8 *(Released Wed, 13 Nov 2013)*
10 - fix bug, that kept nodes offline when readding
11 - when verifying DRBD versions, ignore unavailable nodes
12 - fix bug that made the console unavailable on kvm in split-user
14 - DRBD: ensure peers are UpToDate for dual-primary (inherited 2.8.2)
20 *(Released Tue, 5 Nov 2013)*
22 Incompatible/important changes
23 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
25 - hroller now also plans for capacity to move non-redundant instances off
26 any node to be rebooted; the old behavior of completely ignoring any
27 non-redundant instances can be restored by adding the --ignore-non-redundant
29 - The cluster option '--no-lvm-storage' was removed in favor of the new option
30 '--enabled-disk-templates'.
31 - On instance creation, disk templates no longer need to be specified
32 with '-t'. The default disk template will be taken from the list of
33 enabled disk templates.
34 - The monitoring daemon is now running as root, in order to be able to collect
35 information only available to root (such as the state of Xen instances).
36 - The ConfD client is now IPv6 compatible.
37 - File and shared file storage is no longer dis/enabled at configure time,
38 but using the option '--enabled-disk-templates' at cluster initialization and
40 - The default directories for file and shared file storage are not anymore
41 specified at configure time, but taken from the cluster's configuration.
42 They can be set at cluster initialization and modification with
43 '--file-storage-dir' and '--shared-file-storage-dir'.
44 - Cluster verification now includes stricter checks regarding the
45 default file and shared file storage directories. It now checks that
46 the directories are explicitely allowed in the 'file-storage-paths' file and
47 that the directories exist on all nodes.
48 - The list of allowed disk templates in the instance policy and the list
49 of cluster-wide enabled disk templates is now checked for consistency
50 on cluster or group modification. On cluster initialization, the ipolicy
51 disk templates are ensured to be a subset of the cluster-wide enabled
57 - DRBD 8.4 support. Depending on the installed DRBD version, Ganeti now uses
58 the correct command syntax. It is possible to use different DRBD versions
59 on different nodes as long as they are compatible to each other. This
60 enables rolling upgrades of DRBD with no downtime. As permanent operation
61 of different DRBD versions within a node group is discouraged,
62 ``gnt-cluster verify`` will emit a warning if it detects such a situation.
63 - New "inst-status-xen" data collector for the monitoring daemon, providing
64 information about the state of the xen instances on the nodes.
65 - New "lv" data collector for the monitoring daemon, collecting data about the
66 logical volumes on the nodes, and pairing them with the name of the instances
68 - New "diskstats" data collector, collecting the data from /proc/diskstats and
69 presenting them over the monitoring daemon interface.
70 - The ConfD client is now IPv6 compatible.
74 The following new dependencies have been added.
78 - ``python-mock`` (http://www.voidspace.org.uk/python/mock/) is now a required
79 for the unit tests (and only used for testing).
83 - ``hslogger`` (http://software.complete.org/hslogger) is now always
84 required, even if confd is not enabled.
89 - Correctly start/stop luxid during gnt-cluster master-failover (inherited
91 - Improved error messsages (inherited from stable-2.8)
97 *(Released Tue, 15 Oct 2013)*
99 The third release candidate in the 2.9 series. Since 2.9.0 rc2:
101 - in implicit configuration upgrade, match ipolicy with enabled disk templates
102 - improved harep documentation (inherited from stable-2.8)
108 *(Released Wed, 9 Oct 2013)*
110 The second release candidate in the 2.9 series. Since 2.9.0 rc1:
112 - Fix bug in cfgupgrade that led to failure when upgrading from 2.8 with
113 at least one DRBD instance.
114 - Fix bug in cfgupgrade that led to an invalid 2.8 configuration after
121 *(Released Tue, 1 Oct 2013)*
123 The first release candidate in the 2.9 series. Since 2.9.0 beta1:
126 - update of the documentation, in particular installation instructions
127 - merging of LD_* constants into DT_* constants
128 - python style changes to be compatible with newer versions of pylint
134 *(Released Thu, 29 Aug 2013)*
136 This was the first beta release of the 2.9 series. All important changes
137 are listed in the latest 2.9 entry.
143 *(Released Thu, 07 Nov 2013)*
145 - DRBD: ensure peers are UpToDate for dual-primary
146 - Improve error message for replace-disks
147 - More dependency checks at configure time
148 - Placate warnings on ganeti.outils_unittest.py
154 *(Released Thu, 17 Oct 2013)*
156 - Correctly start/stop luxid during gnt-cluster master-failover
157 - Don't attempt IPv6 ssh in case of IPv4 cluster (Issue 595)
158 - Fix path for the job queue serial file
159 - Improved harep man page
160 - Minor documentation improvements
166 *(Released Mon, 30 Sep 2013)*
168 Incompatible/important changes
169 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
171 - Instance policy can contain multiple instance specs, as described in
172 the “Constrained instance sizes” section of :doc:`Partitioned Ganeti
173 <design-partitioned>`. As a consequence, it's not possible to partially change
174 or override instance specs. Bounding specs (min and max) can be specified as a
175 whole using the new option ``--ipolicy-bounds-specs``, while standard
176 specs use the new option ``--ipolicy-std-specs``.
177 - The output of the info command of gnt-cluster, gnt-group, gnt-node,
178 gnt-instance is a valid YAML object.
179 - hail now honors network restrictions when allocating nodes. This led to an
180 update of the IAllocator protocol. See the IAllocator documentation for
182 - confd now only answers static configuration request over the network. luxid
183 was extracted, listens on the local LUXI socket and responds to live queries.
184 This allows finer grained permissions if using separate users.
189 - The :doc:`Remote API <rapi>` daemon now supports a command line flag
190 to always require authentication, ``--require-authentication``. It can
191 be specified in ``$sysconfdir/default/ganeti``.
192 - A new cluster attribute 'enabled_disk_templates' is introduced. It will
193 be used to manage the disk templates to be used by instances in the cluster.
194 Initially, it will be set to a list that includes plain, drbd, if they were
195 enabled by specifying a volume group name, and file and sharedfile, if those
196 were enabled at configure time. Additionally, it will include all disk
197 templates that are currently used by instances. The order of disk templates
198 will be based on Ganeti's history of supporting them. In the future, the
199 first entry of the list will be used as a default disk template on instance
201 - ``cfgupgrade`` now supports a ``--downgrade`` option to bring the
202 configuration back to the previous stable version.
203 - Disk templates in group ipolicy can be restored to the default value.
204 - Initial support for diskless instances and virtual clusters in QA.
205 - More QA and unit tests for instance policies.
206 - Every opcode now contains a reason trail (visible through ``gnt-job info``)
207 describing why the opcode itself was executed.
208 - The monitoring daemon is now available. It allows users to query the cluster
209 for obtaining information about the status of the system. The daemon is only
210 responsible for providing the information over the network: the actual data
211 gathering is performed by data collectors (currently, only the DRBD status
212 collector is available).
213 - In order to help developers work on Ganeti, a new script
214 (``devel/build_chroot``) is provided, for building a chroot that contains all
215 the required development libraries and tools for compiling Ganeti on a Debian
217 - A new tool, ``harep``, for performing self-repair and recreation of instances
218 in Ganeti has been added.
219 - Split queries are enabled for tags, network, exports, cluster info, groups,
221 - New command ``show-ispecs-cmd`` for ``gnt-cluster`` and ``gnt-group``.
222 It prints the command line to set the current policies, to ease
224 - Add the ``vnet_hdr`` HV parameter for KVM, to control whether the tap
225 devices for KVM virtio-net interfaces will get created with VNET_HDR
226 (IFF_VNET_HDR) support. If set to false, it disables offloading on the
227 virtio-net interfaces, which prevents host kernel tainting and log
228 flooding, when dealing with broken or malicious virtio-net drivers.
229 It's set to true by default.
230 - Instance failover now supports a ``--cleanup`` parameter for fixing previous
232 - Support 'viridian' parameter in Xen HVM
233 - Support DSA SSH keys in bootstrap
234 - To simplify the work of packaging frameworks that want to add the needed users
235 and groups in a split-user setup themselves, at build time three files in
236 ``doc/users`` will be generated. The ``groups`` files contains, one per line,
237 the groups to be generated, the ``users`` file contains, one per line, the
238 users to be generated, optionally followed by their primary group, where
239 important. The ``groupmemberships`` file contains, one per line, additional
240 user-group membership relations that need to be established. The syntax of
241 these files will remain stable in all future versions.
246 The following new dependencies have been added:
249 - The ``curl`` library is not optional anymore for compiling the Haskell code.
250 - ``snap-server`` library (if monitoring is enabled).
253 - The minimum Python version needed to run Ganeti is now 2.6.
254 - ``yaml`` library (only for running the QA).
258 - Perform proper cleanup on termination of Haskell daemons
259 - Fix corner-case in handling of remaining retry time
265 *(Released Tue, 17 Sep 2013)*
267 - To simplify the work of packaging frameworks that want to add the needed users
268 and groups in a split-user setup themselves, at build time three files in
269 ``doc/users`` will be generated. The ``groups`` files contains, one per line,
270 the groups to be generated, the ``users`` file contains, one per line, the
271 users to be generated, optionally followed by their primary group, where
272 important. The ``groupmemberships`` file contains, one per line, additional
273 user-group membership relations that need to be established. The syntax of
274 these files will remain stable in all future versions.
275 - Add a default to file-driver when unspecified over RAPI (Issue 571)
276 - Mark the DSA host pubkey as optional, and remove it during config downgrade
278 - Some documentation fixes
284 *(Released Tue, 27 Aug 2013)*
286 The second release candidate of the 2.8 series. Since 2.8.0. rc1:
288 - Support 'viridian' parameter in Xen HVM (Issue 233)
289 - Include VCS version in ``gnt-cluster version``
290 - Support DSA SSH keys in bootstrap (Issue 338)
291 - Fix batch creation of instances
292 - Use FQDN to check master node status (Issue 551)
293 - Make the DRBD collector more failure-resilient
299 *(Released Fri, 2 Aug 2013)*
301 The first release candidate of the 2.8 series. Since 2.8.0 beta1:
303 - Fix upgrading/downgrading from 2.7
304 - Increase maximum RAPI message size
305 - Documentation updates
306 - Split ``confd`` between ``luxid`` and ``confd``
307 - Merge 2.7 series up to the 2.7.1 release
308 - Allow the ``modify_etc_hosts`` option to be changed
309 - Add better debugging for ``luxid`` queries
310 - Expose bulk parameter for GetJobs in RAPI client
311 - Expose missing ``network`` fields in RAPI
312 - Add some ``cluster verify`` tests
313 - Some unittest fixes
314 - Fix a malfunction in ``hspace``'s tiered allocation
315 - Fix query compatibility between haskell and python implementations
316 - Add the ``vnet_hdr`` HV parameter for KVM
317 - Add ``--cleanup`` to instance failover
318 - Change the connected groups format in ``gnt-network info`` output; it
319 was previously displayed as a raw list by mistake. (Merged from 2.7)
325 *(Released Mon, 24 Jun 2013)*
327 This was the first beta release of the 2.8 series. All important changes
328 are listed in the latest 2.8 entry.
334 *(Released Thu, 26 Sep 2013)*
336 - Change the connected groups format in ``gnt-network info`` output; it
337 was previously displayed as a raw list by mistake
338 - Check disk template in right dict when copying
339 - Support multi-instance allocs without iallocator
340 - Fix some errors in the documentation
341 - Fix formatting of tuple in an error message
347 *(Released Thu, 25 Jul 2013)*
349 - Add logrotate functionality in daemon-util
350 - Add logrotate example file
351 - Add missing fields to network queries over rapi
352 - Fix network object timestamps
353 - Add support for querying network timestamps
354 - Fix a typo in the example crontab
355 - Fix a documentation typo
361 *(Released Thu, 04 Jul 2013)*
363 Incompatible/important changes
364 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
366 - Instance policies for disk size were documented to be on a per-disk
367 basis, but hail applied them to the sum of all disks. This has been
369 - ``hbal`` will now exit with status 0 if, during job execution over
370 LUXI, early exit has been requested and all jobs are successful;
371 before, exit status 1 was used, which cannot be differentiated from
373 - Compatibility with newer versions of rbd has been fixed
374 - ``gnt-instance batch-create`` has been changed to use the bulk create
375 opcode from Ganeti. This lead to incompatible changes in the format of
376 the JSON file. It's now not a custom dict anymore but a dict
377 compatible with the ``OpInstanceCreate`` opcode.
