization_Format>`_ (both to and
814 This relies on the ``qemu-img`` tool to convert the disk formats, so the
815 actual compatibility with other virtualization solutions depends on it.
820 The configuration query daemon (``ganeti-confd``) is now optional, and
821 has been rewritten in Haskell; whether to use the daemon at all, use the
822 Python (default) or the Haskell version is selectable at configure time
823 via the ``--enable-confd`` parameter, which can take one of the
824 ``haskell``, ``python`` or ``no`` values. If not used, disabling the
825 daemon will result in a smaller footprint; for larger systems, we
826 welcome feedback on the Haskell version which might become the default
829 If you want to use ``gnt-node list-drbd`` you need to have the Haskell
830 daemon running. The Python version doesn't implement the new call.
833 User interface changes
834 ~~~~~~~~~~~~~~~~~~~~~~
836 We have replaced the ``--disks`` option of ``gnt-instance
837 replace-disks`` with a more flexible ``--disk`` option, which allows
838 adding and removing disks at arbitrary indices (Issue 188). Furthermore,
839 disk size and mode can be changed upon recreation (via ``gnt-instance
840 recreate-disks``, which accepts the same ``--disk`` option).
842 As many people are used to a ``show`` command, we have added that as an
843 alias to ``info`` on all ``gnt-*`` commands.
845 The ``gnt-instance grow-disk`` command has a new mode in which it can
846 accept the target size of the disk, instead of the delta; this can be
847 more safe since two runs in absolute mode will be idempotent, and
848 sometimes it's also easier to specify the desired size directly.
850 Also the handling of instances with regard to offline secondaries has
851 been improved. Instance operations should not fail because one of it's
852 secondary nodes is offline, even though it's safe to proceed.
854 A new command ``list-drbd`` has been added to the ``gnt-node`` script to
855 support debugging of DRBD issues on nodes. It provides a mapping of DRBD
856 minors to instance name.
861 RAPI coverage has improved, with (for example) new resources for
862 recreate-disks, node power-cycle, etc.
867 There is partial support for ``xl`` in the Xen hypervisor; feedback is
870 Python 2.7 is better supported, and after Ganeti 2.6 we will investigate
871 whether to still support Python 2.4 or move to Python 2.6 as minimum
874 Support for Fedora has been slightly improved; the provided example
875 init.d script should work better on it and the INSTALL file should
876 document the needed dependencies.
881 The deprecated ``QueryLocks`` LUXI request has been removed. Use
882 ``Query(what=QR_LOCK, ...)`` instead.
884 The LUXI requests :pyeval:`luxi.REQ_QUERY_JOBS`,
885 :pyeval:`luxi.REQ_QUERY_INSTANCES`, :pyeval:`luxi.REQ_QUERY_NODES`,
886 :pyeval:`luxi.REQ_QUERY_GROUPS`, :pyeval:`luxi.REQ_QUERY_EXPORTS` and
887 :pyeval:`luxi.REQ_QUERY_TAGS` are deprecated and will be removed in a
888 future version. :pyeval:`luxi.REQ_QUERY` should be used instead.
890 RAPI client: ``CertificateError`` now derives from
891 ``GanetiApiError``. This should make it more easy to handle Ganeti
894 Deprecation warnings due to PyCrypto/paramiko import in
895 ``tools/setup-ssh`` have been silenced, as usually they are safe; please
896 make sure to run an up-to-date paramiko version, if you use this tool.
898 The QA scripts now depend on Python 2.5 or above (the main code base
899 still works with Python 2.4).
901 The configuration file (``config.data``) is now written without
902 indentation for performance reasons; if you want to edit it, it can be
903 re-formatted via ``tools/fmtjson``.
905 A number of bugs has been fixed in the cluster merge tool.
907 ``x509`` certification verification (used in import-export) has been
908 changed to allow the same clock skew as permitted by the cluster
909 verification. This will remove some rare but hard to diagnose errors in
916 *(Released Thu, 19 Jul 2012)*
918 Very few changes from rc4 to the final release, only bugfixes:
920 - integrated fixes from release 2.5.2 (fix general boot flag for KVM
921 instance, fix CDROM booting for KVM instances)
922 - fixed node group modification of node parameters
923 - fixed issue in LUClusterVerifyGroup with multi-group clusters
924 - fixed generation of bash completion to ensure a stable ordering
931 *(Released Fri, 13 Jul 2012)*
933 Third release candidate for 2.6. The following changes were done from
936 - Fixed ``UpgradeConfig`` w.r.t. to disk parameters on disk objects.
937 - Fixed an inconsistency in the LUXI protocol with the provided
938 arguments (NOT backwards compatible)
939 - Fixed a bug with node groups ipolicy where ``min`` was greater than
940 the cluster ``std`` value
941 - Implemented a new ``gnt-node list-drbd`` call to list DRBD minors for
942 easier instance debugging on nodes (requires ``hconfd`` to work)
948 *(Released Tue, 03 Jul 2012)*
950 Second release candidate for 2.6. The following changes were done from
953 - Fixed ``gnt-cluster verify`` regarding ``master-ip-script`` on non
955 - Fixed a RAPI regression on missing beparams/memory
956 - Fixed redistribution of files on offline nodes
957 - Added possibility to run activate-disks even though secondaries are
958 offline. With this change it relaxes also the strictness on some other
959 commands which use activate disks internally:
960 * ``gnt-instance start|reboot|rename|backup|export``
961 - Made it possible to remove safely an instance if its secondaries are
963 - Made it possible to reinstall even though secondaries are offline
969 *(Released Mon, 25 Jun 2012)*
971 First release candidate for 2.6. The following changes were done from
974 - Fixed bugs with disk parameters and ``rbd`` templates as well as
976 - Made ``gnt-instance modify`` more consistent regarding new NIC/Disk
977 behaviour. It supports now the modify operation
978 - ``hcheck`` implemented to analyze cluster health and possibility of
979 improving health by rebalance
980 - ``hbal`` has been improved in dealing with split instances
986 *(Released Mon, 11 Jun 2012)*
988 Second beta release of 2.6. The following changes were done from beta2
991 - Fixed ``daemon-util`` with non-root user models
992 - Fixed creation of plain instances with ``--no-wait-for-sync``
993 - Fix wrong iv_names when running ``cfgupgrade``
994 - Export more information in RAPI group queries
995 - Fixed bug when changing instance network interfaces
996 - Extended burnin to do NIC changes
997 - query: Added ``<``, ``>``, ``<=``, ``>=`` comparison operators
998 - Changed default for DRBD barriers
999 - Fixed DRBD error reporting for syncer rate
1000 - Verify the options on disk parameters
1002 And of course various fixes to documentation and improved unittests and
1009 *(Released Wed, 23 May 2012)*
1011 First beta release of 2.6. The following changes were done from beta1 to
1014 - integrated patch for distributions without ``start-stop-daemon``
1015 - adapted example init.d script to work on Fedora
1016 - fixed log handling in Haskell daemons
1017 - adapted checks in the watcher for pycurl linked against libnss
1018 - add partial support for ``xl`` instead of ``xm`` for Xen
1019 - fixed a type issue in cluster verification
1020 - fixed ssconf handling in the Haskell code (was breaking confd in IPv6
1023 Plus integrated fixes from the 2.5 branch:
1025 - fixed ``kvm-ifup`` to use ``/bin/bash``
1026 - fixed parallel build failures
1027 - KVM live migration when using a custom keymap
1033 *(Released Tue, 24 Jul 2012)*
1035 A small bugfix release, with no new features:
1037 - fixed bash-isms in kvm-ifup, for compatibility with systems which use a
1038 different default shell (e.g. Debian, Ubuntu)
1039 - fixed KVM startup and live migration with a custom keymap (fixes Issue
1040 243 and Debian bug #650664)
1041 - fixed compatibility with KVM versions that don't support multiple boot
1042 devices (fixes Issue 230 and Debian bug #624256)
1044 Additionally, a few fixes were done to the build system (fixed parallel
1045 build failures) and to the unittests (fixed race condition in test for
1046 FileID functions, and the default enable/disable mode for QA test is now
1053 *(Released Fri, 11 May 2012)*
1055 A small bugfix release.
1057 The main issues solved are on the topic of compatibility with newer LVM
1060 - fixed parsing of ``lv_attr`` field
1061 - adapted to new ``vgreduce --removemissing`` behaviour where sometimes
1062 the ``--force`` flag is needed
1064 Also on the topic of compatibility, ``tools/lvmstrap`` has been changed
1065 to accept kernel 3.x too (was hardcoded to 2.6.*).
1067 A regression present in 2.5.0 that broke handling (in the gnt-* scripts)
1068 of hook results and that also made display of other errors suboptimal
1069 was fixed; the code behaves now like 2.4 and earlier.
1071 Another change in 2.5, the cleanup of the OS scripts environment, is too
1072 aggressive: it removed even the ``PATH`` variable, which requires the OS
1073 scripts to *always* need to export it. Since this is a bit too strict,
1074 we now export a minimal PATH, the same that we export for hooks.
