8 *(Released Fri, 3 May 2013)*
10 Incompatible/important changes
11 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
13 - Instance policies for disk size were documented to be on a per-disk
14 basis, but hail applied them to the sum of all disks. This has been
16 - ``hbal`` will now exit with status 0 if, during job execution over
17 LUXI, early exit has been requested and all jobs are successful;
18 before, exit status 1 was used, which cannot be differentiated from
20 - Compatibility with newer versions of rbd has been fixed
21 - ``gnt-instance batch-create`` has been changed to use the bulk create
22 opcode from Ganeti. This lead to incompatible changes in the format of
23 the JSON file. It's now not a custom dict anymore but a dict
24 compatible with the ``OpInstanceCreate`` opcode.
25 - Parent directories for file storage need to be listed in
26 ``$sysconfdir/ganeti/file-storage-paths`` now. ``cfgupgrade`` will
27 write the file automatically based on old configuration values, but it
28 can not distribute it across all nodes and the file contents should be
29 verified. Use ``gnt-cluster copyfile
30 $sysconfdir/ganeti/file-storage-paths`` once the cluster has been
31 upgraded. The reason for requiring this list of paths now is that
32 before it would have been possible to inject new paths via RPC,
33 allowing files to be created in arbitrary locations. The RPC protocol
34 is protected using SSL/X.509 certificates, but as a design principle
35 Ganeti does not permit arbitrary paths to be passed.
36 - The parsing of the variants file for OSes (see
37 :manpage:`ganeti-os-interface(7)`) has been slightly changed: now empty
38 lines and comment lines (starting with ``#``) are ignored for better
40 - The ``setup-ssh`` tool added in Ganeti 2.2 has been replaced and is no
41 longer available. ``gnt-node add`` now invokes a new tool on the
42 destination node, named ``prepare-node-join``, to configure the SSH
43 daemon. Paramiko is no longer necessary to configure nodes' SSH
44 daemons via ``gnt-node add``.
45 - Draining (``gnt-cluster queue drain``) and un-draining the job queue
46 (``gnt-cluster queue undrain``) now affects all nodes in a cluster and
47 the flag is not reset after a master failover.
48 - Python 2.4 has *not* been tested with this release. Using 2.6 or above
49 is recommended. 2.6 will be mandatory from the 2.8 series.
55 - New network management functionality to support automatic allocation
56 of IP addresses and managing of network parameters. See
57 :manpage:`gnt-network(8)` for more details.
58 - New external storage backend, to allow managing arbitrary storage
59 systems external to the cluster. See
60 :manpage:`ganeti-extstorage-interface(7)`.
61 - New ``exclusive-storage`` node parameter added, restricted to
62 nodegroup level. When it's set to true, physical disks are assigned in
63 an exclusive fashion to instances, as documented in :doc:`Partitioned
64 Ganeti <design-partitioned>`. Currently, only instances using the
65 ``plain`` disk template are supported.
66 - The KVM hypervisor has been updated with many new hypervisor
67 parameters, including a generic one for passing arbitrary command line
68 values. See a complete list in :manpage:`gnt-instance(8)`. It is now
69 compatible up to qemu 1.4.
70 - A new tool, called ``mon-collector``, is the stand-alone executor of
71 the data collectors for a monitoring system. As of this version, it
72 just includes the DRBD data collector, that can be executed by calling
73 ``mon-collector`` using the ``drbd`` parameter. See
74 :manpage:`mon-collector(7)`.
75 - A new user option, :pyeval:`rapi.RAPI_ACCESS_READ`, has been added
76 for RAPI users. It allows granting permissions to query for
77 information to a specific user without giving
78 :pyeval:`rapi.RAPI_ACCESS_WRITE` permissions.
79 - A new tool named ``node-cleanup`` has been added. It cleans remains of
80 a cluster from a machine by stopping all daemons, removing
81 certificates and ssconf files. Unless the ``--no-backup`` option is
82 given, copies of the certificates are made.
83 - Instance creations now support the use of opportunistic locking,
84 potentially speeding up the (parallel) creation of multiple instances.
85 This feature is currently only available via the :doc:`RAPI
86 <rapi>` interface and when an instance allocator is used. If the
87 ``opportunistic_locking`` parameter is set the opcode will try to
88 acquire as many locks as possible, but will not wait for any locks
89 held by other opcodes. If not enough resources can be found to
90 allocate the instance, the temporary error code
91 :pyeval:`errors.ECODE_TEMP_NORES` is returned. The operation can be
92 retried thereafter, with or without opportunistic locking.
93 - New experimental linux-ha resource scripts.
94 - Restricted-commands support: ganeti can now be asked (via command line
95 or rapi) to perform commands on a node. These are passed via ganeti
96 RPC rather than ssh. This functionality is restricted to commands
97 specified on the ``$sysconfdir/ganeti/restricted-commands`` for security
98 reasons. The file is not copied automatically.
104 - Diskless instances are now externally mirrored (Issue 237). This for
105 now has only been tested in conjunction with explicit target nodes for
107 - Queries not needing locks or RPC access to the node can now be
108 performed by the confd daemon, making them independent from jobs, and
109 thus faster to execute. This is selectable at configure time.