378 - Parent directories for file storage need to be listed in
379 ``$sysconfdir/ganeti/file-storage-paths`` now. ``cfgupgrade`` will
380 write the file automatically based on old configuration values, but it
381 can not distribute it across all nodes and the file contents should be
382 verified. Use ``gnt-cluster copyfile
383 $sysconfdir/ganeti/file-storage-paths`` once the cluster has been
384 upgraded. The reason for requiring this list of paths now is that
385 before it would have been possible to inject new paths via RPC,
386 allowing files to be created in arbitrary locations. The RPC protocol
387 is protected using SSL/X.509 certificates, but as a design principle
388 Ganeti does not permit arbitrary paths to be passed.
389 - The parsing of the variants file for OSes (see
390 :manpage:`ganeti-os-interface(7)`) has been slightly changed: now empty
391 lines and comment lines (starting with ``#``) are ignored for better
393 - The ``setup-ssh`` tool added in Ganeti 2.2 has been replaced and is no
394 longer available. ``gnt-node add`` now invokes a new tool on the
395 destination node, named ``prepare-node-join``, to configure the SSH
396 daemon. Paramiko is no longer necessary to configure nodes' SSH
397 daemons via ``gnt-node add``.
398 - Draining (``gnt-cluster queue drain``) and un-draining the job queue
399 (``gnt-cluster queue undrain``) now affects all nodes in a cluster and
400 the flag is not reset after a master failover.
401 - Python 2.4 has *not* been tested with this release. Using 2.6 or above
402 is recommended. 2.6 will be mandatory from the 2.8 series.
408 - New network management functionality to support automatic allocation
409 of IP addresses and managing of network parameters. See
410 :manpage:`gnt-network(8)` for more details.
411 - New external storage backend, to allow managing arbitrary storage
412 systems external to the cluster. See
413 :manpage:`ganeti-extstorage-interface(7)`.
414 - New ``exclusive-storage`` node parameter added, restricted to
415 nodegroup level. When it's set to true, physical disks are assigned in
416 an exclusive fashion to instances, as documented in :doc:`Partitioned
417 Ganeti <design-partitioned>`. Currently, only instances using the
418 ``plain`` disk template are supported.
419 - The KVM hypervisor has been updated with many new hypervisor
420 parameters, including a generic one for passing arbitrary command line
421 values. See a complete list in :manpage:`gnt-instance(8)`. It is now
422 compatible up to qemu 1.4.
423 - A new tool, called ``mon-collector``, is the stand-alone executor of
424 the data collectors for a monitoring system. As of this version, it
425 just includes the DRBD data collector, that can be executed by calling
426 ``mon-collector`` using the ``drbd`` parameter. See
427 :manpage:`mon-collector(7)`.
428 - A new user option, :pyeval:`rapi.RAPI_ACCESS_READ`, has been added
429 for RAPI users. It allows granting permissions to query for
430 information to a specific user without giving
431 :pyeval:`rapi.RAPI_ACCESS_WRITE` permissions.
432 - A new tool named ``node-cleanup`` has been added. It cleans remains of
433 a cluster from a machine by stopping all daemons, removing
434 certificates and ssconf files. Unless the ``--no-backup`` option is
435 given, copies of the certificates are made.
436 - Instance creations now support the use of opportunistic locking,
437 potentially speeding up the (parallel) creation of multiple instances.
438 This feature is currently only available via the :doc:`RAPI
439 <rapi>` interface and when an instance allocator is used. If the
440 ``opportunistic_locking`` parameter is set the opcode will try to
441 acquire as many locks as possible, but will not wait for any locks
442 held by other opcodes. If not enough resources can be found to
443 allocate the instance, the temporary error code
444 :pyeval:`errors.ECODE_TEMP_NORES` is returned. The operation can be
445 retried thereafter, with or without opportunistic locking.
446 - New experimental linux-ha resource scripts.
447 - Restricted-commands support: ganeti can now be asked (via command line
448 or rapi) to perform commands on a node. These are passed via ganeti
449 RPC rather than ssh. This functionality is restricted to commands
450 specified on the ``$sysconfdir/ganeti/restricted-commands`` for security
451 reasons. The file is not copied automatically.
457 - Diskless instances are now externally mirrored (Issue 237). This for
458 now has only been tested in conjunction with explicit target nodes for
460 - Queries not needing locks or RPC access to the node can now be
461 performed by the confd daemon, making them independent from jobs, and
462 thus faster to execute. This is selectable at configure time.
463 - The functionality for allocating multiple instances at once has been
464 overhauled and is now also available through :doc:`RAPI <rapi>`.
466 There are no significant changes from version 2.7.0~rc3.
472 *(Released Tue, 25 Jun 2013)*
474 - Fix permissions on the confd query socket (Issue 477)
475 - Fix permissions on the job archive dir (Issue 498)
476 - Fix handling of an internal exception in replace-disks (Issue 472)
477 - Fix gnt-node info handling of shortened names (Issue 497)
478 - Fix gnt-instance grow-disk when wiping is enabled
479 - Documentation improvements, and support for newer pandoc
480 - Fix hspace honoring ipolicy for disks (Issue 484)
481 - Improve handling of the ``kvm_extra`` HV parameter
487 *(Released Fri, 24 May 2013)*
489 - ``devel/upload`` now works when ``/var/run`` on the target nodes is a
491 - Disks added through ``gnt-instance modify`` or created through
492 ``gnt-instance recreate-disks`` are wiped, if the
493 ``prealloc_wipe_disks`` flag is set.
494 - If wiping newly created disks fails, the disks are removed. Also,
495 partial failures in creating disks through ``gnt-instance modify``
496 triggers a cleanup of the partially-created disks.
497 - Removing the master IP address doesn't fail if the address has been
499 - Fix ownership of the OS log dir
500 - Workaround missing SO_PEERCRED constant (Issue 191)
506 *(Released Fri, 3 May 2013)*
508 This was the first release candidate of the 2.7 series. Since beta3:
510 - Fix kvm compatibility with qemu 1.4 (Issue 389)
511 - Documentation updates (admin guide, upgrade notes, install
512 instructions) (Issue 372)
513 - Fix gnt-group list nodes and instances count (Issue 436)
514 - Fix compilation without non-mandatory libraries (Issue 441)
515 - Fix xen-hvm hypervisor forcing nics to type 'ioemu' (Issue 247)
516 - Make confd logging more verbose at INFO level (Issue 435)
517 - Improve "networks" documentation in :manpage:`gnt-instance(8)`
518 - Fix failure path for instance storage type conversion (Issue 229)
519 - Update htools text backend documentation
520 - Improve the renew-crypto section of :manpage:`gnt-cluster(8)`
521 - Disable inter-cluster instance move for file-based instances, because
522 it is dependant on instance export, which is not supported for
523 file-based instances. (Issue 414)
524 - Fix gnt-job crashes on non-ascii characters (Issue 427)
525 - Fix volume group checks on non-vm-capable nodes (Issue 432)
531 *(Released Mon, 22 Apr 2013)*
533 This was the third beta release of the 2.7 series. Since beta2:
535 - Fix hail to verify disk instance policies on a per-disk basis (Issue 418).
536 - Fix data loss on wrong usage of ``gnt-instance move``
537 - Properly export errors in confd-based job queries
538 - Add ``users-setup`` tool
539 - Fix iallocator protocol to report 0 as a disk size for diskless
540 instances. This avoids hail breaking when a diskless instance is
542 - Fix job queue directory permission problem that made confd job queries
543 fail. This requires running an ``ensure-dirs --full-run`` on upgrade
544 for access to archived jobs (Issue 406).
545 - Limit the sizes of networks supported by ``gnt-network`` to something
546 between a ``/16`` and a ``/30`` to prevent memory bloat and crashes.
547 - Fix bugs in instance disk template conversion
548 - Fix GHC 7 compatibility
549 - Fix ``burnin`` install path (Issue 426).
550 - Allow very small disk grows (Issue 347).
551 - Fix a ``ganeti-noded`` memory bloat introduced in 2.5, by making sure
552 that noded doesn't import masterd code (Issue 419).
553 - Make sure the default metavg at cluster init is the same as the vg, if
554 unspecified (Issue 358).
555 - Fix cleanup of partially created disks (part of Issue 416)
561 *(Released Tue, 2 Apr 2013)*
563 This was the second beta release of the 2.7 series. Since beta1:
565 - Networks no longer have a "type" slot, since this information was
566 unused in Ganeti: instead of it tags should be used.
567 - The rapi client now has a ``target_node`` option to MigrateInstance.
568 - Fix early exit return code for hbal (Issue 386).
569 - Fix ``gnt-instance migrate/failover -n`` (Issue 396).
570 - Fix ``rbd showmapped`` output parsing (Issue 312).
571 - Networks are now referenced indexed by UUID, rather than name. This
572 will require running cfgupgrade, from 2.7.0beta1, if networks are in
574 - The OS environment now includes network information.
575 - Deleting of a network is now disallowed if any instance nic is using
576 it, to prevent dangling references.
577 - External storage is now documented in man pages.
578 - The exclusive_storage flag can now only be set at nodegroup level.
579 - Hbal can now submit an explicit priority with its jobs.
580 - Many network related locking fixes.
581 - Bump up the required pylint version to 0.25.1.
582 - Fix the ``no_remember`` option in RAPI client.
583 - Many ipolicy related tests, qa, and fixes.
584 - Many documentation improvements and fixes.
585 - Fix building with ``--disable-file-storage``.
586 - Fix ``-q`` option in htools, which was broken if passed more than
588 - Some haskell/python interaction improvements and fixes.
589 - Fix iallocator in case of missing LVM storage.
590 - Fix confd config load in case of ``--no-lvm-storage``.
591 - The confd/query functionality is now mentioned in the security
598 *(Released Wed, 6 Feb 2013)*
600 This was the first beta release of the 2.7 series. All important changes
601 are listed in the latest 2.7 entry.
607 *(Released Fri, 21 Dec 2012)*
609 Important behaviour change: hbal won't rebalance anymore instances which
610 have the ``auto_balance`` attribute set to false. This was the intention
611 all along, but until now it only skipped those from the N+1 memory
612 reservation (DRBD-specific).
614 A significant number of bug fixes in this release:
616 - Fixed disk adoption interaction with ipolicy checks.
617 - Fixed networking issues when instances are started, stopped or
618 migrated, by forcing the tap device's MAC prefix to "fe" (issue 217).
619 - Fixed the warning in cluster verify for shared storage instances not
621 - Fixed removal of storage directory on shared file storage (issue 262).
622 - Fixed validation of LVM volume group name in OpClusterSetParams
623 (``gnt-cluster modify``) (issue 285).
624 - Fixed runtime memory increases (``gnt-instance modify -m``).
625 - Fixed live migration under Xen's ``xl`` mode.
626 - Fixed ``gnt-instance console`` with ``xl``.
627 - Fixed building with newer Haskell compiler/libraries.
628 - Fixed PID file writing in Haskell daemons (confd); this prevents
629 restart issues if confd was launched manually (outside of
630 ``daemon-util``) while another copy of it was running
631 - Fixed a type error when doing live migrations with KVM (issue 297) and
632 the error messages for failing migrations have been improved.
633 - Fixed opcode validation for the out-of-band commands (``gnt-node
635 - Fixed a type error when unsetting OS hypervisor parameters (issue
636 311); now it's possible to unset all OS-specific hypervisor
638 - Fixed the ``dry-run`` mode for many operations: verification of
639 results was over-zealous but didn't take into account the ``dry-run``
640 operation, resulting in "wrong" failures.
641 - Fixed bash completion in ``gnt-job list`` when the job queue has
642 hundreds of entries; especially with older ``bash`` versions, this
643 results in significant CPU usage.
645 And lastly, a few other improvements have been made:
647 - Added option to force master-failover without voting (issue 282).
648 - Clarified error message on lock conflict (issue 287).
649 - Logging of newly submitted jobs has been improved (issue 290).
650 - Hostname checks have been made uniform between instance rename and
652 - The ``--submit`` option is now supported by ``gnt-debug delay``.
653 - Shutting down the master daemon by sending SIGTERM now stops it from
654 processing jobs waiting for locks; instead, those jobs will be started
655 once again after the master daemon is started the next time (issue
657 - Support for Xen's ``xl`` program has been improved (besides the fixes
659 - Reduced logging noise in the Haskell confd daemon (only show one log
660 entry for each config reload, instead of two).
661 - Several man page updates and typo fixes.
667 *(Released Fri, 12 Oct 2012)*
669 A small bugfix release. Among the bugs fixed:
671 - Fixed double use of ``PRIORITY_OPT`` in ``gnt-node migrate``, that
672 made the command unusable.
673 - Commands that issue many jobs don't fail anymore just because some jobs
674 take so long that other jobs are archived.
675 - Failures during ``gnt-instance reinstall`` are reflected by the exit
677 - Issue 190 fixed. Check for DRBD in cluster verify is enabled only when
679 - When ``always_failover`` is set, ``--allow-failover`` is not required
680 in migrate commands anymore.
681 - ``bash_completion`` works even if extglob is disabled.
682 - Fixed bug with locks that made failover for RDB-based instances fail.
683 - Fixed bug in non-mirrored instance allocation that made Ganeti choose
684 a random node instead of one based on the allocator metric.
685 - Support for newer versions of pylint and pep8.
686 - Hail doesn't fail anymore when trying to add an instance of type
687 ``file``, ``sharedfile`` or ``rbd``.
688 - Added new Makefile target to rebuild the whole distribution, so that
689 all files are included.