1076 The fix for issue 201 (Preserve bridge MTU in KVM ifup script) was
1077 integrated into this release.
1079 Finally, a few other miscellaneous changes were done (no new features,
1080 just small improvements):
1082 - Fix ``gnt-group --help`` display
1083 - Fix hardcoded Xen kernel path
1084 - Fix grow-disk handling of invalid units
1085 - Update synopsis for ``gnt-cluster repair-disk-sizes``
1086 - Accept both PUT and POST in noded (makes future upgrade to 2.6 easier)
1092 *(Released Thu, 12 Apr 2012)*
1094 Incompatible/important changes and bugfixes
1095 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1097 - The default of the ``/2/instances/[instance_name]/rename`` RAPI
1098 resource's ``ip_check`` parameter changed from ``True`` to ``False``
1099 to match the underlying LUXI interface.
1100 - The ``/2/nodes/[node_name]/evacuate`` RAPI resource was changed to use
1101 body parameters, see :doc:`RAPI documentation <rapi>`. The server does
1102 not maintain backwards-compatibility as the underlying operation
1103 changed in an incompatible way. The RAPI client can talk to old
1104 servers, but it needs to be told so as the return value changed.
1105 - When creating file-based instances via RAPI, the ``file_driver``
1106 parameter no longer defaults to ``loop`` and must be specified.
1107 - The deprecated ``bridge`` NIC parameter is no longer supported. Use
1109 - Support for the undocumented and deprecated RAPI instance creation
1110 request format version 0 has been dropped. Use version 1, supported
1111 since Ganeti 2.1.3 and :doc:`documented <rapi>`, instead.
1112 - Pyparsing 1.4.6 or above is required, see :doc:`installation
1113 documentation <install>`.
1114 - The "cluster-verify" hooks are now executed per group by the
1115 ``OP_CLUSTER_VERIFY_GROUP`` opcode. This maintains the same behavior
1116 if you just run ``gnt-cluster verify``, which generates one opcode per
1118 - The environment as passed to the OS scripts is cleared, and thus no
1119 environment variables defined in the node daemon's environment will be
1120 inherited by the scripts.
1121 - The :doc:`iallocator <iallocator>` mode ``multi-evacuate`` has been
1123 - :doc:`New iallocator modes <design-multi-reloc>` have been added to
1124 support operations involving multiple node groups.
1125 - Offline nodes are ignored when failing over an instance.
1126 - Support for KVM version 1.0, which changed the version reporting format
1128 - TCP/IP ports used by DRBD disks are returned to a pool upon instance
1130 - ``Makefile`` is now compatible with Automake 1.11.2
1131 - Includes all bugfixes made in the 2.4 series
1136 - The ganeti-htools project has been merged into the ganeti-core source
1137 tree and will be built as part of Ganeti (see :doc:`install-quick`).
1138 - Implemented support for :doc:`shared storage <design-shared-storage>`.
1139 - Add support for disks larger than 2 TB in ``lvmstrap`` by supporting
1140 GPT-style partition tables (requires `parted
1141 <http://www.gnu.org/s/parted/>`_).
1142 - Added support for floppy drive and 2nd CD-ROM drive in KVM hypervisor.
1143 - Allowed adding tags on instance creation.
1144 - Export instance tags to hooks (``INSTANCE_TAGS``, see :doc:`hooks`)
1145 - Allow instances to be started in a paused state, enabling the user to
1146 see the complete console output on boot using the console.
1147 - Added new hypervisor flag to control default reboot behaviour
1148 (``reboot_behavior``).
1149 - Added support for KVM keymaps (hypervisor parameter ``keymap``).
1150 - Improved out-of-band management support:
1152 - Added ``gnt-node health`` command reporting the health status of
1154 - Added ``gnt-node power`` command to manage power status of nodes.
1155 - Added command for emergency power-off (EPO), ``gnt-cluster epo``.
1157 - Instance migration can fall back to failover if instance is not
1159 - Filters can be used when listing nodes, instances, groups and locks;
1160 see :manpage:`ganeti(7)` manpage.
1161 - Added post-execution status as variables to :doc:`hooks <hooks>`
1163 - Instance tags are exported/imported together with the instance.
1164 - When given an explicit job ID, ``gnt-job info`` will work for archived
1166 - Jobs can define dependencies on other jobs (not yet supported via
1167 RAPI or command line, but used by internal commands and usable via
1170 - Lock monitor (``gnt-debug locks``) shows jobs waiting for
1173 - Instance failover is now available as a RAPI resource
1174 (``/2/instances/[instance_name]/failover``).
1175 - ``gnt-instance info`` defaults to static information if primary node
1177 - Opcodes have a new ``comment`` attribute.
1178 - Added basic SPICE support to KVM hypervisor.
1179 - ``tools/ganeti-listrunner`` allows passing of arguments to executable.
1181 Node group improvements
1182 ~~~~~~~~~~~~~~~~~~~~~~~
1184 - ``gnt-cluster verify`` has been modified to check groups separately,
1185 thereby improving performance.
1186 - Node group support has been added to ``gnt-cluster verify-disks``,
1187 which now operates per node group.
1188 - Watcher has been changed to work better with node groups.
1190 - One process and state file per node group.
1191 - Slow watcher in one group doesn't block other group's watcher.
1193 - Added new command, ``gnt-group evacuate``, to move all instances in a
1194 node group to other groups.
1195 - Added ``gnt-instance change-group`` to move an instance to another
1197 - ``gnt-cluster command`` and ``gnt-cluster copyfile`` now support
1198 per-group operations.
1199 - Node groups can be tagged.
1200 - Some operations switch from an exclusive to a shared lock as soon as
1202 - Instance's primary and secondary nodes' groups are now available as
1203 query fields (``pnode.group``, ``pnode.group.uuid``, ``snodes.group``
1204 and ``snodes.group.uuid``).
1209 - Numerous updates to documentation and manpages.
1211 - :doc:`RAPI <rapi>` documentation now has detailed parameter
1213 - Some opcode/job results are now also documented, see :doc:`RAPI
1216 - A lockset's internal lock is now also visible in lock monitor.
1217 - Log messages from job queue workers now contain information about the
1218 opcode they're processing.
1219 - ``gnt-instance console`` no longer requires the instance lock.
1220 - A short delay when waiting for job changes reduces the number of LUXI
1221 requests significantly.
1222 - DRBD metadata volumes are overwritten with zeros during disk creation.
1223 - Out-of-band commands no longer acquire the cluster lock in exclusive
1225 - ``devel/upload`` now uses correct permissions for directories.
1231 *(Released Fri, 23 Mar 2012)*
1233 This was the sixth release candidate of the 2.5 series.
1239 *(Released Mon, 9 Jan 2012)*
1241 This was the fifth release candidate of the 2.5 series.
1247 *(Released Thu, 27 Oct 2011)*
1249 This was the fourth release candidate of the 2.5 series.
1255 *(Released Wed, 26 Oct 2011)*
1257 This was the third release candidate of the 2.5 series.
1263 *(Released Tue, 18 Oct 2011)*
1265 This was the second release candidate of the 2.5 series.
1271 *(Released Tue, 4 Oct 2011)*
1273 This was the first release candidate of the 2.5 series.
1279 *(Released Wed, 31 Aug 2011)*
1281 This was the third beta release of the 2.5 series.
1287 *(Released Mon, 22 Aug 2011)*
1289 This was the second beta release of the 2.5 series.
1295 *(Released Fri, 12 Aug 2011)*
1297 This was the first beta release of the 2.5 series.
1303 *(Released Thu, 27 Oct 2011)*
1305 - Fixed bug when parsing command line parameter values ending in
1307 - Fixed assertion error after unclean master shutdown
1308 - Disable HTTP client pool for RPC, significantly reducing memory usage
1310 - Fixed queue archive creation with wrong permissions
1316 *(Released Tue, 23 Aug 2011)*
1320 - Fixed documentation for importing with ``--src-dir`` option
1321 - Fixed a bug in ``ensure-dirs`` with queue/archive permissions
1322 - Fixed a parsing issue with DRBD 8.3.11 in the Linux kernel
1328 *(Released Fri, 5 Aug 2011)*
1330 Many bug-fixes and a few small features:
1332 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
1333 issues when you tried to add an instance with two MAC addresses in one
1335 - KVM: fixed per-instance stored UID value
1336 - KVM: configure bridged NICs at migration start
1337 - KVM: Fix a bug where instance will not start with never KVM versions
1339 - Added OS search path to ``gnt-cluster info``
1340 - Fixed an issue with ``file_storage_dir`` where you were forced to
1341 provide an absolute path, but the documentation states it is a
1342 relative path, the documentation was right
1343 - Added a new parameter to instance stop/start called ``--no-remember``
1344 that will make the state change to not be remembered
1345 - Implemented ``no_remember`` at RAPI level
1346 - Improved the documentation
1347 - Node evacuation: don't call IAllocator if node is already empty
1348 - Fixed bug in DRBD8 replace disks on current nodes
1349 - Fixed bug in recreate-disks for DRBD instances
1350 - Moved assertion checking locks in ``gnt-instance replace-disks``
1351 causing it to abort with not owning the right locks for some situation
1352 - Job queue: Fixed potential race condition when cancelling queued jobs
1353 - Fixed off-by-one bug in job serial generation
1354 - ``gnt-node volumes``: Fix instance names
1355 - Fixed aliases in bash completion
1356 - Fixed a bug in reopening log files after being sent a SIGHUP
1357 - Added a flag to burnin to allow specifying VCPU count
1358 - Bugfixes to non-root Ganeti configuration
1364 *(Released Thu, 12 May 2011)*
1366 Many bug-fixes and a few new small features:
1368 - Fixed a bug related to log opening failures
1369 - Fixed a bug in instance listing with orphan instances
1370 - Fixed a bug which prevented resetting the cluster-level node parameter
1371 ``oob_program`` to the default
1372 - Many fixes related to the ``cluster-merge`` tool
1373 - Fixed a race condition in the lock monitor, which caused failures
1374 during (at least) creation of many instances in parallel
1375 - Improved output for gnt-job info
1376 - Removed the quiet flag on some ssh calls which prevented debugging
1378 - Improved the N+1 failure messages in cluster verify by actually
1379 showing the memory values (needed and available)
1380 - Increased lock attempt timeouts so that when executing long operations
1381 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
1382 too early and thus prevent the use of the 'fair' mechanism
1383 - Changed instance query data (``gnt-instance info``) to not acquire
1384 locks unless needed, thus allowing its use on locked instance if only
1385 static information is asked for
1386 - Improved behaviour with filesystems that do not support rename on an
1388 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
1389 kept locks on all nodes during the wipe, which is unneeded
1390 - Fixed ``gnt-watcher`` handling of errors during hooks execution
1391 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
1392 10GiB) which caused the wipe to fail right at the end in some cases
1393 - Fixed master IP activation when doing master failover with no-voting
1394 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
1395 the master node itself
1396 - Fixed potential data-loss in under disk full conditions, where Ganeti
1397 wouldn't check correctly the return code and would consider
1398 partially-written files 'correct'
1399 - Fixed bug related to multiple VGs and DRBD disk replacing
1400 - Added new disk parameter ``metavg`` that allows placement of the meta
1401 device for DRBD in a different volume group
1402 - Fixed error handling in the node daemon when the system libc doesn't
1403 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
1404 - Fixed lock release during replace-disks, which kept cluster-wide locks
1405 when doing disk replaces with an iallocator script
1406 - Added check for missing bridges in cluster verify
1407 - Handle EPIPE errors while writing to the terminal better, so that
1408 piping the output to e.g. ``less`` doesn't cause a backtrace
1409 - Fixed rare case where a ^C during Luxi calls could have been
1410 interpreted as server errors, instead of simply terminating
1411 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
1413 - Added a few more parameters to the KVM hypervisor, allowing a second
1414 CDROM, custom disk type for CDROMs and a floppy image
1415 - Removed redundant message in instance rename when the name is given
1417 - Added option to ``gnt-instance recreate-disks`` to allow creating the
1418 disks on new nodes, allowing recreation when the original instance
1419 nodes are completely gone
1420 - Added option when converting disk templates to DRBD to skip waiting
1421 for the resync, in order to make the instance available sooner
1422 - Added two new variables to the OS scripts environment (containing the
1424 - Made the root_path and optional parameter for the xen-pvm hypervisor,
1425 to allow use of ``pvgrub`` as bootloader
1426 - Changed the instance memory modifications to only check out-of-memory
1427 conditions on memory increases, and turned the secondary node warnings
1428 into errors (they can still be overridden via ``--force``)
1429 - Fixed the handling of a corner case when the Python installation gets
1430 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
1431 to execute a command that doesn't exist
1432 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
1433 node of the instance returned failures during instance shutdown; this
1434 adds the option ``--ignore-consistency`` to gnt-instance move
1436 And as usual, various improvements to the error messages, documentation
1443 *(Released Wed, 09 Mar 2011)*
1445 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
1446 the RAPI users file if run twice (even with ``--dry-run``).
1448 The release fixes that bug (nothing else changed).
1454 *(Released Mon, 07 Mar 2011)*
1456 Final 2.4.0 release. Just a few small fixes:
1458 - Fixed RAPI node evacuate
1459 - Fixed the kvm-ifup script
1460 - Fixed internal error handling for special job cases
1461 - Updated man page to specify the escaping feature for options
1467 *(Released Mon, 28 Feb 2011)*
1469 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
1470 that this feature wiped the disks of the wrong instance, leading to loss
1475 - Fixed title of query field containing instance name
1476 - Expanded the glossary in the documentation
1477 - Fixed one unittest (internal issue)
1483 *(Released Mon, 21 Feb 2011)*
1485 A number of bug fixes plus just a couple functionality changes.
1487 On the user-visible side, the ``gnt-* list`` command output has changed
1488 with respect to "special" field states. The current rc1 style of display
1489 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
1490 the default output mode special fields states are displayed as follows:
1492 - Offline resource: ``*``
1493 - Unavailable/not applicable: ``-``
1494 - Data missing (RPC failure): ``?``
1495 - Unknown field: ``??``
1497 Another user-visible change is the addition of ``--force-join`` to
1502 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
1503 - Fixed regression in RAPI/instance reinstall where all parameters were
1504 required (instead of optional)
1505 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
1506 - Fixed iallocator usage (offline nodes were not considered offline)
1507 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
1508 - Fixed hypervisor and OS parameter validation with respect to
1509 non-vm_capable nodes
1510 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
1512 - Fixed ``tools/listrunner`` with respect to agent-based usage
1518 *(Released Fri, 4 Feb 2011)*
1520 Many changes and fixes since the beta1 release. While there were some
1521 internal changes, the code has been mostly stabilised for the RC
1524 Note: the dumb allocator was removed in this release, as it was not kept
1525 up-to-date with the IAllocator protocol changes. It is recommended to
1526 use the ``hail`` command from the ganeti-htools package.
1528 Note: the 2.4 and up versions of Ganeti are not compatible with the
1529 0.2.x branch of ganeti-htools. You need to upgrade to
1530 ganeti-htools-0.3.0 (or later).
1532 Regressions fixed from 2.3
1533 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1535 - Fixed the ``gnt-cluster verify-disks`` command
1536 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
1538 - Fixed disk adoption breakage
1539 - Fixed wrong headers in instance listing for field aliases
1544 - Fixed corner case in KVM handling of NICs
1545 - Fixed many cases of wrong handling of non-vm_capable nodes
1546 - Fixed a bug where a missing instance symlink was not possible to
1547 recreate with any ``gnt-*`` command (now ``gnt-instance
1548 activate-disks`` does it)
1549 - Fixed the volume group name as reported by ``gnt-cluster
1551 - Increased timeouts for the import-export code, hopefully leading to
1552 fewer aborts due network or instance timeouts
1553 - Fixed bug in ``gnt-node list-storage``
1554 - Fixed bug where not all daemons were started on cluster
1555 initialisation, but only at the first watcher run
1556 - Fixed many bugs in the OOB implementation
1557 - Fixed watcher behaviour in presence of instances with offline
1559 - Fixed instance list output for instances running on the wrong node
1560 - a few fixes to the cluster-merge tool, but it still cannot merge
1561 multi-node groups (currently it is not recommended to use this tool)
1567 - Improved network configuration for the KVM hypervisor
1568 - Added e1000 as a supported NIC for Xen-HVM
1569 - Improved the lvmstrap tool to also be able to use partitions, as
1570 opposed to full disks
1571 - Improved speed of disk wiping (the cluster parameter
1572 ``prealloc_wipe_disks``, so that it has a low impact on the total time
1573 of instance creations
1574 - Added documentation for the OS parameters
1575 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
1576 hypervisor is not responding
1577 - Added display of blacklisted and hidden OS information in
1578 ``gnt-cluster info``
1579 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
1580 NIC and node parameters, which might create problems with currently
1581 invalid (but undetected) configuration files, but prevents validation
1582 failures when unrelated parameters are modified
1583 - Changed cluster initialisation to wait for the master daemon to become
1585 - Expanded the RAPI interface:
1587 - Added config redistribution resource
1588 - Added activation/deactivation of instance disks
1589 - Added export of console information
1591 - Implemented log file reopening on SIGHUP, which allows using
1592 logrotate(8) for the Ganeti log files
1593 - Added a basic OOB helper script as an example
1599 *(Released Fri, 14 Jan 2011)*
1604 - Fixed timezone issues when formatting timestamps
1605 - Added support for node groups, available via ``gnt-group`` and other
1607 - Added out-of-band framework and management, see :doc:`design
1608 document <design-oob>`
1609 - Removed support for roman numbers from ``gnt-node list`` and
1610 ``gnt-instance list``.