695 *(Released Fri, 27 Jul 2012)*
698 .. attention:: The ``LUXI`` protocol has been made more consistent
699 regarding its handling of command arguments. This, however, leads to
700 incompatibility issues with previous versions. Please ensure that you
701 restart Ganeti daemons soon after the upgrade, otherwise most
702 ``LUXI`` calls (job submission, setting/resetting the drain flag,
703 pausing/resuming the watcher, cancelling and archiving jobs, querying
704 the cluster configuration) will fail.
713 The current ``admin_up`` field, which used to denote whether an instance
714 should be running or not, has been removed. Instead, ``admin_state`` is
715 introduced, with 3 possible values -- ``up``, ``down`` and ``offline``.
717 The rational behind this is that an instance being “down” can have
720 - it could be down during a reboot
721 - it could be temporarily be down for a reinstall
722 - or it could be down because it is deprecated and kept just for its
725 The previous Boolean state was making it difficult to do capacity
726 calculations: should Ganeti reserve memory for a down instance? Now, the
727 tri-state field makes it clear:
729 - in ``up`` and ``down`` state, all resources are reserved for the
730 instance, and it can be at any time brought up if it is down
731 - in ``offline`` state, only disk space is reserved for it, but not
734 The field can have an extra use: since the transition between ``up`` and
735 ``down`` and vice-versus is done via ``gnt-instance start/stop``, but
736 transition between ``offline`` and ``down`` is done via ``gnt-instance
737 modify``, it is possible to given different rights to users. For
738 example, owners of an instance could be allowed to start/stop it, but
739 not transition it out of the offline state.
741 Instance policies and specs
742 +++++++++++++++++++++++++++
744 In previous Ganeti versions, an instance creation request was not
745 limited on the minimum size and on the maximum size just by the cluster
746 resources. As such, any policy could be implemented only in third-party
747 clients (RAPI clients, or shell wrappers over ``gnt-*``
748 tools). Furthermore, calculating cluster capacity via ``hspace`` again
749 required external input with regards to instance sizes.
751 In order to improve these workflows and to allow for example better
752 per-node group differentiation, we introduced instance specs, which
755 - minimum instance disk size, disk count, memory size, cpu count
756 - maximum values for the above metrics
757 - and “standard” values (used in ``hspace`` to calculate the standard
760 The minimum/maximum values can be also customised at node-group level,
761 for example allowing more powerful hardware to support bigger instance
764 Beside the instance specs, there are a few other settings belonging to
765 the instance policy framework. It is possible now to customise, per
766 cluster and node-group:
768 - the list of allowed disk templates
769 - the maximum ratio of VCPUs per PCPUs (to control CPU oversubscription)
770 - the maximum ratio of instance to spindles (see below for more
771 information) for local storage
773 All these together should allow all tools that talk to Ganeti to know
774 what are the ranges of allowed values for instances and the
775 over-subscription that is allowed.
777 For the VCPU/PCPU ratio, we already have the VCPU configuration from the
778 instance configuration, and the physical CPU configuration from the
779 node. For the spindle ratios however, we didn't track before these
780 values, so new parameters have been added:
782 - a new node parameter ``spindle_count``, defaults to 1, customisable at
783 node group or node level
784 - at new backend parameter (for instances), ``spindle_use`` defaults to 1
786 Note that spindles in this context doesn't need to mean actual
787 mechanical hard-drives; it's just a relative number for both the node
788 I/O capacity and instance I/O consumption.
790 Instance migration behaviour
791 ++++++++++++++++++++++++++++
793 While live-migration is in general desirable over failover, it is
794 possible that for some workloads it is actually worse, due to the
795 variable time of the “suspend” phase during live migration.
797 To allow the tools to work consistently over such instances (without
798 having to hard-code instance names), a new backend parameter
799 ``always_failover`` has been added to control the migration/failover
800 behaviour. When set to True, all migration requests for an instance will
801 instead fall-back to failover.
803 Instance memory ballooning
804 ++++++++++++++++++++++++++
806 Initial support for memory ballooning has been added. The memory for an
807 instance is no longer fixed (backend parameter ``memory``), but instead
808 can vary between minimum and maximum values (backend parameters
809 ``minmem`` and ``maxmem``). Currently we only change an instance's
812 - live migrating or failing over and instance and the target node
813 doesn't have enough memory
814 - user requests changing the memory via ``gnt-instance modify
820 In order to control the use of specific CPUs by instance, support for
821 controlling CPU pinning has been added for the Xen, HVM and LXC
822 hypervisors. This is controlled by a new hypervisor parameter
823 ``cpu_mask``; details about possible values for this are in the
824 :manpage:`gnt-instance(8)`. Note that use of the most specific (precise
825 VCPU-to-CPU mapping) form will work well only when all nodes in your
826 cluster have the same amount of CPUs.
831 Another area in which Ganeti was not customisable were the parameters
832 used for storage configuration, e.g. how many stripes to use for LVM,
833 DRBD resync configuration, etc.
835 To improve this area, we've added disks parameters, which are
836 customisable at cluster and node group level, and which allow to
837 specify various parameters for disks (DRBD has the most parameters
838 currently), for example:
840 - DRBD resync algorithm and parameters (e.g. speed)
841 - the default VG for meta-data volumes for DRBD
842 - number of stripes for LVM (plain disk template)
845 These parameters can be modified via ``gnt-cluster modify -D …`` and
846 ``gnt-group modify -D …``, and are used at either instance creation (in
847 case of LVM stripes, for example) or at disk “activation” time
850 Rados block device support
851 ++++++++++++++++++++++++++
853 A Rados (http://ceph.com/wiki/Rbd) storage backend has been added,
854 denoted by the ``rbd`` disk template type. This is considered
855 experimental, feedback is welcome. For details on configuring it, see
856 the :doc:`install` document and the :manpage:`gnt-cluster(8)` man page.
861 The existing master IP functionality works well only in simple setups (a
862 single network shared by all nodes); however, if nodes belong to
863 different networks, then the ``/32`` setup and lack of routing
864 information is not enough.
866 To allow the master IP to function well in more complex cases, the
867 system was reworked as follows:
869 - a master IP netmask setting has been added
870 - the master IP activation/turn-down code was moved from the node daemon
872 - whether to run the Ganeti-supplied master IP script or a user-supplied
873 on is a ``gnt-cluster init`` setting
875 Details about the location of the standard and custom setup scripts are
876 in the man page :manpage:`gnt-cluster(8)`; for information about the
877 setup script protocol, look at the Ganeti-supplied script.
882 The `SPICE <http://www.linux-kvm.org/page/SPICE>`_ support has been
885 It is now possible to use TLS-protected connections, and when renewing
886 or changing the cluster certificates (via ``gnt-cluster renew-crypto``,
887 it is now possible to specify spice or spice CA certificates. Also, it
888 is possible to configure a password for SPICE sessions via the
889 hypervisor parameter ``spice_password_file``.
891 There are also new parameters to control the compression and streaming
892 options (e.g. ``spice_image_compression``, ``spice_streaming_video``,
893 etc.). For details, see the man page :manpage:`gnt-instance(8)` and look
894 for the spice parameters.
896 Lastly, it is now possible to see the SPICE connection information via
897 ``gnt-instance console``.
902 A new tool (``tools/ovfconverter``) has been added that supports
903 conversion between Ganeti and the `Open Virtualization Format
904 <http://en.wikipedia.org/wiki/Open_Virtualization_Format>`_ (both to and
907 This relies on the ``qemu-img`` tool to convert the disk formats, so the
908 actual compatibility with other virtualization solutions depends on it.
913 The configuration query daemon (``ganeti-confd``) is now optional, and
914 has been rewritten in Haskell; whether to use the daemon at all, use the
915 Python (default) or the Haskell version is selectable at configure time
916 via the ``--enable-confd`` parameter, which can take one of the
917 ``haskell``, ``python`` or ``no`` values. If not used, disabling the
918 daemon will result in a smaller footprint; for larger systems, we
919 welcome feedback on the Haskell version which might become the default
922 If you want to use ``gnt-node list-drbd`` you need to have the Haskell
923 daemon running. The Python version doesn't implement the new call.
926 User interface changes
927 ~~~~~~~~~~~~~~~~~~~~~~
929 We have replaced the ``--disks`` option of ``gnt-instance
930 replace-disks`` with a more flexible ``--disk`` option, which allows
931 adding and removing disks at arbitrary indices (Issue 188). Furthermore,
932 disk size and mode can be changed upon recreation (via ``gnt-instance
933 recreate-disks``, which accepts the same ``--disk`` option).
935 As many people are used to a ``show`` command, we have added that as an
936 alias to ``info`` on all ``gnt-*`` commands.
938 The ``gnt-instance grow-disk`` command has a new mode in which it can
939 accept the target size of the disk, instead of the delta; this can be
940 more safe since two runs in absolute mode will be idempotent, and
941 sometimes it's also easier to specify the desired size directly.
943 Also the handling of instances with regard to offline secondaries has
944 been improved. Instance operations should not fail because one of it's
945 secondary nodes is offline, even though it's safe to proceed.
947 A new command ``list-drbd`` has been added to the ``gnt-node`` script to
948 support debugging of DRBD issues on nodes. It provides a mapping of DRBD
949 minors to instance name.
954 RAPI coverage has improved, with (for example) new resources for
955 recreate-disks, node power-cycle, etc.
960 There is partial support for ``xl`` in the Xen hypervisor; feedback is
963 Python 2.7 is better supported, and after Ganeti 2.6 we will investigate
964 whether to still support Python 2.4 or move to Python 2.6 as minimum
967 Support for Fedora has been slightly improved; the provided example
968 init.d script should work better on it and the INSTALL file should
969 document the needed dependencies.
974 The deprecated ``QueryLocks`` LUXI request has been removed. Use
975 ``Query(what=QR_LOCK, ...)`` instead.
977 The LUXI requests :pyeval:`luxi.REQ_QUERY_JOBS`,
978 :pyeval:`luxi.REQ_QUERY_INSTANCES`, :pyeval:`luxi.REQ_QUERY_NODES`,
979 :pyeval:`luxi.REQ_QUERY_GROUPS`, :pyeval:`luxi.REQ_QUERY_EXPORTS` and
980 :pyeval:`luxi.REQ_QUERY_TAGS` are deprecated and will be removed in a
981 future version. :pyeval:`luxi.REQ_QUERY` should be used instead.
983 RAPI client: ``CertificateError`` now derives from
984 ``GanetiApiError``. This should make it more easy to handle Ganeti
987 Deprecation warnings due to PyCrypto/paramiko import in
988 ``tools/setup-ssh`` have been silenced, as usually they are safe; please
989 make sure to run an up-to-date paramiko version, if you use this tool.
991 The QA scripts now depend on Python 2.5 or above (the main code base
992 still works with Python 2.4).
994 The configuration file (``config.data``) is now written without
995 indentation for performance reasons; if you want to edit it, it can be
996 re-formatted via ``tools/fmtjson``.
998 A number of bugs has been fixed in the cluster merge tool.
1000 ``x509`` certification verification (used in import-export) has been
1001 changed to allow the same clock skew as permitted by the cluster
1002 verification. This will remove some rare but hard to diagnose errors in
1009 *(Released Thu, 19 Jul 2012)*
1011 Very few changes from rc4 to the final release, only bugfixes:
1013 - integrated fixes from release 2.5.2 (fix general boot flag for KVM
1014 instance, fix CDROM booting for KVM instances)
1015 - fixed node group modification of node parameters
1016 - fixed issue in LUClusterVerifyGroup with multi-group clusters
1017 - fixed generation of bash completion to ensure a stable ordering
1024 *(Released Fri, 13 Jul 2012)*
1026 Third release candidate for 2.6. The following changes were done from
1029 - Fixed ``UpgradeConfig`` w.r.t. to disk parameters on disk objects.