1611 - Allowed modification of master network interface via ``gnt-cluster
1612 modify --master-netdev``
1613 - Accept offline secondaries while shutting down instance disks
1614 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
1615 - Added support for multiple LVM volume groups
1616 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
1618 - Added commands to list available fields:
1620 - ``gnt-node list-fields``
1621 - ``gnt-group list-fields``
1622 - ``gnt-instance list-fields``
1624 - Updated documentation and man pages
1629 - Moved ``rapi_users`` file into separate directory, now named
1630 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
1632 - Added new tool for running commands on many machines,
1633 ``tools/ganeti-listrunner``
1634 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
1635 improving the ``gnt-instance console`` output
1636 - Allowed customisation of disk index separator at ``configure`` time
1637 - Export node group allocation policy to :doc:`iallocator <iallocator>`
1638 - Added support for non-partitioned md disks in ``lvmstrap``
1639 - Added script to gracefully power off KVM instances
1640 - Split ``utils`` module into smaller parts
1641 - Changed query operations to return more detailed information, e.g.
1642 whether an information is unavailable due to an offline node. To use
1643 this new functionality, the LUXI call ``Query`` must be used. Field
1644 information is now stored by the master daemon and can be retrieved
1645 using ``QueryFields``. Instances, nodes and groups can also be queried
1646 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
1647 exposed via RAPI). The following commands make use of this
1648 infrastructure change:
1650 - ``gnt-group list``
1651 - ``gnt-group list-fields``
1653 - ``gnt-node list-fields``
1654 - ``gnt-instance list``
1655 - ``gnt-instance list-fields``
1656 - ``gnt-debug locks``
1661 - New RAPI resources (see :doc:`rapi`):
1665 - ``/2/groups/[group_name]``
1666 - ``/2/groups/[group_name]/assign-nodes``
1667 - ``/2/groups/[group_name]/modify``
1668 - ``/2/groups/[group_name]/rename``
1669 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
1673 - Implemented ``no_install`` for instance creation
1674 - Implemented OS parameters for instance reinstallation, allowing
1675 use of special settings on reinstallation (e.g. for preserving data)
1680 - Added IPv6 support in import/export
1681 - Pause DRBD synchronization while wiping disks on instance creation
1682 - Updated unittests and QA scripts
1683 - Improved network parameters passed to KVM
1684 - Converted man pages from docbook to reStructuredText
1690 *(Released Mon, 20 Dec 2010)*
1692 Released version 2.3.1~rc1 without any changes.
1698 *(Released Wed, 1 Dec 2010)*
1700 - impexpd: Disable OpenSSL compression in socat if possible (backport
1701 from master, commit e90739d625b, see :doc:`installation guide
1702 <install-quick>` for details)
1703 - Changed unittest coverage report to exclude test scripts
1704 - Added script to check version format
1710 *(Released Wed, 1 Dec 2010)*
1712 Released version 2.3.0~rc1 without any changes.
1718 *(Released Fri, 19 Nov 2010)*
1720 A number of bugfixes and documentation updates:
1722 - Update ganeti-os-interface documentation
1723 - Fixed a bug related to duplicate MACs or similar items which should be
1725 - Fix breakage in OS state modify
1726 - Reinstall instance: disallow offline secondaries (fixes bug related to
1727 OS changing but reinstall failing)
1728 - plus all the other fixes between 2.2.1 and 2.2.2
1734 *(Released Tue, 2 Nov 2010)*
1736 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
1737 - Fixed master failover race with watcher
1738 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
1740 - Accept previously stopped instance for export with instance removal
1741 - Simplify and extend the environment variables for instance OS scripts
1742 - Added new node flags, ``master_capable`` and ``vm_capable``
1743 - Added optional instance disk wiping prior during allocation. This is a
1744 cluster-wide option and can be set/modified using
1745 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
1746 - Added IPv6 support, see :doc:`design document <design-2.3>` and
1747 :doc:`install-quick`
1748 - Added a new watcher option (``--ignore-pause``)
1749 - Added option to ignore offline node on instance start/stop
1750 (``--ignore-offline``)
1751 - Allow overriding OS parameters with ``gnt-instance reinstall``
1752 - Added ability to change node's secondary IP address using ``gnt-node
1754 - Implemented privilege separation for all daemons except
1755 ``ganeti-noded``, see ``configure`` options
1756 - Complain if an instance's disk is marked faulty in ``gnt-cluster
1758 - Implemented job priorities (see ``ganeti(7)`` manpage)
1759 - Ignore failures while shutting down instances during failover from
1761 - Exit daemon's bootstrap process only once daemon is ready
1762 - Export more information via ``LUInstanceQuery``/remote API
1763 - Improved documentation, QA and unittests
1764 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
1765 restart if the file was created or changed
1766 - Added LUXI protocol version sent with each request and response,
1767 allowing detection of server/client mismatches
1768 - Moved the Python scripts among gnt-* and ganeti-* into modules
1769 - Moved all code related to setting up SSH to an external script,
1771 - Infrastructure changes for node group support in future versions
1777 *(Released Fri, 19 Nov 2010)*
1779 A few small bugs fixed, and some improvements to the build system:
1781 - Fix documentation regarding conversion to drbd
1782 - Fix validation of parameters in cluster modify (``gnt-cluster modify
1784 - Fix error handling in node modify with multiple changes
1785 - Allow remote imports without checked names
1791 *(Released Tue, 19 Oct 2010)*
1793 - Disable SSL session ID cache in RPC client
1799 *(Released Thu, 14 Oct 2010)*
1801 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
1802 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
1804 - Fix problems with interaction between Curl and Python's HTTP server,
1805 resulting in increased speed in many RPC calls
1806 - Improve our release script to prevent breakage with older aclocal and
1813 *(Released Thu, 7 Oct 2010)*
1815 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
1816 value retrieved from master
1817 - Added support for blacklisted or hidden OS definitions
1818 - Added simple lock monitor (accessible via (``gnt-debug locks``)
1819 - Added support for -mem-path in KVM hypervisor abstraction layer
1820 - Allow overriding instance parameters in tool for inter-cluster
1821 instance moves (``tools/move-instance``)
1822 - Improved opcode summaries (e.g. in ``gnt-job list``)
1823 - Improve consistency of OS listing by sorting it
1824 - Documentation updates
1830 *(Released Fri, 8 Oct 2010)*
1832 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
1838 *(Released Mon, 4 Oct 2010)*
1840 - Fixed regression in ``gnt-instance rename``
1846 *(Released Wed, 22 Sep 2010)*
1848 - Fixed OS_VARIANT variable for OS scripts
1849 - Fixed cluster tag operations via RAPI
1850 - Made ``setup-ssh`` exit with non-zero code if an error occurred
1851 - Disabled RAPI CA checks in watcher
1857 *(Released Mon, 23 Aug 2010)*
1859 - Support DRBD versions of the format "a.b.c.d"
1861 - Re-introduce support for usage from multiple threads in RAPI client
1862 - Instance renames and modify via RAPI
1863 - Work around race condition between processing and archival in job
1865 - Mark opcodes following failed one as failed, too
1866 - Job field ``lock_status`` was removed due to difficulties making it
1867 work with the changed job queue in Ganeti 2.2; a better way to monitor
1868 locks is expected for a later 2.2.x release
1869 - Fixed dry-run behaviour with many commands
1870 - Support ``ssh-agent`` again when adding nodes
1871 - Many additional bugfixes
1877 *(Released Fri, 30 Jul 2010)*
1879 Important change: the internal RPC mechanism between Ganeti nodes has
1880 changed from using a home-grown http library (based on the Python base
1881 libraries) to use the PycURL library. This requires that PycURL is
1882 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
1883 against GnuTLS by default. cURL's support for GnuTLS had known issues
1884 before cURL 7.21.0 and we recommend using the latest cURL release or
1885 linking against OpenSSL. Most other distributions already link PycURL
1886 and cURL against OpenSSL. The command::
1888 python -c 'import pycurl; print pycurl.version'
1890 can be used to determine the libraries PycURL and cURL are linked
1893 Other significant changes:
1895 - Rewrote much of the internals of the job queue, in order to achieve
1896 better parallelism; this decouples job query operations from the job
1897 processing, and it should allow much nicer behaviour of the master
1898 daemon under load, and it also has uncovered some long-standing bugs
1899 related to the job serialisation (now fixed)
1900 - Added a default iallocator setting to the cluster parameters,
1901 eliminating the need to always pass nodes or an iallocator for
1902 operations that require selection of new node(s)
1903 - Added experimental support for the LXC virtualization method
1904 - Added support for OS parameters, which allows the installation of
1905 instances to pass parameter to OS scripts in order to customise the
1907 - Added a hypervisor parameter controlling the migration type (live or
1908 non-live), since hypervisors have various levels of reliability; this
1909 has renamed the 'live' parameter to 'mode'
1910 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
1911 logical volumes, meaning that cluster verify will ignore them and not
1912 flag their presence as errors
1913 - The watcher will now reset the error count for failed instances after
1914 8 hours, thus allowing self-healing if the problem that caused the
1915 instances to be down/fail to start has cleared in the meantime
1916 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
1917 check for, and warn, if the drbd module parameter ``usermode_helper``
1918 is not consistent with the cluster-wide setting; this is needed to
1919 make diagnose easier of failed drbd creations
1920 - Started adding base IPv6 support, but this is not yet
1921 enabled/available for use
1922 - Rename operations (cluster, instance) will now return the new name,
1923 which is especially useful if a short name was passed in
1924 - Added support for instance migration in RAPI
1925 - Added a tool to pre-configure nodes for the SSH setup, before joining
1926 them to the cluster; this will allow in the future a simplified model
1927 for node joining (but not yet fully enabled in 2.2); this needs the
1928 paramiko python library
1929 - Fixed handling of name-resolving errors
1930 - Fixed consistency of job results on the error path
1931 - Fixed master-failover race condition when executed multiple times in
1933 - Fixed many bugs related to the job queue (mostly introduced during the
1934 2.2 development cycle, so not all are impacting 2.1)
1935 - Fixed instance migration with missing disk symlinks
1936 - Fixed handling of unknown jobs in ``gnt-job archive``
1937 - And many other small fixes/improvements
1941 - Enhanced both the unittest and the QA coverage
1942 - Switched the opcode validation to a generic model, and extended the
1943 validation to all opcode parameters
1944 - Changed more parts of the code that write shell scripts to use the
1946 - Switched the master daemon to use the asyncore library for the Luxi
1953 *(Released Thu, 17 Jun 2010)*
1955 - Added tool (``move-instance``) and infrastructure to move instances
1956 between separate clusters (see :doc:`separate documentation
1957 <move-instance>` and :doc:`design document <design-2.2>`)
1958 - Added per-request RPC timeout
1959 - RAPI now requires a Content-Type header for requests with a body (e.g.
1960 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
1961 :rfc:`2616` (HTTP/1.1), section 7.2.1)
1962 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
1964 - Implemented initial support for running Ganeti daemons as separate
1965 users, see configure-time flags ``--with-user-prefix`` and
1966 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
1968 - Instances can be removed after export (``gnt-backup export
1969 --remove-instance``)
1970 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
1971 key (instead of 1024 bit)
1972 - Added new cluster configuration file for cluster domain secret
1973 - Import/export now use SSL instead of SSH
1974 - Added support for showing estimated time when exporting an instance,
1975 see the ``ganeti-os-interface(7)`` manpage and look for
1982 *(Released Tue, 16 Nov 2010)*
1984 Some more bugfixes. Unless critical bugs occur, this will be the last
1987 - Fix case of MAC special-values
1988 - Fix mac checker regex
1989 - backend: Fix typo causing "out of range" error
1990 - Add missing --units in gnt-instance list man page
1996 *(Released Tue, 24 Aug 2010)*
1999 - Don't ignore secondary node silently on non-mirrored disk templates
2001 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
2002 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
2003 - Properly document the usb_mouse parameter
2004 - Fix path in ganeti-rapi(8) (issue 116)
2005 - Adjust error message when the ganeti user's .ssh directory is
2007 - Add same-node-check when changing the disk template to drbd
2013 *(Released Fri, 16 Jul 2010)*
2016 - Add an option to only select some reboot types during qa/burnin.
2017 (on some hypervisors consequent reboots are not supported)
2018 - Fix infrequent race condition in master failover. Sometimes the old
2019 master ip address would be still detected as up for a short time
2020 after it was removed, causing failover to fail.
2021 - Decrease mlockall warnings when the ctypes module is missing. On
2022 Python 2.4 we support running even if no ctypes module is installed,
2023 but we were too verbose about this issue.
2024 - Fix building on old distributions, on which man doesn't have a
2026 - Fix RAPI not to ignore the MAC address on instance creation
2027 - Implement the old instance creation format in the RAPI client.
2033 *(Released Thu, 01 Jul 2010)*
2035 A small bugfix release:
2036 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
2037 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
2038 a non-existing option
2039 - Fix instance create: the --force-variant option was ignored
2040 - Improve pylint 0.21 compatibility and warnings with Python 2.6
2041 - Fix modify node storage with non-FQDN arguments
2042 - Fix RAPI client to authenticate under Python 2.6 when used
2043 for more than 5 requests needing authentication
2044 - Fix gnt-instance modify -t (storage) giving a wrong error message
2045 when converting a non-shutdown drbd instance to plain
2051 *(Released Fri, 18 Jun 2010)*
2053 A small bugfix release:
2055 - Fix live migration of KVM instances started with older Ganeti
2056 versions which had fewer hypervisor parameters
2057 - Fix gnt-instance grow-disk on down instances
2058 - Fix an error-reporting bug during instance migration
2059 - Better checking of the ``--net`` and ``--disk`` values, to avoid
2060 silently ignoring broken ones
2061 - Fix an RPC error reporting bug affecting, for example, RAPI client
2063 - Fix bug triggered by different API version os-es on different nodes
2064 - Fix a bug in instance startup with custom hvparams: OS level
2065 parameters would fail to be applied.
2066 - Fix the RAPI client under Python 2.6 (but more work is needed to
2067 make it work completely well with OpenSSL)
2068 - Fix handling of errors when resolving names from DNS
2074 *(Released Thu, 3 Jun 2010)*
2076 A medium sized development cycle. Some new features, and some
2077 fixes/small improvements/cleanups.
2079 Significant features
2080 ~~~~~~~~~~~~~~~~~~~~
2082 The node deamon now tries to mlock itself into memory, unless the
2083 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
2084 its logs, and falls back to console logging. This allows emergency
2085 features such as ``gnt-node powercycle`` to work even in the event of a
2086 broken node disk (tested offlining the disk hosting the node's
2087 filesystem and dropping its memory caches; don't try this at home)
2089 KVM: add vhost-net acceleration support. It can be tested with a new
2090 enough version of the kernel and of qemu-kvm.
2092 KVM: Add instance chrooting feature. If you use privilege dropping for
2093 your VMs you can also now force them to chroot to an empty directory,
2094 before starting the emulated guest.
2096 KVM: Add maximum migration bandwith and maximum downtime tweaking
2097 support (requires a new-enough version of qemu-kvm).
2099 Cluster verify will now warn if the master node doesn't have the master
2100 ip configured on it.
2102 Add a new (incompatible) instance creation request format to RAPI which
2103 supports all parameters (previously only a subset was supported, and it
2104 wasn't possible to extend the old format to accomodate all the new
2105 features. The old format is still supported, and a client can check for
2106 this feature, before using it, by checking for its presence in the
2107 ``features`` RAPI resource.
2109 Now with ancient latin support. Try it passing the ``--roman`` option to
2110 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
2111 (requires the python-roman module to be installed, in order to work).
2116 As usual many internal code refactorings, documentation updates, and
2119 - Lots of improvements and cleanups to the experimental Remote API
2120 (RAPI) client library.
2121 - A new unit test suite for the core daemon libraries.
2122 - A fix to creating missing directories makes sure the umask is not
2123 applied anymore. This enforces the same directory permissions
2125 - Better handling terminating daemons with ctrl+c (used when running
2126 them in debugging mode).
2127 - Fix a race condition in live migrating a KVM instance, when stat()
2128 on the old proc status file returned EINVAL, which is an unexpected
2130 - Fixed manpage checking with newer man and utf-8 charachters. But now
2131 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
2137 *(Released Fri, 7 May 2010)*
2139 Fix a bug which prevented untagged KVM instances from starting.
2145 *(Released Fri, 7 May 2010)*
2147 Another release with a long development cycle, during which many
2148 different features were added.
2150 Significant features
2151 ~~~~~~~~~~~~~~~~~~~~
2153 The KVM hypervisor now can run the individual instances as non-root, to
2154 reduce the impact of a VM being hijacked due to bugs in the
2155 hypervisor. It is possible to run all instances as a single (non-root)
2156 user, to manually specify a user for each instance, or to dynamically
2157 allocate a user out of a cluster-wide pool to each instance, with the
2158 guarantee that no two instances will run under the same user ID on any
2161 An experimental RAPI client library, that can be used standalone
2162 (without the other Ganeti libraries), is provided in the source tree as
2163 ``lib/rapi/client.py``. Note this client might change its interface in
2164 the future, as we iterate on its capabilities.
2166 A new command, ``gnt-cluster renew-crypto`` has been added to easily
2167 replace the cluster's certificates and crypto keys. This might help in
2168 case they have been compromised, or have simply expired.
2170 A new disk option for instance creation has been added that allows one
2171 to "adopt" currently existing logical volumes, with data
2172 preservation. This should allow easier migration to Ganeti from
2173 unmanaged (or managed via other software) instances.
2175 Another disk improvement is the possibility to convert between redundant
2176 (DRBD) and plain (LVM) disk configuration for an instance. This should
2177 allow better scalability (starting with one node and growing the
2178 cluster, or shrinking a two-node cluster to one node).