1030 - Fixed an inconsistency in the LUXI protocol with the provided
1031 arguments (NOT backwards compatible)
1032 - Fixed a bug with node groups ipolicy where ``min`` was greater than
1033 the cluster ``std`` value
1034 - Implemented a new ``gnt-node list-drbd`` call to list DRBD minors for
1035 easier instance debugging on nodes (requires ``hconfd`` to work)
1041 *(Released Tue, 03 Jul 2012)*
1043 Second release candidate for 2.6. The following changes were done from
1046 - Fixed ``gnt-cluster verify`` regarding ``master-ip-script`` on non
1048 - Fixed a RAPI regression on missing beparams/memory
1049 - Fixed redistribution of files on offline nodes
1050 - Added possibility to run activate-disks even though secondaries are
1051 offline. With this change it relaxes also the strictness on some other
1052 commands which use activate disks internally:
1053 * ``gnt-instance start|reboot|rename|backup|export``
1054 - Made it possible to remove safely an instance if its secondaries are
1056 - Made it possible to reinstall even though secondaries are offline
1062 *(Released Mon, 25 Jun 2012)*
1064 First release candidate for 2.6. The following changes were done from
1067 - Fixed bugs with disk parameters and ``rbd`` templates as well as
1069 - Made ``gnt-instance modify`` more consistent regarding new NIC/Disk
1070 behaviour. It supports now the modify operation
1071 - ``hcheck`` implemented to analyze cluster health and possibility of
1072 improving health by rebalance
1073 - ``hbal`` has been improved in dealing with split instances
1079 *(Released Mon, 11 Jun 2012)*
1081 Second beta release of 2.6. The following changes were done from beta2
1084 - Fixed ``daemon-util`` with non-root user models
1085 - Fixed creation of plain instances with ``--no-wait-for-sync``
1086 - Fix wrong iv_names when running ``cfgupgrade``
1087 - Export more information in RAPI group queries
1088 - Fixed bug when changing instance network interfaces
1089 - Extended burnin to do NIC changes
1090 - query: Added ``<``, ``>``, ``<=``, ``>=`` comparison operators
1091 - Changed default for DRBD barriers
1092 - Fixed DRBD error reporting for syncer rate
1093 - Verify the options on disk parameters
1095 And of course various fixes to documentation and improved unittests and
1102 *(Released Wed, 23 May 2012)*
1104 First beta release of 2.6. The following changes were done from beta1 to
1107 - integrated patch for distributions without ``start-stop-daemon``
1108 - adapted example init.d script to work on Fedora
1109 - fixed log handling in Haskell daemons
1110 - adapted checks in the watcher for pycurl linked against libnss
1111 - add partial support for ``xl`` instead of ``xm`` for Xen
1112 - fixed a type issue in cluster verification
1113 - fixed ssconf handling in the Haskell code (was breaking confd in IPv6
1116 Plus integrated fixes from the 2.5 branch:
1118 - fixed ``kvm-ifup`` to use ``/bin/bash``
1119 - fixed parallel build failures
1120 - KVM live migration when using a custom keymap
1126 *(Released Tue, 24 Jul 2012)*
1128 A small bugfix release, with no new features:
1130 - fixed bash-isms in kvm-ifup, for compatibility with systems which use a
1131 different default shell (e.g. Debian, Ubuntu)
1132 - fixed KVM startup and live migration with a custom keymap (fixes Issue
1133 243 and Debian bug #650664)
1134 - fixed compatibility with KVM versions that don't support multiple boot
1135 devices (fixes Issue 230 and Debian bug #624256)
1137 Additionally, a few fixes were done to the build system (fixed parallel
1138 build failures) and to the unittests (fixed race condition in test for
1139 FileID functions, and the default enable/disable mode for QA test is now
1146 *(Released Fri, 11 May 2012)*
1148 A small bugfix release.
1150 The main issues solved are on the topic of compatibility with newer LVM
1153 - fixed parsing of ``lv_attr`` field
1154 - adapted to new ``vgreduce --removemissing`` behaviour where sometimes
1155 the ``--force`` flag is needed
1157 Also on the topic of compatibility, ``tools/lvmstrap`` has been changed
1158 to accept kernel 3.x too (was hardcoded to 2.6.*).
1160 A regression present in 2.5.0 that broke handling (in the gnt-* scripts)
1161 of hook results and that also made display of other errors suboptimal
1162 was fixed; the code behaves now like 2.4 and earlier.
1164 Another change in 2.5, the cleanup of the OS scripts environment, is too
1165 aggressive: it removed even the ``PATH`` variable, which requires the OS
1166 scripts to *always* need to export it. Since this is a bit too strict,
1167 we now export a minimal PATH, the same that we export for hooks.
1169 The fix for issue 201 (Preserve bridge MTU in KVM ifup script) was
1170 integrated into this release.
1172 Finally, a few other miscellaneous changes were done (no new features,
1173 just small improvements):
1175 - Fix ``gnt-group --help`` display
1176 - Fix hardcoded Xen kernel path
1177 - Fix grow-disk handling of invalid units
1178 - Update synopsis for ``gnt-cluster repair-disk-sizes``
1179 - Accept both PUT and POST in noded (makes future upgrade to 2.6 easier)
1185 *(Released Thu, 12 Apr 2012)*
1187 Incompatible/important changes and bugfixes
1188 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1190 - The default of the ``/2/instances/[instance_name]/rename`` RAPI
1191 resource's ``ip_check`` parameter changed from ``True`` to ``False``
1192 to match the underlying LUXI interface.
1193 - The ``/2/nodes/[node_name]/evacuate`` RAPI resource was changed to use
1194 body parameters, see :doc:`RAPI documentation <rapi>`. The server does
1195 not maintain backwards-compatibility as the underlying operation
1196 changed in an incompatible way. The RAPI client can talk to old
1197 servers, but it needs to be told so as the return value changed.
1198 - When creating file-based instances via RAPI, the ``file_driver``
1199 parameter no longer defaults to ``loop`` and must be specified.
1200 - The deprecated ``bridge`` NIC parameter is no longer supported. Use
1202 - Support for the undocumented and deprecated RAPI instance creation
1203 request format version 0 has been dropped. Use version 1, supported
1204 since Ganeti 2.1.3 and :doc:`documented <rapi>`, instead.
1205 - Pyparsing 1.4.6 or above is required, see :doc:`installation
1206 documentation <install>`.
1207 - The "cluster-verify" hooks are now executed per group by the
1208 ``OP_CLUSTER_VERIFY_GROUP`` opcode. This maintains the same behavior
1209 if you just run ``gnt-cluster verify``, which generates one opcode per
1211 - The environment as passed to the OS scripts is cleared, and thus no
1212 environment variables defined in the node daemon's environment will be
1213 inherited by the scripts.
1214 - The :doc:`iallocator <iallocator>` mode ``multi-evacuate`` has been
1216 - :doc:`New iallocator modes <design-multi-reloc>` have been added to
1217 support operations involving multiple node groups.
1218 - Offline nodes are ignored when failing over an instance.
1219 - Support for KVM version 1.0, which changed the version reporting format
1221 - TCP/IP ports used by DRBD disks are returned to a pool upon instance
1223 - ``Makefile`` is now compatible with Automake 1.11.2
1224 - Includes all bugfixes made in the 2.4 series
1229 - The ganeti-htools project has been merged into the ganeti-core source
1230 tree and will be built as part of Ganeti (see :doc:`install-quick`).
1231 - Implemented support for :doc:`shared storage <design-shared-storage>`.
1232 - Add support for disks larger than 2 TB in ``lvmstrap`` by supporting
1233 GPT-style partition tables (requires `parted
1234 <http://www.gnu.org/s/parted/>`_).
1235 - Added support for floppy drive and 2nd CD-ROM drive in KVM hypervisor.
1236 - Allowed adding tags on instance creation.
1237 - Export instance tags to hooks (``INSTANCE_TAGS``, see :doc:`hooks`)
1238 - Allow instances to be started in a paused state, enabling the user to
1239 see the complete console output on boot using the console.
1240 - Added new hypervisor flag to control default reboot behaviour
1241 (``reboot_behavior``).
1242 - Added support for KVM keymaps (hypervisor parameter ``keymap``).
1243 - Improved out-of-band management support:
1245 - Added ``gnt-node health`` command reporting the health status of
1247 - Added ``gnt-node power`` command to manage power status of nodes.
1248 - Added command for emergency power-off (EPO), ``gnt-cluster epo``.
1250 - Instance migration can fall back to failover if instance is not
1252 - Filters can be used when listing nodes, instances, groups and locks;
1253 see :manpage:`ganeti(7)` manpage.
1254 - Added post-execution status as variables to :doc:`hooks <hooks>`
1256 - Instance tags are exported/imported together with the instance.
1257 - When given an explicit job ID, ``gnt-job info`` will work for archived
1259 - Jobs can define dependencies on other jobs (not yet supported via
1260 RAPI or command line, but used by internal commands and usable via
1263 - Lock monitor (``gnt-debug locks``) shows jobs waiting for
1266 - Instance failover is now available as a RAPI resource
1267 (``/2/instances/[instance_name]/failover``).
1268 - ``gnt-instance info`` defaults to static information if primary node
1270 - Opcodes have a new ``comment`` attribute.
1271 - Added basic SPICE support to KVM hypervisor.
1272 - ``tools/ganeti-listrunner`` allows passing of arguments to executable.
1274 Node group improvements
1275 ~~~~~~~~~~~~~~~~~~~~~~~
1277 - ``gnt-cluster verify`` has been modified to check groups separately,
1278 thereby improving performance.
1279 - Node group support has been added to ``gnt-cluster verify-disks``,
1280 which now operates per node group.
1281 - Watcher has been changed to work better with node groups.
1283 - One process and state file per node group.
1284 - Slow watcher in one group doesn't block other group's watcher.
1286 - Added new command, ``gnt-group evacuate``, to move all instances in a
1287 node group to other groups.
1288 - Added ``gnt-instance change-group`` to move an instance to another
1290 - ``gnt-cluster command`` and ``gnt-cluster copyfile`` now support
1291 per-group operations.
1292 - Node groups can be tagged.
1293 - Some operations switch from an exclusive to a shared lock as soon as
1295 - Instance's primary and secondary nodes' groups are now available as
1296 query fields (``pnode.group``, ``pnode.group.uuid``, ``snodes.group``
1297 and ``snodes.group.uuid``).
1302 - Numerous updates to documentation and manpages.
1304 - :doc:`RAPI <rapi>` documentation now has detailed parameter
1306 - Some opcode/job results are now also documented, see :doc:`RAPI
1309 - A lockset's internal lock is now also visible in lock monitor.
1310 - Log messages from job queue workers now contain information about the
1311 opcode they're processing.
1312 - ``gnt-instance console`` no longer requires the instance lock.
1313 - A short delay when waiting for job changes reduces the number of LUXI
1314 requests significantly.
1315 - DRBD metadata volumes are overwritten with zeros during disk creation.
1316 - Out-of-band commands no longer acquire the cluster lock in exclusive
1318 - ``devel/upload`` now uses correct permissions for directories.
1324 *(Released Fri, 23 Mar 2012)*
1326 This was the sixth release candidate of the 2.5 series.
1332 *(Released Mon, 9 Jan 2012)*
1334 This was the fifth release candidate of the 2.5 series.
1340 *(Released Thu, 27 Oct 2011)*
1342 This was the fourth release candidate of the 2.5 series.
1348 *(Released Wed, 26 Oct 2011)*
1350 This was the third release candidate of the 2.5 series.
1356 *(Released Tue, 18 Oct 2011)*
1358 This was the second release candidate of the 2.5 series.
1364 *(Released Tue, 4 Oct 2011)*
1366 This was the first release candidate of the 2.5 series.
1372 *(Released Wed, 31 Aug 2011)*
1374 This was the third beta release of the 2.5 series.
1380 *(Released Mon, 22 Aug 2011)*
1382 This was the second beta release of the 2.5 series.
1388 *(Released Fri, 12 Aug 2011)*
1390 This was the first beta release of the 2.5 series.
1396 *(Released Thu, 27 Oct 2011)*
1398 - Fixed bug when parsing command line parameter values ending in
1400 - Fixed assertion error after unclean master shutdown
1401 - Disable HTTP client pool for RPC, significantly reducing memory usage
1403 - Fixed queue archive creation with wrong permissions
1409 *(Released Tue, 23 Aug 2011)*
1413 - Fixed documentation for importing with ``--src-dir`` option
1414 - Fixed a bug in ``ensure-dirs`` with queue/archive permissions
1415 - Fixed a parsing issue with DRBD 8.3.11 in the Linux kernel
1421 *(Released Fri, 5 Aug 2011)*
1423 Many bug-fixes and a few small features:
1425 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
1426 issues when you tried to add an instance with two MAC addresses in one
1428 - KVM: fixed per-instance stored UID value
1429 - KVM: configure bridged NICs at migration start
1430 - KVM: Fix a bug where instance will not start with never KVM versions
1432 - Added OS search path to ``gnt-cluster info``
1433 - Fixed an issue with ``file_storage_dir`` where you were forced to
1434 provide an absolute path, but the documentation states it is a
1435 relative path, the documentation was right
1436 - Added a new parameter to instance stop/start called ``--no-remember``
1437 that will make the state change to not be remembered
1438 - Implemented ``no_remember`` at RAPI level
1439 - Improved the documentation
1440 - Node evacuation: don't call IAllocator if node is already empty
1441 - Fixed bug in DRBD8 replace disks on current nodes
1442 - Fixed bug in recreate-disks for DRBD instances
1443 - Moved assertion checking locks in ``gnt-instance replace-disks``
1444 causing it to abort with not owning the right locks for some situation
1445 - Job queue: Fixed potential race condition when cancelling queued jobs
1446 - Fixed off-by-one bug in job serial generation
1447 - ``gnt-node volumes``: Fix instance names
1448 - Fixed aliases in bash completion
1449 - Fixed a bug in reopening log files after being sent a SIGHUP
1450 - Added a flag to burnin to allow specifying VCPU count
1451 - Bugfixes to non-root Ganeti configuration
1457 *(Released Thu, 12 May 2011)*
1459 Many bug-fixes and a few new small features:
1461 - Fixed a bug related to log opening failures
1462 - Fixed a bug in instance listing with orphan instances
1463 - Fixed a bug which prevented resetting the cluster-level node parameter
1464 ``oob_program`` to the default
1465 - Many fixes related to the ``cluster-merge`` tool
1466 - Fixed a race condition in the lock monitor, which caused failures
1467 during (at least) creation of many instances in parallel
1468 - Improved output for gnt-job info
1469 - Removed the quiet flag on some ssh calls which prevented debugging
1471 - Improved the N+1 failure messages in cluster verify by actually
1472 showing the memory values (needed and available)
1473 - Increased lock attempt timeouts so that when executing long operations
1474 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
1475 too early and thus prevent the use of the 'fair' mechanism
1476 - Changed instance query data (``gnt-instance info``) to not acquire
1477 locks unless needed, thus allowing its use on locked instance if only
1478 static information is asked for
1479 - Improved behaviour with filesystems that do not support rename on an
1481 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
1482 kept locks on all nodes during the wipe, which is unneeded
1483 - Fixed ``gnt-watcher`` handling of errors during hooks execution
1484 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
1485 10GiB) which caused the wipe to fail right at the end in some cases
1486 - Fixed master IP activation when doing master failover with no-voting
1487 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
1488 the master node itself
1489 - Fixed potential data-loss in under disk full conditions, where Ganeti
1490 wouldn't check correctly the return code and would consider
1491 partially-written files 'correct'
1492 - Fixed bug related to multiple VGs and DRBD disk replacing
1493 - Added new disk parameter ``metavg`` that allows placement of the meta
1494 device for DRBD in a different volume group
1495 - Fixed error handling in the node daemon when the system libc doesn't
1496 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
1497 - Fixed lock release during replace-disks, which kept cluster-wide locks
1498 when doing disk replaces with an iallocator script
1499 - Added check for missing bridges in cluster verify
1500 - Handle EPIPE errors while writing to the terminal better, so that
1501 piping the output to e.g. ``less`` doesn't cause a backtrace
1502 - Fixed rare case where a ^C during Luxi calls could have been
1503 interpreted as server errors, instead of simply terminating
1504 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
1506 - Added a few more parameters to the KVM hypervisor, allowing a second
1507 CDROM, custom disk type for CDROMs and a floppy image
1508 - Removed redundant message in instance rename when the name is given
1510 - Added option to ``gnt-instance recreate-disks`` to allow creating the
1511 disks on new nodes, allowing recreation when the original instance
1512 nodes are completely gone
1513 - Added option when converting disk templates to DRBD to skip waiting
1514 for the resync, in order to make the instance available sooner
1515 - Added two new variables to the OS scripts environment (containing the
1517 - Made the root_path and optional parameter for the xen-pvm hypervisor,
1518 to allow use of ``pvgrub`` as bootloader
1519 - Changed the instance memory modifications to only check out-of-memory
1520 conditions on memory increases, and turned the secondary node warnings
1521 into errors (they can still be overridden via ``--force``)
1522 - Fixed the handling of a corner case when the Python installation gets
1523 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
1524 to execute a command that doesn't exist
1525 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
1526 node of the instance returned failures during instance shutdown; this
1527 adds the option ``--ignore-consistency`` to gnt-instance move
1529 And as usual, various improvements to the error messages, documentation
1536 *(Released Wed, 09 Mar 2011)*
1538 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
1539 the RAPI users file if run twice (even with ``--dry-run``).
1541 The release fixes that bug (nothing else changed).
1547 *(Released Mon, 07 Mar 2011)*
1549 Final 2.4.0 release. Just a few small fixes:
1551 - Fixed RAPI node evacuate
1552 - Fixed the kvm-ifup script
1553 - Fixed internal error handling for special job cases
1554 - Updated man page to specify the escaping feature for options
1560 *(Released Mon, 28 Feb 2011)*
1562 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
1563 that this feature wiped the disks of the wrong instance, leading to loss
1568 - Fixed title of query field containing instance name
1569 - Expanded the glossary in the documentation
1570 - Fixed one unittest (internal issue)
1576 *(Released Mon, 21 Feb 2011)*
1578 A number of bug fixes plus just a couple functionality changes.
1580 On the user-visible side, the ``gnt-* list`` command output has changed
1581 with respect to "special" field states. The current rc1 style of display
1582 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
1583 the default output mode special fields states are displayed as follows:
1585 - Offline resource: ``*``
1586 - Unavailable/not applicable: ``-``
1587 - Data missing (RPC failure): ``?``
1588 - Unknown field: ``??``
1590 Another user-visible change is the addition of ``--force-join`` to
1595 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
1596 - Fixed regression in RAPI/instance reinstall where all parameters were
1597 required (instead of optional)
1598 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
1599 - Fixed iallocator usage (offline nodes were not considered offline)
1600 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
1601 - Fixed hypervisor and OS parameter validation with respect to
1602 non-vm_capable nodes
1603 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
1605 - Fixed ``tools/listrunner`` with respect to agent-based usage
1611 *(Released Fri, 4 Feb 2011)*
1613 Many changes and fixes since the beta1 release. While there were some
1614 internal changes, the code has been mostly stabilised for the RC
1617 Note: the dumb allocator was removed in this release, as it was not kept
1618 up-to-date with the IAllocator protocol changes. It is recommended to
1619 use the ``hail`` command from the ganeti-htools package.
1621 Note: the 2.4 and up versions of Ganeti are not compatible with the
1622 0.2.x branch of ganeti-htools. You need to upgrade to
1623 ganeti-htools-0.3.0 (or later).
1625 Regressions fixed from 2.3
1626 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1628 - Fixed the ``gnt-cluster verify-disks`` command
1629 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
1631 - Fixed disk adoption breakage
1632 - Fixed wrong headers in instance listing for field aliases
1637 - Fixed corner case in KVM handling of NICs
1638 - Fixed many cases of wrong handling of non-vm_capable nodes
1639 - Fixed a bug where a missing instance symlink was not possible to
1640 recreate with any ``gnt-*`` command (now ``gnt-instance
1641 activate-disks`` does it)
1642 - Fixed the volume group name as reported by ``gnt-cluster
1644 - Increased timeouts for the import-export code, hopefully leading to
1645 fewer aborts due network or instance timeouts
1646 - Fixed bug in ``gnt-node list-storage``
1647 - Fixed bug where not all daemons were started on cluster
1648 initialisation, but only at the first watcher run
1649 - Fixed many bugs in the OOB implementation
1650 - Fixed watcher behaviour in presence of instances with offline
1652 - Fixed instance list output for instances running on the wrong node
1653 - a few fixes to the cluster-merge tool, but it still cannot merge
1654 multi-node groups (currently it is not recommended to use this tool)
1660 - Improved network configuration for the KVM hypervisor
1661 - Added e1000 as a supported NIC for Xen-HVM
1662 - Improved the lvmstrap tool to also be able to use partitions, as
1663 opposed to full disks
1664 - Improved speed of disk wiping (the cluster parameter
1665 ``prealloc_wipe_disks``, so that it has a low impact on the total time
1666 of instance creations
1667 - Added documentation for the OS parameters
1668 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
1669 hypervisor is not responding
1670 - Added display of blacklisted and hidden OS information in
1671 ``gnt-cluster info``
1672 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
1673 NIC and node parameters, which might create problems with currently
1674 invalid (but undetected) configuration files, but prevents validation
1675 failures when unrelated parameters are modified
1676 - Changed cluster initialisation to wait for the master daemon to become
1678 - Expanded the RAPI interface:
1680 - Added config redistribution resource
1681 - Added activation/deactivation of instance disks
1682 - Added export of console information
1684 - Implemented log file reopening on SIGHUP, which allows using
1685 logrotate(8) for the Ganeti log files
1686 - Added a basic OOB helper script as an example
1692 *(Released Fri, 14 Jan 2011)*
1697 - Fixed timezone issues when formatting timestamps
1698 - Added support for node groups, available via ``gnt-group`` and other
1700 - Added out-of-band framework and management, see :doc:`design
1701 document <design-oob>`
1702 - Removed support for roman numbers from ``gnt-node list`` and
1703 ``gnt-instance list``.
1704 - Allowed modification of master network interface via ``gnt-cluster
1705 modify --master-netdev``
1706 - Accept offline secondaries while shutting down instance disks
1707 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
1708 - Added support for multiple LVM volume groups
1709 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
1711 - Added commands to list available fields:
1713 - ``gnt-node list-fields``
1714 - ``gnt-group list-fields``
1715 - ``gnt-instance list-fields``
1717 - Updated documentation and man pages
1722 - Moved ``rapi_users`` file into separate directory, now named
1723 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
1725 - Added new tool for running commands on many machines,
1726 ``tools/ganeti-listrunner``
1727 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
1728 improving the ``gnt-instance console`` output
1729 - Allowed customisation of disk index separator at ``configure`` time
1730 - Export node group allocation policy to :doc:`iallocator <iallocator>`
1731 - Added support for non-partitioned md disks in ``lvmstrap``
1732 - Added script to gracefully power off KVM instances
1733 - Split ``utils`` module into smaller parts
1734 - Changed query operations to return more detailed information, e.g.
1735 whether an information is unavailable due to an offline node. To use
1736 this new functionality, the LUXI call ``Query`` must be used. Field
1737 information is now stored by the master daemon and can be retrieved
1738 using ``QueryFields``. Instances, nodes and groups can also be queried
1739 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
1740 exposed via RAPI). The following commands make use of this
1741 infrastructure change:
1743 - ``gnt-group list``
1744 - ``gnt-group list-fields``
1746 - ``gnt-node list-fields``
1747 - ``gnt-instance list``
1748 - ``gnt-instance list-fields``
1749 - ``gnt-debug locks``
1754 - New RAPI resources (see :doc:`rapi`):
1758 - ``/2/groups/[group_name]``
1759 - ``/2/groups/[group_name]/assign-nodes``
1760 - ``/2/groups/[group_name]/modify``
1761 - ``/2/groups/[group_name]/rename``
1762 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
1766 - Implemented ``no_install`` for instance creation
1767 - Implemented OS parameters for instance reinstallation, allowing
1768 use of special settings on reinstallation (e.g. for preserving data)
1773 - Added IPv6 support in import/export
1774 - Pause DRBD synchronization while wiping disks on instance creation
1775 - Updated unittests and QA scripts
1776 - Improved network parameters passed to KVM
1777 - Converted man pages from docbook to reStructuredText
1783 *(Released Mon, 20 Dec 2010)*
1785 Released version 2.3.1~rc1 without any changes.
1791 *(Released Wed, 1 Dec 2010)*
1793 - impexpd: Disable OpenSSL compression in socat if possible (backport
1794 from master, commit e90739d625b, see :doc:`installation guide
1795 <install-quick>` for details)