2180 A new feature that could help with automated node failovers has been
2181 implemented: if a node sees itself as offline (by querying the master
2182 candidates), it will try to shutdown (hard) all instances and any active
2183 DRBD devices. This reduces the risk of duplicate instances if an
2184 external script automatically failovers the instances on such nodes. To
2185 enable this, the cluster parameter ``maintain_node_health`` should be
2186 enabled; in the future this option (per the name) will enable other
2187 automatic maintenance features.
2189 Instance export/import now will reuse the original instance
2190 specifications for all parameters; that means exporting an instance,
2191 deleting it and the importing it back should give an almost identical
2192 instance. Note that the default import behaviour has changed from
2193 before, where it created only one NIC; now it recreates the original
2196 Cluster verify has added a few new checks: SSL certificates validity,
2197 /etc/hosts consistency across the cluster, etc.
2202 As usual, many internal changes were done, documentation fixes,
2205 - Fixed cluster initialization with disabled cluster storage (regression
2206 introduced in 2.1.1)
2207 - File-based storage supports growing the disks
2208 - Fixed behaviour of node role changes
2209 - Fixed cluster verify for some corner cases, plus a general rewrite of
2210 cluster verify to allow future extension with more checks
2211 - Fixed log spamming by watcher and node daemon (regression introduced
2213 - Fixed possible validation issues when changing the list of enabled
2215 - Fixed cleanup of /etc/hosts during node removal
2216 - Fixed RAPI response for invalid methods
2217 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
2218 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
2219 from ide disks, etc.)
2220 - Allow OS changes without re-installation (to record a changed OS
2221 outside of Ganeti, or to allow OS renames)
2222 - Allow instance creation without OS installation (useful for example if
2223 the OS will be installed manually, or restored from a backup not in
2225 - Implemented option to make cluster ``copyfile`` use the replication
2227 - Added list of enabled hypervisors to ssconf (possibly useful for
2229 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
2231 - A partial form of node re-IP is possible via node readd, which now
2232 allows changed node primary IP
2233 - Command line utilities now show an informational message if the job is
2235 - The logs of the master daemon now show the PID/UID/GID of the
2242 *(Released Fri, 12 Mar 2010)*
2244 During the 2.1.0 long release candidate cycle, a lot of improvements and
2245 changes have accumulated with were released later as 2.1.1.
2250 The node evacuate command (``gnt-node evacuate``) was significantly
2251 rewritten, and as such the IAllocator protocol was changed - a new
2252 request type has been added. This unfortunate change during a stable
2253 series is designed to improve performance of node evacuations; on
2254 clusters with more than about five nodes and which are well-balanced,
2255 evacuation should proceed in parallel for all instances of the node
2256 being evacuated. As such, any existing IAllocator scripts need to be
2257 updated, otherwise the above command will fail due to the unknown
2258 request. The provided "dumb" allocator has not been updated; but the
2259 ganeti-htools package supports the new protocol since version 0.2.4.
2261 Another important change is increased validation of node and instance
2262 names. This might create problems in special cases, if invalid host
2263 names are being used.
2265 Also, a new layer of hypervisor parameters has been added, that sits at
2266 OS level between the cluster defaults and the instance ones. This allows
2267 customisation of virtualization parameters depending on the installed
2268 OS. For example instances with OS 'X' may have a different KVM kernel
2269 (or any other parameter) than the cluster defaults. This is intended to
2270 help managing a multiple OSes on the same cluster, without manual
2271 modification of each instance's parameters.
2273 A tool for merging clusters, ``cluster-merge``, has been added in the
2274 tools sub-directory.
2279 - Improved the int/float conversions that should make the code more
2280 robust in face of errors from the node daemons
2281 - Fixed the remove node code in case of internal configuration errors
2282 - Fixed the node daemon behaviour in face of inconsistent queue
2283 directory (e.g. read-only file-system where we can't open the files
2285 - Fixed the behaviour of gnt-node modify for master candidate demotion;
2286 now it either aborts cleanly or, if given the new "auto_promote"
2287 parameter, will automatically promote other nodes as needed
2288 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
2289 completely prevent Ganeti from working
2290 - Fixed bug for instance export when not all disks were successfully
2292 - Fixed behaviour of node add when the new node is slow in starting up
2294 - Fixed handling of signals in the LUXI client, which should improve
2295 behaviour of command-line scripts
2296 - Added checks for invalid node/instance names in the configuration (now
2297 flagged during cluster verify)
2298 - Fixed watcher behaviour for disk activation errors
2299 - Fixed two potentially endless loops in http library, which led to the
2300 RAPI daemon hanging and consuming 100% CPU in some cases
2301 - Fixed bug in RAPI daemon related to hashed passwords
2302 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
2304 - Enhanced compatibility with non-Debian OSes, but not using absolute
2305 path in some commands and allowing customisation of the ssh
2306 configuration directory
2307 - Fixed possible future issue with new Python versions by abiding to the
2308 proper use of ``__slots__`` attribute on classes
2309 - Added checks that should prevent directory traversal attacks
2310 - Many documentation fixes based on feedback from users
2315 - Added an "early_release" more for instance replace disks and node
2316 evacuate, where we release locks earlier and thus allow higher
2317 parallelism within the cluster
2318 - Added watcher hooks, intended to allow the watcher to restart other
2319 daemons (e.g. from the ganeti-nbma project), but they can be used of
2320 course for any other purpose
2321 - Added a compile-time disable for DRBD barriers, to increase
2322 performance if the administrator trusts the power supply or the
2323 storage system to not lose writes
2324 - Added the option of using syslog for logging instead of, or in
2325 addition to, Ganeti's own log files
2326 - Removed boot restriction for paravirtual NICs for KVM, recent versions
2327 can indeed boot from a paravirtual NIC
2328 - Added a generic debug level for many operations; while this is not
2329 used widely yet, it allows one to pass the debug value all the way to
2331 - Enhanced the hooks environment for instance moves (failovers,
2332 migrations) where the primary/secondary nodes changed during the
2333 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
2334 - Enhanced data validations for many user-supplied values; one important
2335 item is the restrictions imposed on instance and node names, which
2336 might reject some (invalid) host names
2337 - Add a configure-time option to disable file-based storage, if it's not
2338 needed; this allows greater security separation between the master
2339 node and the other nodes from the point of view of the inter-node RPC
2341 - Added user notification in interactive tools if job is waiting in the
2342 job queue or trying to acquire locks
2343 - Added log messages when a job is waiting for locks
2344 - Added filtering by node tags in instance operations which admit
2345 multiple instances (start, stop, reboot, reinstall)
2346 - Added a new tool for cluster mergers, ``cluster-merge``
2347 - Parameters from command line which are of the form ``a=b,c=d`` can now
2348 use backslash escapes to pass in values which contain commas,
2349 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
2351 - For KVM, the instance name is the first parameter passed to KVM, so
2352 that it's more visible in the process list
2358 *(Released Tue, 2 Mar 2010)*
2360 Ganeti 2.1 brings many improvements with it. Major changes:
2362 - Added infrastructure to ease automated disk repairs
2363 - Added new daemon to export configuration data in a cheaper way than
2364 using the remote API
2365 - Instance NICs can now be routed instead of being associated with a
2367 - Improved job locking logic to reduce impact of jobs acquiring multiple
2368 locks waiting for other long-running jobs
2370 In-depth implementation details can be found in the Ganeti 2.1 design
2376 - Added chroot hypervisor
2377 - Added more options to xen-hvm hypervisor (``kernel_path`` and
2379 - Added more options to xen-pvm hypervisor (``use_bootloader``,
2380 ``bootloader_path`` and ``bootloader_args``)
2381 - Added the ``use_localtime`` option for the xen-hvm and kvm
2382 hypervisors, and the default value for this has changed to false (in
2383 2.0 xen-hvm always enabled it)
2384 - Added luxi call to submit multiple jobs in one go
2385 - Added cluster initialization option to not modify ``/etc/hosts``
2387 - Added network interface parameters
2388 - Added dry run mode to some LUs
2389 - Added RAPI resources:
2391 - ``/2/instances/[instance_name]/info``
2392 - ``/2/instances/[instance_name]/replace-disks``
2393 - ``/2/nodes/[node_name]/evacuate``
2394 - ``/2/nodes/[node_name]/migrate``
2395 - ``/2/nodes/[node_name]/role``
2396 - ``/2/nodes/[node_name]/storage``
2397 - ``/2/nodes/[node_name]/storage/modify``
2398 - ``/2/nodes/[node_name]/storage/repair``
2400 - Added OpCodes to evacuate or migrate all instances on a node
2401 - Added new command to list storage elements on nodes (``gnt-node
2402 list-storage``) and modify them (``gnt-node modify-storage``)
2403 - Added new ssconf files with master candidate IP address
2404 (``ssconf_master_candidates_ips``), node primary IP address
2405 (``ssconf_node_primary_ips``) and node secondary IP address
2406 (``ssconf_node_secondary_ips``)
2407 - Added ``ganeti-confd`` and a client library to query the Ganeti
2408 configuration via UDP
2409 - Added ability to run hooks after cluster initialization and before
2411 - Added automatic mode for disk replace (``gnt-instance replace-disks
2413 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
2414 after catastrophic data-loss
2415 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
2417 - Added ``gnt-instance move`` command to move instances
2418 - Added ``gnt-cluster watcher`` command to control watcher
2419 - Added ``gnt-node powercycle`` command to powercycle nodes
2420 - Added new job status field ``lock_status``
2421 - Added parseable error codes to cluster verification (``gnt-cluster
2422 verify --error-codes``) and made output less verbose (use
2423 ``--verbose`` to restore previous behaviour)
2424 - Added UUIDs to the main config entities (cluster, nodes, instances)
2425 - Added support for OS variants
2426 - Added support for hashed passwords in the Ganeti remote API users file
2428 - Added option to specify maximum timeout on instance shutdown
2429 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
2430 - Added new helper script to start and stop Ganeti daemons
2431 (``daemon-util``), with the intent to reduce the work necessary to
2432 adjust Ganeti for non-Debian distributions and to start/stop daemons
2434 - Added more unittests
2435 - Fixed critical bug in ganeti-masterd startup
2436 - Removed the configure-time ``kvm-migration-port`` parameter, this is
2437 now customisable at the cluster level for both the KVM and Xen
2438 hypervisors using the new ``migration_port`` parameter
2439 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
2440 reinstalling an instance
2441 - Allowed ``@`` in tag names
2442 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
2443 - Many documentation updates
2444 - Distribute hypervisor files on ``gnt-cluster redist-conf``
2445 - ``gnt-instance reinstall`` can now reinstall multiple instances
2446 - Updated many command line parameters
2447 - Introduced new OS API version 15
2448 - No longer support a default hypervisor
2449 - Treat virtual LVs as inexistent
2450 - Improved job locking logic to reduce lock contention
2451 - Match instance and node names case insensitively
2452 - Reimplemented bash completion script to be more complete
2459 *(Released Thu, 4 Feb 2010)*
2461 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
2462 - Fix a string formatting bug
2463 - Improve safety of the code in some error paths
2464 - Improve data validation in the master of values returned from nodes
2470 *(Released Thu, 17 Dec 2009)*
2472 - Fix security issue due to missing validation of iallocator names; this
2473 allows local and remote execution of arbitrary executables
2474 - Fix failure of gnt-node list during instance removal
2475 - Ship the RAPI documentation in the archive
2481 *(Released Wed, 30 Sep 2009)*
2483 - Fixed many wrong messages
2484 - Fixed a few bugs related to the locking library
2485 - Fixed MAC checking at instance creation time
2486 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
2487 - Fixed a few issues related to signal handling in both daemons and
2489 - Fixed the example startup script provided
2490 - Fixed insserv dependencies in the example startup script (patch from
2492 - Fixed handling of drained nodes in the iallocator framework
2493 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
2495 - Fixed error related to invalid job IDs in job polling
2496 - Fixed job/opcode persistence on unclean master shutdown
2497 - Fixed handling of partial job processing after unclean master
2499 - Fixed error reporting from LUs, previously all errors were converted
2500 into execution errors
2501 - Fixed error reporting from burnin
2502 - Decreased significantly the memory usage of the job queue
2503 - Optimised slightly multi-job submission
2504 - Optimised slightly opcode loading
2505 - Backported the multi-job submit framework from the development
2506 branch; multi-instance start and stop should be faster
2507 - Added script to clean archived jobs after 21 days; this will reduce
2508 the size of the queue directory
2509 - Added some extra checks in disk size tracking
2510 - Added an example ethers hook script
2511 - Added a cluster parameter that prevents Ganeti from modifying of
2513 - Added more node information to RAPI responses
2514 - Added a ``gnt-job watch`` command that allows following the ouput of a
2516 - Added a bind-address option to ganeti-rapi
2517 - Added more checks to the configuration verify
2518 - Enhanced the burnin script such that some operations can be retried
2520 - Converted instance reinstall to multi-instance model
2526 *(Released Fri, 7 Aug 2009)*
2528 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
2529 to allow using the pre-2.0.2 behaviour in activation, if any existing
2530 instances have mismatched disk sizes in the configuration
2531 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
2532 any configuration mismatches for disk sizes
2533 - Added ``gnt-master cluste-failover --no-voting`` to allow master
2534 failover to work on two-node clusters
2535 - Fixed the ``--net`` option of ``gnt-backup import``, which was
2537 - Fixed detection of OS script errors in ``gnt-backup export``
2538 - Fixed exit code of ``gnt-backup export``
2544 *(Released Fri, 17 Jul 2009)*
2546 - Added experimental support for stripped logical volumes; this should
2547 enhance performance but comes with a higher complexity in the block
2548 device handling; stripping is only enabled when passing
2549 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
2550 affected even in the non-stripped mode
2551 - Improved resiliency against transient failures at the end of DRBD
2552 resyncs, and in general of DRBD resync checks
2553 - Fixed a couple of issues with exports and snapshot errors
2554 - Fixed a couple of issues in instance listing
2555 - Added display of the disk size in ``gnt-instance info``
2556 - Fixed checking for valid OSes in instance creation
2557 - Fixed handling of the "vcpus" parameter in instance listing and in
2558 general of invalid parameters
2559 - Fixed http server library, and thus RAPI, to handle invalid
2560 username/password combinations correctly; this means that now they
2561 report unauthorized for queries too, not only for modifications,
2562 allowing earlier detect of configuration problems
2563 - Added a new "role" node list field, equivalent to the master/master
2564 candidate/drained/offline flags combinations
2565 - Fixed cluster modify and changes of candidate pool size
2566 - Fixed cluster verify error messages for wrong files on regular nodes
2567 - Fixed a couple of issues with node demotion from master candidate role
2568 - Fixed node readd issues
2569 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
2570 - Added a new ``--no-voting`` option for masterfailover to fix failover
2571 on two-nodes clusters when the former master node is unreachable
2572 - Added instance reinstall over RAPI
2578 *(Released Tue, 16 Jun 2009)*
2580 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
2581 allow re-adding the start in single-user option (regression from 1.2)
2582 - the watcher writes the instance status to a file, to allow monitoring
2583 to report the instance status (from the master) based on cached
2584 results of the watcher's queries; while this can get stale if the
2585 watcher is being locked due to other work on the cluster, this is
2586 still an improvement
2587 - the watcher now also restarts the node daemon and the rapi daemon if
2589 - fixed the watcher to handle full and drained queue cases
2590 - hooks export more instance data in the environment, which helps if
2591 hook scripts need to take action based on the instance's properties
2592 (no longer need to query back into ganeti)
2593 - instance failovers when the instance is stopped do not check for free
2594 RAM, so that failing over a stopped instance is possible in low memory
2596 - rapi uses queries for tags instead of jobs (for less job traffic), and
2597 for cluster tags it won't talk to masterd at all but read them from
2599 - a couple of error handling fixes in RAPI
2600 - drbd handling: improved the error handling of inconsistent disks after
2601 resync to reduce the frequency of "there are some degraded disks for
2602 this instance" messages
2603 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
2604 error handling path called a wrong function name)
2610 *(Released Wed, 27 May 2009)*
2612 - no changes from rc5
2618 *(Released Wed, 20 May 2009)*
2620 - fix a couple of bugs (validation, argument checks)
2621 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
2622 - some small improvements to RAPI and IAllocator
2623 - make watcher automatically start the master daemon if down
2629 *(Released Mon, 27 Apr 2009)*
2631 - change the OS list to not require locks; this helps with big clusters
2632 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
2633 volume group is broken
2634 - ``gnt-instance info``, without any arguments, doesn't run for all
2635 instances anymore; either pass ``--all`` or pass the desired
2636 instances; this helps against mistakes on big clusters where listing
2637 the information for all instances takes a long time
2638 - miscellaneous doc and man pages fixes
2644 *(Released Wed, 8 Apr 2009)*
2646 - Change the internal locking model of some ``gnt-node`` commands, in
2647 order to reduce contention (and blocking of master daemon) when
2648 batching many creation/reinstall jobs
2649 - Fixes to Xen soft reboot
2650 - No longer build documentation at build time, instead distribute it in
2651 the archive, in order to reduce the need for the whole docbook/rst
2658 *(Released Fri, 27 Mar 2009)*
2660 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
2661 - Fix watcher startup sequence, improves the behaviour of busy clusters
2662 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
2663 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
2664 bind address and other places
2665 - Some documentation fixes and updates
2671 *(Released Mon, 2 Mar 2009)*
2673 - More documentation updates, now all docs should be more-or-less
2675 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
2677 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
2679 - Fix an issue related to $libdir/run/ganeti and cluster creation
2685 *(Released Thu, 19 Feb 2009)*
2687 - Xen PVM and KVM have switched the default value for the instance root
2688 disk to the first partition on the first drive, instead of the whole
2689 drive; this means that the OS installation scripts must be changed
2691 - Man pages have been updated
2692 - RAPI has been switched by default to HTTPS, and the exported functions
2693 should all work correctly
2694 - RAPI v1 has been removed
2695 - Many improvements to the KVM hypervisor
2696 - Block device errors are now better reported
2697 - Many other bugfixes and small improvements
2703 *(Released Mon, 26 Jan 2009)*
2705 - Version 2 is a general rewrite of the code and therefore the
2706 differences are too many to list, see the design document for 2.0 in
2707 the ``doc/`` subdirectory for more details
2708 - In this beta version there is not yet a migration path from 1.2 (there
2709 will be one in the final 2.0 release)
2710 - A few significant changes are:
2712 - all commands are executed by a daemon (``ganeti-masterd``) and the
2713 various ``gnt-*`` commands are just front-ends to it
2714 - all the commands are entered into, and executed from a job queue,
2715 see the ``gnt-job(8)`` manpage
2716 - the RAPI daemon supports read-write operations, secured by basic
2717 HTTP authentication on top of HTTPS
2718 - DRBD version 0.7 support has been removed, DRBD 8 is the only
2719 supported version (when migrating from Ganeti 1.2 to 2.0, you need
2720 to migrate to DRBD 8 first while still running Ganeti 1.2)
2721 - DRBD devices are using statically allocated minor numbers, which
2722 will be assigned to existing instances during the migration process
2723 - there is support for both Xen PVM and Xen HVM instances running on
2725 - KVM virtualization is supported too
2726 - file-based storage has been implemented, which means that it is
2727 possible to run the cluster without LVM and DRBD storage, for
2728 example using a shared filesystem exported from shared storage (and
2729 still have live migration)
2735 *(Released Tue, 13 Jan 2009)*
2737 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
2738 - Reuse the old instance mac address by default on instance import, if
2739 the instance name is the same.