1796 - Changed unittest coverage report to exclude test scripts
1797 - Added script to check version format
1803 *(Released Wed, 1 Dec 2010)*
1805 Released version 2.3.0~rc1 without any changes.
1811 *(Released Fri, 19 Nov 2010)*
1813 A number of bugfixes and documentation updates:
1815 - Update ganeti-os-interface documentation
1816 - Fixed a bug related to duplicate MACs or similar items which should be
1818 - Fix breakage in OS state modify
1819 - Reinstall instance: disallow offline secondaries (fixes bug related to
1820 OS changing but reinstall failing)
1821 - plus all the other fixes between 2.2.1 and 2.2.2
1827 *(Released Tue, 2 Nov 2010)*
1829 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
1830 - Fixed master failover race with watcher
1831 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
1833 - Accept previously stopped instance for export with instance removal
1834 - Simplify and extend the environment variables for instance OS scripts
1835 - Added new node flags, ``master_capable`` and ``vm_capable``
1836 - Added optional instance disk wiping prior during allocation. This is a
1837 cluster-wide option and can be set/modified using
1838 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
1839 - Added IPv6 support, see :doc:`design document <design-2.3>` and
1840 :doc:`install-quick`
1841 - Added a new watcher option (``--ignore-pause``)
1842 - Added option to ignore offline node on instance start/stop
1843 (``--ignore-offline``)
1844 - Allow overriding OS parameters with ``gnt-instance reinstall``
1845 - Added ability to change node's secondary IP address using ``gnt-node
1847 - Implemented privilege separation for all daemons except
1848 ``ganeti-noded``, see ``configure`` options
1849 - Complain if an instance's disk is marked faulty in ``gnt-cluster
1851 - Implemented job priorities (see ``ganeti(7)`` manpage)
1852 - Ignore failures while shutting down instances during failover from
1854 - Exit daemon's bootstrap process only once daemon is ready
1855 - Export more information via ``LUInstanceQuery``/remote API
1856 - Improved documentation, QA and unittests
1857 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
1858 restart if the file was created or changed
1859 - Added LUXI protocol version sent with each request and response,
1860 allowing detection of server/client mismatches
1861 - Moved the Python scripts among gnt-* and ganeti-* into modules
1862 - Moved all code related to setting up SSH to an external script,
1864 - Infrastructure changes for node group support in future versions
1870 *(Released Fri, 19 Nov 2010)*
1872 A few small bugs fixed, and some improvements to the build system:
1874 - Fix documentation regarding conversion to drbd
1875 - Fix validation of parameters in cluster modify (``gnt-cluster modify
1877 - Fix error handling in node modify with multiple changes
1878 - Allow remote imports without checked names
1884 *(Released Tue, 19 Oct 2010)*
1886 - Disable SSL session ID cache in RPC client
1892 *(Released Thu, 14 Oct 2010)*
1894 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
1895 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
1897 - Fix problems with interaction between Curl and Python's HTTP server,
1898 resulting in increased speed in many RPC calls
1899 - Improve our release script to prevent breakage with older aclocal and
1906 *(Released Thu, 7 Oct 2010)*
1908 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
1909 value retrieved from master
1910 - Added support for blacklisted or hidden OS definitions
1911 - Added simple lock monitor (accessible via (``gnt-debug locks``)
1912 - Added support for -mem-path in KVM hypervisor abstraction layer
1913 - Allow overriding instance parameters in tool for inter-cluster
1914 instance moves (``tools/move-instance``)
1915 - Improved opcode summaries (e.g. in ``gnt-job list``)
1916 - Improve consistency of OS listing by sorting it
1917 - Documentation updates
1923 *(Released Fri, 8 Oct 2010)*
1925 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
1931 *(Released Mon, 4 Oct 2010)*
1933 - Fixed regression in ``gnt-instance rename``
1939 *(Released Wed, 22 Sep 2010)*
1941 - Fixed OS_VARIANT variable for OS scripts
1942 - Fixed cluster tag operations via RAPI
1943 - Made ``setup-ssh`` exit with non-zero code if an error occurred
1944 - Disabled RAPI CA checks in watcher
1950 *(Released Mon, 23 Aug 2010)*
1952 - Support DRBD versions of the format "a.b.c.d"
1954 - Re-introduce support for usage from multiple threads in RAPI client
1955 - Instance renames and modify via RAPI
1956 - Work around race condition between processing and archival in job
1958 - Mark opcodes following failed one as failed, too
1959 - Job field ``lock_status`` was removed due to difficulties making it
1960 work with the changed job queue in Ganeti 2.2; a better way to monitor
1961 locks is expected for a later 2.2.x release
1962 - Fixed dry-run behaviour with many commands
1963 - Support ``ssh-agent`` again when adding nodes
1964 - Many additional bugfixes
1970 *(Released Fri, 30 Jul 2010)*
1972 Important change: the internal RPC mechanism between Ganeti nodes has
1973 changed from using a home-grown http library (based on the Python base
1974 libraries) to use the PycURL library. This requires that PycURL is
1975 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
1976 against GnuTLS by default. cURL's support for GnuTLS had known issues
1977 before cURL 7.21.0 and we recommend using the latest cURL release or
1978 linking against OpenSSL. Most other distributions already link PycURL
1979 and cURL against OpenSSL. The command::
1981 python -c 'import pycurl; print pycurl.version'
1983 can be used to determine the libraries PycURL and cURL are linked
1986 Other significant changes:
1988 - Rewrote much of the internals of the job queue, in order to achieve
1989 better parallelism; this decouples job query operations from the job
1990 processing, and it should allow much nicer behaviour of the master
1991 daemon under load, and it also has uncovered some long-standing bugs
1992 related to the job serialisation (now fixed)
1993 - Added a default iallocator setting to the cluster parameters,
1994 eliminating the need to always pass nodes or an iallocator for
1995 operations that require selection of new node(s)
1996 - Added experimental support for the LXC virtualization method
1997 - Added support for OS parameters, which allows the installation of
1998 instances to pass parameter to OS scripts in order to customise the
2000 - Added a hypervisor parameter controlling the migration type (live or
2001 non-live), since hypervisors have various levels of reliability; this
2002 has renamed the 'live' parameter to 'mode'
2003 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
2004 logical volumes, meaning that cluster verify will ignore them and not
2005 flag their presence as errors
2006 - The watcher will now reset the error count for failed instances after
2007 8 hours, thus allowing self-healing if the problem that caused the
2008 instances to be down/fail to start has cleared in the meantime
2009 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
2010 check for, and warn, if the drbd module parameter ``usermode_helper``
2011 is not consistent with the cluster-wide setting; this is needed to
2012 make diagnose easier of failed drbd creations
2013 - Started adding base IPv6 support, but this is not yet
2014 enabled/available for use
2015 - Rename operations (cluster, instance) will now return the new name,
2016 which is especially useful if a short name was passed in
2017 - Added support for instance migration in RAPI
2018 - Added a tool to pre-configure nodes for the SSH setup, before joining
2019 them to the cluster; this will allow in the future a simplified model
2020 for node joining (but not yet fully enabled in 2.2); this needs the
2021 paramiko python library
2022 - Fixed handling of name-resolving errors
2023 - Fixed consistency of job results on the error path
2024 - Fixed master-failover race condition when executed multiple times in
2026 - Fixed many bugs related to the job queue (mostly introduced during the
2027 2.2 development cycle, so not all are impacting 2.1)
2028 - Fixed instance migration with missing disk symlinks
2029 - Fixed handling of unknown jobs in ``gnt-job archive``
2030 - And many other small fixes/improvements
2034 - Enhanced both the unittest and the QA coverage
2035 - Switched the opcode validation to a generic model, and extended the
2036 validation to all opcode parameters
2037 - Changed more parts of the code that write shell scripts to use the
2039 - Switched the master daemon to use the asyncore library for the Luxi
2046 *(Released Thu, 17 Jun 2010)*
2048 - Added tool (``move-instance``) and infrastructure to move instances
2049 between separate clusters (see :doc:`separate documentation
2050 <move-instance>` and :doc:`design document <design-2.2>`)
2051 - Added per-request RPC timeout
2052 - RAPI now requires a Content-Type header for requests with a body (e.g.
2053 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
2054 :rfc:`2616` (HTTP/1.1), section 7.2.1)
2055 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
2057 - Implemented initial support for running Ganeti daemons as separate
2058 users, see configure-time flags ``--with-user-prefix`` and
2059 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
2061 - Instances can be removed after export (``gnt-backup export
2062 --remove-instance``)
2063 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
2064 key (instead of 1024 bit)
2065 - Added new cluster configuration file for cluster domain secret
2066 - Import/export now use SSL instead of SSH
2067 - Added support for showing estimated time when exporting an instance,
2068 see the ``ganeti-os-interface(7)`` manpage and look for
2075 *(Released Tue, 16 Nov 2010)*
2077 Some more bugfixes. Unless critical bugs occur, this will be the last
2080 - Fix case of MAC special-values
2081 - Fix mac checker regex
2082 - backend: Fix typo causing "out of range" error
2083 - Add missing --units in gnt-instance list man page
2089 *(Released Tue, 24 Aug 2010)*
2092 - Don't ignore secondary node silently on non-mirrored disk templates
2094 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
2095 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
2096 - Properly document the usb_mouse parameter
2097 - Fix path in ganeti-rapi(8) (issue 116)
2098 - Adjust error message when the ganeti user's .ssh directory is
2100 - Add same-node-check when changing the disk template to drbd
2106 *(Released Fri, 16 Jul 2010)*
2109 - Add an option to only select some reboot types during qa/burnin.
2110 (on some hypervisors consequent reboots are not supported)
2111 - Fix infrequent race condition in master failover. Sometimes the old
2112 master ip address would be still detected as up for a short time
2113 after it was removed, causing failover to fail.
2114 - Decrease mlockall warnings when the ctypes module is missing. On
2115 Python 2.4 we support running even if no ctypes module is installed,
2116 but we were too verbose about this issue.
2117 - Fix building on old distributions, on which man doesn't have a
2119 - Fix RAPI not to ignore the MAC address on instance creation
2120 - Implement the old instance creation format in the RAPI client.
2126 *(Released Thu, 01 Jul 2010)*
2128 A small bugfix release:
2129 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
2130 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
2131 a non-existing option
2132 - Fix instance create: the --force-variant option was ignored
2133 - Improve pylint 0.21 compatibility and warnings with Python 2.6
2134 - Fix modify node storage with non-FQDN arguments
2135 - Fix RAPI client to authenticate under Python 2.6 when used
2136 for more than 5 requests needing authentication
2137 - Fix gnt-instance modify -t (storage) giving a wrong error message
2138 when converting a non-shutdown drbd instance to plain
2144 *(Released Fri, 18 Jun 2010)*
2146 A small bugfix release:
2148 - Fix live migration of KVM instances started with older Ganeti
2149 versions which had fewer hypervisor parameters
2150 - Fix gnt-instance grow-disk on down instances
2151 - Fix an error-reporting bug during instance migration
2152 - Better checking of the ``--net`` and ``--disk`` values, to avoid
2153 silently ignoring broken ones
2154 - Fix an RPC error reporting bug affecting, for example, RAPI client
2156 - Fix bug triggered by different API version os-es on different nodes
2157 - Fix a bug in instance startup with custom hvparams: OS level
2158 parameters would fail to be applied.
2159 - Fix the RAPI client under Python 2.6 (but more work is needed to
2160 make it work completely well with OpenSSL)
2161 - Fix handling of errors when resolving names from DNS
2167 *(Released Thu, 3 Jun 2010)*
2169 A medium sized development cycle. Some new features, and some
2170 fixes/small improvements/cleanups.
2172 Significant features
2173 ~~~~~~~~~~~~~~~~~~~~
2175 The node deamon now tries to mlock itself into memory, unless the
2176 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
2177 its logs, and falls back to console logging. This allows emergency
2178 features such as ``gnt-node powercycle`` to work even in the event of a
2179 broken node disk (tested offlining the disk hosting the node's
2180 filesystem and dropping its memory caches; don't try this at home)
2182 KVM: add vhost-net acceleration support. It can be tested with a new
2183 enough version of the kernel and of qemu-kvm.
2185 KVM: Add instance chrooting feature. If you use privilege dropping for
2186 your VMs you can also now force them to chroot to an empty directory,
2187 before starting the emulated guest.
2189 KVM: Add maximum migration bandwith and maximum downtime tweaking
2190 support (requires a new-enough version of qemu-kvm).
2192 Cluster verify will now warn if the master node doesn't have the master
2193 ip configured on it.
2195 Add a new (incompatible) instance creation request format to RAPI which
2196 supports all parameters (previously only a subset was supported, and it
2197 wasn't possible to extend the old format to accomodate all the new
2198 features. The old format is still supported, and a client can check for
2199 this feature, before using it, by checking for its presence in the
2200 ``features`` RAPI resource.
2202 Now with ancient latin support. Try it passing the ``--roman`` option to
2203 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
2204 (requires the python-roman module to be installed, in order to work).
2209 As usual many internal code refactorings, documentation updates, and
2212 - Lots of improvements and cleanups to the experimental Remote API
2213 (RAPI) client library.
2214 - A new unit test suite for the core daemon libraries.
2215 - A fix to creating missing directories makes sure the umask is not
2216 applied anymore. This enforces the same directory permissions
2218 - Better handling terminating daemons with ctrl+c (used when running
2219 them in debugging mode).
2220 - Fix a race condition in live migrating a KVM instance, when stat()
2221 on the old proc status file returned EINVAL, which is an unexpected
2223 - Fixed manpage checking with newer man and utf-8 charachters. But now
2224 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
2230 *(Released Fri, 7 May 2010)*
2232 Fix a bug which prevented untagged KVM instances from starting.
2238 *(Released Fri, 7 May 2010)*
2240 Another release with a long development cycle, during which many
2241 different features were added.
2243 Significant features
2244 ~~~~~~~~~~~~~~~~~~~~
2246 The KVM hypervisor now can run the individual instances as non-root, to
2247 reduce the impact of a VM being hijacked due to bugs in the
2248 hypervisor. It is possible to run all instances as a single (non-root)
2249 user, to manually specify a user for each instance, or to dynamically
2250 allocate a user out of a cluster-wide pool to each instance, with the
2251 guarantee that no two instances will run under the same user ID on any
2254 An experimental RAPI client library, that can be used standalone
2255 (without the other Ganeti libraries), is provided in the source tree as
2256 ``lib/rapi/client.py``. Note this client might change its interface in
2257 the future, as we iterate on its capabilities.
2259 A new command, ``gnt-cluster renew-crypto`` has been added to easily
2260 replace the cluster's certificates and crypto keys. This might help in
2261 case they have been compromised, or have simply expired.
2263 A new disk option for instance creation has been added that allows one
2264 to "adopt" currently existing logical volumes, with data
2265 preservation. This should allow easier migration to Ganeti from
2266 unmanaged (or managed via other software) instances.
2268 Another disk improvement is the possibility to convert between redundant
2269 (DRBD) and plain (LVM) disk configuration for an instance. This should
2270 allow better scalability (starting with one node and growing the
2271 cluster, or shrinking a two-node cluster to one node).
2273 A new feature that could help with automated node failovers has been
2274 implemented: if a node sees itself as offline (by querying the master
2275 candidates), it will try to shutdown (hard) all instances and any active
2276 DRBD devices. This reduces the risk of duplicate instances if an
2277 external script automatically failovers the instances on such nodes. To
2278 enable this, the cluster parameter ``maintain_node_health`` should be
2279 enabled; in the future this option (per the name) will enable other
2280 automatic maintenance features.