2740 - Handle situations in which the node info rpc returns incomplete
2742 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
2743 - Improved version of batcher:
2745 - state file support
2746 - instance mac address support
2747 - support for HVM clusters/instances
2749 - Add an option to show the number of cpu sockets and nodes in
2751 - Support OSes that handle more than one version of the OS api (but do
2752 not change the current API in any other way)
2753 - Fix ``gnt-node migrate``
2754 - ``gnt-debug`` man page
2755 - Fixes various more typos and small issues
2756 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
2762 *(Released Wed, 24 Sep 2008)*
2764 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
2765 type of disk exported to fully virtualized instances.
2766 - provide access to the serial console of HVM instances
2767 - instance auto_balance flag, set by default. If turned off it will
2768 avoid warnings on cluster verify if there is not enough memory to fail
2769 over an instance. in the future it will prevent automatically failing
2770 it over when we will support that.
2771 - batcher tool for instance creation, see ``tools/README.batcher``
2772 - ``gnt-instance reinstall --select-os`` to interactively select a new
2773 operating system when reinstalling an instance.
2774 - when changing the memory amount on instance modify a check has been
2775 added that the instance will be able to start. also warnings are
2776 emitted if the instance will not be able to fail over, if auto_balance
2778 - documentation fixes
2779 - sync fields between ``gnt-instance list/modify/add/import``
2780 - fix a race condition in drbd when the sync speed was set after giving
2781 the device a remote peer.
2787 *(Released Tue, 22 Jul 2008)*
2789 - note: the allowed size and number of tags per object were reduced
2790 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
2791 - fixed twisted 8.x compatibility
2792 - fixed ``gnt-instance replace-disks`` with iallocator
2793 - add TCP keepalives on twisted connections to detect restarted nodes
2794 - disk increase support, see ``gnt-instance grow-disk``
2795 - implement bulk node/instance query for RAPI
2796 - add tags in node/instance listing (optional)
2797 - experimental migration (and live migration) support, read the man page
2798 for ``gnt-instance migrate``
2799 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
2800 has some small improvements in handling its state file
2806 *(Released Fri, 13 Jun 2008)*
2808 - Experimental readonly, REST-based remote API implementation;
2809 automatically started on master node, TCP port 5080, if enabled by
2810 ``--enable-rapi`` parameter to configure script.
2811 - Instance allocator support. Add and import instance accept a
2812 ``--iallocator`` parameter, and call that instance allocator to decide
2813 which node to use for the instance. The iallocator document describes
2814 what's expected from an allocator script.
2815 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
2816 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
2817 node is lost there is still enough memory to fail over the instances
2819 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
2820 ``gnt-cluster verify``, to check for site-specific compliance. All the
2821 hooks will run, and their output, if any, will be displayed. Any
2822 failing hook will make the verification return an error value.
2823 - ``gnt-cluster verify`` now checks that its peers are reachable on the
2824 primary and secondary interfaces
2825 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
2826 that is still declared as part of the cluster and has failed.
2827 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
2828 specifying output fields, that just adds the chosen fields to the
2830 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
2831 export from the filesystem.
2832 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
2833 have been added. Using them you can enable/disable acpi and pae
2834 support, and specify a path for a cd image to be exported to the
2835 instance. These parameters as the name suggest only work on HVM
2837 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
2838 PAE support will be set to the previously hardcoded values, but the
2839 (previously hardcoded) path to the CDROM ISO image will be unset and
2840 if required, needs to be set manually with ``gnt-instance modify``
2842 - The address to which an instance's VNC console is bound is now
2843 selectable per-instance, rather than being cluster wide. Of course
2844 this only applies to instances controlled via VNC, so currently just
2845 applies to HVM clusters.
2851 *(Released Mon, 18 Feb 2008)*
2853 - more tweaks to the disk activation code (especially helpful for DRBD)
2854 - change the default ``gnt-instance list`` output format, now there is
2855 one combined status field (see the manpage for the exact values this
2857 - some more fixes for the mac export to hooks change
2858 - make Ganeti not break with DRBD 8.2.x (which changed the version
2859 format in ``/proc/drbd``) (issue 24)
2860 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
2861 template, allowing migration from DRBD0.7+MD to DRBD8
2867 *(Released Wed, 30 Jan 2008)*
2869 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
2871 - add command aliases infrastructure and a few aliases
2872 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
2873 man pages and the ``--help`` option of ``gnt-node
2874 list``/``gnt-instance list``
2875 - fix ``gnt-backup list`` with down nodes (issue 21)
2876 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
2877 - fix the dist archive and add a check for including svn/git files in
2879 - some developer-related changes: improve the burnin and the QA suite,
2880 add an upload script for testing during development
2886 *(Released Wed, 16 Jan 2008)*
2888 - experimental HVM support, read the install document, section
2889 "Initializing the cluster"
2890 - allow for the PVM hypervisor per-instance kernel and initrd paths
2891 - add a new command ``gnt-cluster verify-disks`` which uses a new
2892 algorithm to improve the reconnection of the DRBD pairs if the device
2893 on the secondary node has gone away
2894 - make logical volume code auto-activate LVs at disk activation time
2895 - slightly improve the speed of activating disks
2896 - allow specification of the MAC address at instance creation time, and
2897 changing it later via ``gnt-instance modify``
2898 - fix handling of external commands that generate lots of output on
2900 - update documentation with regard to minimum version of DRBD8 supported
2906 *(Released Tue, 4 Dec 2007)*
2908 - Log the ``xm create`` output to the node daemon log on failure (to
2909 help diagnosing the error)
2910 - In debug mode, log all external commands output if failed to the logs
2911 - Change parsing of lvm commands to ignore stderr
2917 *(Released Wed, 28 Nov 2007)*
2919 - Another round of updates to the DRBD 8 code to deal with more failures
2920 in the replace secondary node operation
2921 - Some more logging of failures in disk operations (lvm, drbd)
2922 - A few documentation updates
2929 *(Released Tue, 13 Nov 2007)*
2931 - Change configuration file format from Python's Pickle to JSON.
2932 Upgrading is possible using the cfgupgrade utility.
2933 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
2934 faster replace disks and is more stable (DRBD 8 has many improvements
2935 compared to DRBD 0.7)
2936 - Added command line tags support (see man pages for ``gnt-instance``,
2937 ``gnt-node``, ``gnt-cluster``)
2938 - Added instance rename support
2939 - Added multi-instance startup/shutdown
2940 - Added cluster rename support
2941 - Added ``gnt-node evacuate`` to simplify some node operations
2942 - Added instance reboot operation that can speedup reboot as compared to
2944 - Soften the requirement that hostnames are in FQDN format
2945 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
2947 - Removed dependency on debian's patched fping that uses the
2948 non-standard ``-S`` option
2949 - Now the OS definitions are searched for in multiple, configurable
2950 paths (easier for distros to package)
2951 - Some changes to the hooks infrastructure (especially the new
2952 post-configuration update hook)
2953 - Other small bugfixes
2955 .. vim: set textwidth=72 syntax=rst :