2282 Instance export/import now will reuse the original instance
2283 specifications for all parameters; that means exporting an instance,
2284 deleting it and the importing it back should give an almost identical
2285 instance. Note that the default import behaviour has changed from
2286 before, where it created only one NIC; now it recreates the original
2289 Cluster verify has added a few new checks: SSL certificates validity,
2290 /etc/hosts consistency across the cluster, etc.
2295 As usual, many internal changes were done, documentation fixes,
2298 - Fixed cluster initialization with disabled cluster storage (regression
2299 introduced in 2.1.1)
2300 - File-based storage supports growing the disks
2301 - Fixed behaviour of node role changes
2302 - Fixed cluster verify for some corner cases, plus a general rewrite of
2303 cluster verify to allow future extension with more checks
2304 - Fixed log spamming by watcher and node daemon (regression introduced
2306 - Fixed possible validation issues when changing the list of enabled
2308 - Fixed cleanup of /etc/hosts during node removal
2309 - Fixed RAPI response for invalid methods
2310 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
2311 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
2312 from ide disks, etc.)
2313 - Allow OS changes without re-installation (to record a changed OS
2314 outside of Ganeti, or to allow OS renames)
2315 - Allow instance creation without OS installation (useful for example if
2316 the OS will be installed manually, or restored from a backup not in
2318 - Implemented option to make cluster ``copyfile`` use the replication
2320 - Added list of enabled hypervisors to ssconf (possibly useful for
2322 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
2324 - A partial form of node re-IP is possible via node readd, which now
2325 allows changed node primary IP
2326 - Command line utilities now show an informational message if the job is
2328 - The logs of the master daemon now show the PID/UID/GID of the
2335 *(Released Fri, 12 Mar 2010)*
2337 During the 2.1.0 long release candidate cycle, a lot of improvements and
2338 changes have accumulated with were released later as 2.1.1.
2343 The node evacuate command (``gnt-node evacuate``) was significantly
2344 rewritten, and as such the IAllocator protocol was changed - a new
2345 request type has been added. This unfortunate change during a stable
2346 series is designed to improve performance of node evacuations; on
2347 clusters with more than about five nodes and which are well-balanced,
2348 evacuation should proceed in parallel for all instances of the node
2349 being evacuated. As such, any existing IAllocator scripts need to be
2350 updated, otherwise the above command will fail due to the unknown
2351 request. The provided "dumb" allocator has not been updated; but the
2352 ganeti-htools package supports the new protocol since version 0.2.4.
2354 Another important change is increased validation of node and instance
2355 names. This might create problems in special cases, if invalid host
2356 names are being used.
2358 Also, a new layer of hypervisor parameters has been added, that sits at
2359 OS level between the cluster defaults and the instance ones. This allows
2360 customisation of virtualization parameters depending on the installed
2361 OS. For example instances with OS 'X' may have a different KVM kernel
2362 (or any other parameter) than the cluster defaults. This is intended to
2363 help managing a multiple OSes on the same cluster, without manual
2364 modification of each instance's parameters.
2366 A tool for merging clusters, ``cluster-merge``, has been added in the
2367 tools sub-directory.
2372 - Improved the int/float conversions that should make the code more
2373 robust in face of errors from the node daemons
2374 - Fixed the remove node code in case of internal configuration errors
2375 - Fixed the node daemon behaviour in face of inconsistent queue
2376 directory (e.g. read-only file-system where we can't open the files
2378 - Fixed the behaviour of gnt-node modify for master candidate demotion;
2379 now it either aborts cleanly or, if given the new "auto_promote"
2380 parameter, will automatically promote other nodes as needed
2381 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
2382 completely prevent Ganeti from working
2383 - Fixed bug for instance export when not all disks were successfully
2385 - Fixed behaviour of node add when the new node is slow in starting up
2387 - Fixed handling of signals in the LUXI client, which should improve
2388 behaviour of command-line scripts
2389 - Added checks for invalid node/instance names in the configuration (now
2390 flagged during cluster verify)
2391 - Fixed watcher behaviour for disk activation errors
2392 - Fixed two potentially endless loops in http library, which led to the
2393 RAPI daemon hanging and consuming 100% CPU in some cases
2394 - Fixed bug in RAPI daemon related to hashed passwords
2395 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
2397 - Enhanced compatibility with non-Debian OSes, but not using absolute
2398 path in some commands and allowing customisation of the ssh
2399 configuration directory
2400 - Fixed possible future issue with new Python versions by abiding to the
2401 proper use of ``__slots__`` attribute on classes
2402 - Added checks that should prevent directory traversal attacks
2403 - Many documentation fixes based on feedback from users
2408 - Added an "early_release" more for instance replace disks and node
2409 evacuate, where we release locks earlier and thus allow higher
2410 parallelism within the cluster
2411 - Added watcher hooks, intended to allow the watcher to restart other
2412 daemons (e.g. from the ganeti-nbma project), but they can be used of
2413 course for any other purpose
2414 - Added a compile-time disable for DRBD barriers, to increase
2415 performance if the administrator trusts the power supply or the
2416 storage system to not lose writes
2417 - Added the option of using syslog for logging instead of, or in
2418 addition to, Ganeti's own log files
2419 - Removed boot restriction for paravirtual NICs for KVM, recent versions
2420 can indeed boot from a paravirtual NIC
2421 - Added a generic debug level for many operations; while this is not
2422 used widely yet, it allows one to pass the debug value all the way to
2424 - Enhanced the hooks environment for instance moves (failovers,
2425 migrations) where the primary/secondary nodes changed during the
2426 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
2427 - Enhanced data validations for many user-supplied values; one important
2428 item is the restrictions imposed on instance and node names, which
2429 might reject some (invalid) host names
2430 - Add a configure-time option to disable file-based storage, if it's not
2431 needed; this allows greater security separation between the master
2432 node and the other nodes from the point of view of the inter-node RPC
2434 - Added user notification in interactive tools if job is waiting in the
2435 job queue or trying to acquire locks
2436 - Added log messages when a job is waiting for locks
2437 - Added filtering by node tags in instance operations which admit
2438 multiple instances (start, stop, reboot, reinstall)
2439 - Added a new tool for cluster mergers, ``cluster-merge``
2440 - Parameters from command line which are of the form ``a=b,c=d`` can now
2441 use backslash escapes to pass in values which contain commas,
2442 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
2444 - For KVM, the instance name is the first parameter passed to KVM, so
2445 that it's more visible in the process list
2451 *(Released Tue, 2 Mar 2010)*
2453 Ganeti 2.1 brings many improvements with it. Major changes:
2455 - Added infrastructure to ease automated disk repairs
2456 - Added new daemon to export configuration data in a cheaper way than
2457 using the remote API
2458 - Instance NICs can now be routed instead of being associated with a
2460 - Improved job locking logic to reduce impact of jobs acquiring multiple
2461 locks waiting for other long-running jobs
2463 In-depth implementation details can be found in the Ganeti 2.1 design
2469 - Added chroot hypervisor
2470 - Added more options to xen-hvm hypervisor (``kernel_path`` and
2472 - Added more options to xen-pvm hypervisor (``use_bootloader``,
2473 ``bootloader_path`` and ``bootloader_args``)
2474 - Added the ``use_localtime`` option for the xen-hvm and kvm
2475 hypervisors, and the default value for this has changed to false (in
2476 2.0 xen-hvm always enabled it)
2477 - Added luxi call to submit multiple jobs in one go
2478 - Added cluster initialization option to not modify ``/etc/hosts``
2480 - Added network interface parameters
2481 - Added dry run mode to some LUs
2482 - Added RAPI resources:
2484 - ``/2/instances/[instance_name]/info``
2485 - ``/2/instances/[instance_name]/replace-disks``
2486 - ``/2/nodes/[node_name]/evacuate``
2487 - ``/2/nodes/[node_name]/migrate``
2488 - ``/2/nodes/[node_name]/role``
2489 - ``/2/nodes/[node_name]/storage``
2490 - ``/2/nodes/[node_name]/storage/modify``
2491 - ``/2/nodes/[node_name]/storage/repair``
2493 - Added OpCodes to evacuate or migrate all instances on a node
2494 - Added new command to list storage elements on nodes (``gnt-node
2495 list-storage``) and modify them (``gnt-node modify-storage``)
2496 - Added new ssconf files with master candidate IP address
2497 (``ssconf_master_candidates_ips``), node primary IP address
2498 (``ssconf_node_primary_ips``) and node secondary IP address
2499 (``ssconf_node_secondary_ips``)
2500 - Added ``ganeti-confd`` and a client library to query the Ganeti
2501 configuration via UDP
2502 - Added ability to run hooks after cluster initialization and before
2504 - Added automatic mode for disk replace (``gnt-instance replace-disks
2506 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
2507 after catastrophic data-loss
2508 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
2510 - Added ``gnt-instance move`` command to move instances
2511 - Added ``gnt-cluster watcher`` command to control watcher
2512 - Added ``gnt-node powercycle`` command to powercycle nodes
2513 - Added new job status field ``lock_status``
2514 - Added parseable error codes to cluster verification (``gnt-cluster
2515 verify --error-codes``) and made output less verbose (use
2516 ``--verbose`` to restore previous behaviour)
2517 - Added UUIDs to the main config entities (cluster, nodes, instances)
2518 - Added support for OS variants
2519 - Added support for hashed passwords in the Ganeti remote API users file
2521 - Added option to specify maximum timeout on instance shutdown
2522 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
2523 - Added new helper script to start and stop Ganeti daemons
2524 (``daemon-util``), with the intent to reduce the work necessary to
2525 adjust Ganeti for non-Debian distributions and to start/stop daemons
2527 - Added more unittests
2528 - Fixed critical bug in ganeti-masterd startup
2529 - Removed the configure-time ``kvm-migration-port`` parameter, this is
2530 now customisable at the cluster level for both the KVM and Xen
2531 hypervisors using the new ``migration_port`` parameter
2532 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
2533 reinstalling an instance
2534 - Allowed ``@`` in tag names
2535 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
2536 - Many documentation updates
2537 - Distribute hypervisor files on ``gnt-cluster redist-conf``
2538 - ``gnt-instance reinstall`` can now reinstall multiple instances
2539 - Updated many command line parameters
2540 - Introduced new OS API version 15
2541 - No longer support a default hypervisor
2542 - Treat virtual LVs as inexistent
2543 - Improved job locking logic to reduce lock contention
2544 - Match instance and node names case insensitively
2545 - Reimplemented bash completion script to be more complete
2552 *(Released Thu, 4 Feb 2010)*
2554 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
2555 - Fix a string formatting bug
2556 - Improve safety of the code in some error paths
2557 - Improve data validation in the master of values returned from nodes
2563 *(Released Thu, 17 Dec 2009)*
2565 - Fix security issue due to missing validation of iallocator names; this
2566 allows local and remote execution of arbitrary executables
2567 - Fix failure of gnt-node list during instance removal
2568 - Ship the RAPI documentation in the archive
2574 *(Released Wed, 30 Sep 2009)*
2576 - Fixed many wrong messages
2577 - Fixed a few bugs related to the locking library
2578 - Fixed MAC checking at instance creation time
2579 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
2580 - Fixed a few issues related to signal handling in both daemons and
2582 - Fixed the example startup script provided
2583 - Fixed insserv dependencies in the example startup script (patch from
2585 - Fixed handling of drained nodes in the iallocator framework
2586 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
2588 - Fixed error related to invalid job IDs in job polling
2589 - Fixed job/opcode persistence on unclean master shutdown
2590 - Fixed handling of partial job processing after unclean master
2592 - Fixed error reporting from LUs, previously all errors were converted
2593 into execution errors
2594 - Fixed error reporting from burnin
2595 - Decreased significantly the memory usage of the job queue
2596 - Optimised slightly multi-job submission
2597 - Optimised slightly opcode loading
2598 - Backported the multi-job submit framework from the development
2599 branch; multi-instance start and stop should be faster
2600 - Added script to clean archived jobs after 21 days; this will reduce
2601 the size of the queue directory
2602 - Added some extra checks in disk size tracking
2603 - Added an example ethers hook script
2604 - Added a cluster parameter that prevents Ganeti from modifying of
2606 - Added more node information to RAPI responses
2607 - Added a ``gnt-job watch`` command that allows following the ouput of a
2609 - Added a bind-address option to ganeti-rapi
2610 - Added more checks to the configuration verify
2611 - Enhanced the burnin script such that some operations can be retried
2613 - Converted instance reinstall to multi-instance model
2619 *(Released Fri, 7 Aug 2009)*
2621 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
2622 to allow using the pre-2.0.2 behaviour in activation, if any existing
2623 instances have mismatched disk sizes in the configuration
2624 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
2625 any configuration mismatches for disk sizes
2626 - Added ``gnt-master cluste-failover --no-voting`` to allow master
2627 failover to work on two-node clusters
2628 - Fixed the ``--net`` option of ``gnt-backup import``, which was
2630 - Fixed detection of OS script errors in ``gnt-backup export``
2631 - Fixed exit code of ``gnt-backup export``
2637 *(Released Fri, 17 Jul 2009)*
2639 - Added experimental support for stripped logical volumes; this should
2640 enhance performance but comes with a higher complexity in the block
2641 device handling; stripping is only enabled when passing
2642 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
2643 affected even in the non-stripped mode
2644 - Improved resiliency against transient failures at the end of DRBD
2645 resyncs, and in general of DRBD resync checks
2646 - Fixed a couple of issues with exports and snapshot errors
2647 - Fixed a couple of issues in instance listing
2648 - Added display of the disk size in ``gnt-instance info``
2649 - Fixed checking for valid OSes in instance creation
2650 - Fixed handling of the "vcpus" parameter in instance listing and in
2651 general of invalid parameters
2652 - Fixed http server library, and thus RAPI, to handle invalid
2653 username/password combinations correctly; this means that now they
2654 report unauthorized for queries too, not only for modifications,
2655 allowing earlier detect of configuration problems
2656 - Added a new "role" node list field, equivalent to the master/master
2657 candidate/drained/offline flags combinations
2658 - Fixed cluster modify and changes of candidate pool size
2659 - Fixed cluster verify error messages for wrong files on regular nodes
2660 - Fixed a couple of issues with node demotion from master candidate role
2661 - Fixed node readd issues
2662 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
2663 - Added a new ``--no-voting`` option for masterfailover to fix failover
2664 on two-nodes clusters when the former master node is unreachable
2665 - Added instance reinstall over RAPI
2671 *(Released Tue, 16 Jun 2009)*
2673 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
2674 allow re-adding the start in single-user option (regression from 1.2)
2675 - the watcher writes the instance status to a file, to allow monitoring
2676 to report the instance status (from the master) based on cached
2677 results of the watcher's queries; while this can get stale if the
2678 watcher is being locked due to other work on the cluster, this is
2679 still an improvement
2680 - the watcher now also restarts the node daemon and the rapi daemon if
2682 - fixed the watcher to handle full and drained queue cases
2683 - hooks export more instance data in the environment, which helps if
2684 hook scripts need to take action based on the instance's properties
2685 (no longer need to query back into ganeti)
2686 - instance failovers when the instance is stopped do not check for free
2687 RAM, so that failing over a stopped instance is possible in low memory
2689 - rapi uses queries for tags instead of jobs (for less job traffic), and
2690 for cluster tags it won't talk to masterd at all but read them from
2692 - a couple of error handling fixes in RAPI
2693 - drbd handling: improved the error handling of inconsistent disks after
2694 resync to reduce the frequency of "there are some degraded disks for
2695 this instance" messages
2696 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
2697 error handling path called a wrong function name)
2703 *(Released Wed, 27 May 2009)*
2705 - no changes from rc5
2711 *(Released Wed, 20 May 2009)*
2713 - fix a couple of bugs (validation, argument checks)
2714 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
2715 - some small improvements to RAPI and IAllocator
2716 - make watcher automatically start the master daemon if down
2722 *(Released Mon, 27 Apr 2009)*
2724 - change the OS list to not require locks; this helps with big clusters
2725 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
2726 volume group is broken
2727 - ``gnt-instance info``, without any arguments, doesn't run for all
2728 instances anymore; either pass ``--all`` or pass the desired
2729 instances; this helps against mistakes on big clusters where listing
2730 the information for all instances takes a long time
2731 - miscellaneous doc and man pages fixes
2737 *(Released Wed, 8 Apr 2009)*
2739 - Change the internal locking model of some ``gnt-node`` commands, in
2740 order to reduce contention (and blocking of master daemon) when
2741 batching many creation/reinstall jobs
2742 - Fixes to Xen soft reboot
2743 - No longer build documentation at build time, instead distribute it in
2744 the archive, in order to reduce the need for the whole docbook/rst
2751 *(Released Fri, 27 Mar 2009)*
2753 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
2754 - Fix watcher startup sequence, improves the behaviour of busy clusters
2755 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
2756 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
2757 bind address and other places
2758 - Some documentation fixes and updates
2764 *(Released Mon, 2 Mar 2009)*
2766 - More documentation updates, now all docs should be more-or-less
2768 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
2770 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
2772 - Fix an issue related to $libdir/run/ganeti and cluster creation
2778 *(Released Thu, 19 Feb 2009)*
2780 - Xen PVM and KVM have switched the default value for the instance root
2781 disk to the first partition on the first drive, instead of the whole
2782 drive; this means that the OS installation scripts must be changed
2784 - Man pages have been updated
2785 - RAPI has been switched by default to HTTPS, and the exported functions
2786 should all work correctly
2787 - RAPI v1 has been removed
2788 - Many improvements to the KVM hypervisor
2789 - Block device errors are now better reported
2790 - Many other bugfixes and small improvements
2796 *(Released Mon, 26 Jan 2009)*
2798 - Version 2 is a general rewrite of the code and therefore the
2799 differences are too many to list, see the design document for 2.0 in
2800 the ``doc/`` subdirectory for more details
2801 - In this beta version there is not yet a migration path from 1.2 (there
2802 will be one in the final 2.0 release)
2803 - A few significant changes are:
2805 - all commands are executed by a daemon (``ganeti-masterd``) and the
2806 various ``gnt-*`` commands are just front-ends to it
2807 - all the commands are entered into, and executed from a job queue,
2808 see the ``gnt-job(8)`` manpage
2809 - the RAPI daemon supports read-write operations, secured by basic
2810 HTTP authentication on top of HTTPS
2811 - DRBD version 0.7 support has been removed, DRBD 8 is the only
2812 supported version (when migrating from Ganeti 1.2 to 2.0, you need
2813 to migrate to DRBD 8 first while still running Ganeti 1.2)
2814 - DRBD devices are using statically allocated minor numbers, which
2815 will be assigned to existing instances during the migration process
2816 - there is support for both Xen PVM and Xen HVM instances running on
2818 - KVM virtualization is supported too
2819 - file-based storage has been implemented, which means that it is
2820 possible to run the cluster without LVM and DRBD storage, for
2821 example using a shared filesystem exported from shared storage (and
2822 still have live migration)
2828 *(Released Tue, 13 Jan 2009)*
2830 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
2831 - Reuse the old instance mac address by default on instance import, if
2832 the instance name is the same.
2833 - Handle situations in which the node info rpc returns incomplete
2835 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
2836 - Improved version of batcher:
2838 - state file support
2839 - instance mac address support
2840 - support for HVM clusters/instances
2842 - Add an option to show the number of cpu sockets and nodes in
2844 - Support OSes that handle more than one version of the OS api (but do
2845 not change the current API in any other way)
2846 - Fix ``gnt-node migrate``
2847 - ``gnt-debug`` man page
2848 - Fixes various more typos and small issues
2849 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
2855 *(Released Wed, 24 Sep 2008)*
2857 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
2858 type of disk exported to fully virtualized instances.
2859 - provide access to the serial console of HVM instances
2860 - instance auto_balance flag, set by default. If turned off it will
2861 avoid warnings on cluster verify if there is not enough memory to fail
2862 over an instance. in the future it will prevent automatically failing
2863 it over when we will support that.
2864 - batcher tool for instance creation, see ``tools/README.batcher``
2865 - ``gnt-instance reinstall --select-os`` to interactively select a new
2866 operating system when reinstalling an instance.
2867 - when changing the memory amount on instance modify a check has been
2868 added that the instance will be able to start. also warnings are
2869 emitted if the instance will not be able to fail over, if auto_balance
2871 - documentation fixes
2872 - sync fields between ``gnt-instance list/modify/add/import``
2873 - fix a race condition in drbd when the sync speed was set after giving
2874 the device a remote peer.
2880 *(Released Tue, 22 Jul 2008)*
2882 - note: the allowed size and number of tags per object were reduced
2883 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
2884 - fixed twisted 8.x compatibility
2885 - fixed ``gnt-instance replace-disks`` with iallocator
2886 - add TCP keepalives on twisted connections to detect restarted nodes
2887 - disk increase support, see ``gnt-instance grow-disk``
2888 - implement bulk node/instance query for RAPI
2889 - add tags in node/instance listing (optional)
2890 - experimental migration (and live migration) support, read the man page
2891 for ``gnt-instance migrate``
2892 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
2893 has some small improvements in handling its state file
2899 *(Released Fri, 13 Jun 2008)*
2901 - Experimental readonly, REST-based remote API implementation;
2902 automatically started on master node, TCP port 5080, if enabled by
2903 ``--enable-rapi`` parameter to configure script.
2904 - Instance allocator support. Add and import instance accept a
2905 ``--iallocator`` parameter, and call that instance allocator to decide
2906 which node to use for the instance. The iallocator document describes
2907 what's expected from an allocator script.
2908 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
2909 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
2910 node is lost there is still enough memory to fail over the instances
2912 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
2913 ``gnt-cluster verify``, to check for site-specific compliance. All the
2914 hooks will run, and their output, if any, will be displayed. Any
2915 failing hook will make the verification return an error value.
2916 - ``gnt-cluster verify`` now checks that its peers are reachable on the
2917 primary and secondary interfaces
2918 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
2919 that is still declared as part of the cluster and has failed.
2920 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
2921 specifying output fields, that just adds the chosen fields to the
2923 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
2924 export from the filesystem.
2925 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
2926 have been added. Using them you can enable/disable acpi and pae
2927 support, and specify a path for a cd image to be exported to the
2928 instance. These parameters as the name suggest only work on HVM
2930 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
2931 PAE support will be set to the previously hardcoded values, but the
2932 (previously hardcoded) path to the CDROM ISO image will be unset and
2933 if required, needs to be set manually with ``gnt-instance modify``
2935 - The address to which an instance's VNC console is bound is now
2936 selectable per-instance, rather than being cluster wide. Of course
2937 this only applies to instances controlled via VNC, so currently just
2938 applies to HVM clusters.
2944 *(Released Mon, 18 Feb 2008)*
2946 - more tweaks to the disk activation code (especially helpful for DRBD)
2947 - change the default ``gnt-instance list`` output format, now there is
2948 one combined status field (see the manpage for the exact values this
2950 - some more fixes for the mac export to hooks change
2951 - make Ganeti not break with DRBD 8.2.x (which changed the version
2952 format in ``/proc/drbd``) (issue 24)
2953 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
2954 template, allowing migration from DRBD0.7+MD to DRBD8
2960 *(Released Wed, 30 Jan 2008)*
2962 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
2964 - add command aliases infrastructure and a few aliases
2965 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
2966 man pages and the ``--help`` option of ``gnt-node
2967 list``/``gnt-instance list``
2968 - fix ``gnt-backup list`` with down nodes (issue 21)
2969 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
2970 - fix the dist archive and add a check for including svn/git files in
2972 - some developer-related changes: improve the burnin and the QA suite,
2973 add an upload script for testing during development
2979 *(Released Wed, 16 Jan 2008)*
2981 - experimental HVM support, read the install document, section
2982 "Initializing the cluster"
2983 - allow for the PVM hypervisor per-instance kernel and initrd paths
2984 - add a new command ``gnt-cluster verify-disks`` which uses a new
2985 algorithm to improve the reconnection of the DRBD pairs if the device
2986 on the secondary node has gone away
2987 - make logical volume code auto-activate LVs at disk activation time
2988 - slightly improve the speed of activating disks
2989 - allow specification of the MAC address at instance creation time, and
2990 changing it later via ``gnt-instance modify``
2991 - fix handling of external commands that generate lots of output on
2993 - update documentation with regard to minimum version of DRBD8 supported
2999 *(Released Tue, 4 Dec 2007)*
3001 - Log the ``xm create`` output to the node daemon log on failure (to
3002 help diagnosing the error)
3003 - In debug mode, log all external commands output if failed to the logs
3004 - Change parsing of lvm commands to ignore stderr
3010 *(Released Wed, 28 Nov 2007)*
3012 - Another round of updates to the DRBD 8 code to deal with more failures
3013 in the replace secondary node operation
3014 - Some more logging of failures in disk operations (lvm, drbd)
3015 - A few documentation updates
3022 *(Released Tue, 13 Nov 2007)*
3024 - Change configuration file format from Python's Pickle to JSON.
3025 Upgrading is possible using the cfgupgrade utility.
3026 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
3027 faster replace disks and is more stable (DRBD 8 has many improvements
3028 compared to DRBD 0.7)
3029 - Added command line tags support (see man pages for ``gnt-instance``,
3030 ``gnt-node``, ``gnt-cluster``)
3031 - Added instance rename support
3032 - Added multi-instance startup/shutdown
3033 - Added cluster rename support
3034 - Added ``gnt-node evacuate`` to simplify some node operations
3035 - Added instance reboot operation that can speedup reboot as compared to
3037 - Soften the requirement that hostnames are in FQDN format
3038 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
3040 - Removed dependency on debian's patched fping that uses the
3041 non-standard ``-S`` option
3042 - Now the OS definitions are searched for in multiple, configurable
3043 paths (easier for distros to package)
3044 - Some changes to the hooks infrastructure (especially the new
3045 post-configuration update hook)
3046 - Other small bugfixes
3048 .. vim: set textwidth=72 syntax=rst :