110 - The functionality for allocating multiple instances at once has been
111 overhauled and is now also available through :doc:`RAPI <rapi>`.
116 - Fix kvm compatibility with qemu 1.4 (Issue 389)
117 - Documentation updates (admin guide, upgrade notes, install
118 instructions) (Issue 372)
119 - Fix gnt-group list nodes and instances count (Issue 436)
120 - Fix compilation without non-mandatory libraries (Issue 441)
121 - Fix xen-hvm hypervisor forcing nics to type 'ioemu' (Issue 247)
122 - Make confd logging more verbose at INFO level (Issue 435)
123 - Improve "networks" documentation in :manpage:`gnt-instance(8)`
124 - Fix failure path for instance storage type conversion (Issue 229)
125 - Update htools text backend documentation
126 - Improve the renew-crypto section of :manpage:`gnt-cluster(8)`
127 - Disable inter-cluster instance move for file-based instances, because
128 it is dependant on instance export, which is not supported for
129 file-based instances. (Issue 414)
130 - Fix gnt-job crashes on non-ascii characters (Issue 427)
131 - Fix volume group checks on non-vm-capable nodes (Issue 432)
137 *(Released Mon, 22 Apr 2013)*
139 This was the third beta release of the 2.7 series. Since beta2:
141 - Fix hail to verify disk instance policies on a per-disk basis (Issue 418).
142 - Fix data loss on wrong usage of ``gnt-instance move``
143 - Properly export errors in confd-based job queries
144 - Add ``users-setup`` tool
145 - Fix iallocator protocol to report 0 as a disk size for diskless
146 instances. This avoids hail breaking when a diskless instance is
148 - Fix job queue directory permission problem that made confd job queries
149 fail. This requires running an ``ensure-dirs --full-run`` on upgrade
150 for access to archived jobs (Issue 406).
151 - Limit the sizes of networks supported by ``gnt-network`` to something
152 between a ``/16`` and a ``/30`` to prevent memory bloat and crashes.
153 - Fix bugs in instance disk template conversion
154 - Fix GHC 7 compatibility
155 - Fix ``burnin`` install path (Issue 426).
156 - Allow very small disk grows (Issue 347).
157 - Fix a ``ganeti-noded`` memory bloat introduced in 2.5, by making sure
158 that noded doesn't import masterd code (Issue 419).
159 - Make sure the default metavg at cluster init is the same as the vg, if
160 unspecified (Issue 358).
161 - Fix cleanup of partially created disks (part of Issue 416)
167 *(Released Tue, 2 Apr 2013)*
169 This was the second beta release of the 2.7 series. Since beta1:
171 - Networks no longer have a "type" slot, since this information was
172 unused in Ganeti: instead of it tags should be used.
173 - The rapi client now has a ``target_node`` option to MigrateInstance.
174 - Fix early exit return code for hbal (Issue 386).
175 - Fix ``gnt-instance migrate/failover -n`` (Issue 396).
176 - Fix ``rbd showmapped`` output parsing (Issue 312).
177 - Networks are now referenced indexed by UUID, rather than name. This
178 will require running cfgupgrade, from 2.7.0beta1, if networks are in
180 - The OS environment now includes network information.
181 - Deleting of a network is now disallowed if any instance nic is using
182 it, to prevent dangling references.
183 - External storage is now documented in man pages.
184 - The exclusive_storage flag can now only be set at nodegroup level.
185 - Hbal can now submit an explicit priority with its jobs.
186 - Many network related locking fixes.
187 - Bump up the required pylint version to 0.25.1.
188 - Fix the ``no_remember`` option in RAPI client.
189 - Many ipolicy related tests, qa, and fixes.
190 - Many documentation improvements and fixes.
191 - Fix building with ``--disable-file-storage``.
192 - Fix ``-q`` option in htools, which was broken if passed more than
194 - Some haskell/python interaction improvements and fixes.
195 - Fix iallocator in case of missing LVM storage.
196 - Fix confd config load in case of ``--no-lvm-storage``.
197 - The confd/query functionality is now mentioned in the security
204 *(Released Wed, 6 Feb 2013)*
206 This was the first beta release of the 2.7 series. All important changes
207 are listed in the latest 2.7 entry.
213 *(Released Fri, 21 Dec 2012)*
215 Important behaviour change: hbal won't rebalance anymore instances which
216 have the ``auto_balance`` attribute set to false. This was the intention
217 all along, but until now it only skipped those from the N+1 memory
218 reservation (DRBD-specific).
220 A significant number of bug fixes in this release:
222 - Fixed disk adoption interaction with ipolicy checks.
223 - Fixed networking issues when instances are started, stopped or
224 migrated, by forcing the tap device's MAC prefix to "fe" (issue 217).
225 - Fixed the warning in cluster verify for shared storage instances not
227 - Fixed removal of storage directory on shared file storage (issue 262).
228 - Fixed validation of LVM volume group name in OpClusterSetParams
229 (``gnt-cluster modify``) (issue 285).
230 - Fixed runtime memory increases (``gnt-instance modify -m``).
231 - Fixed live migration under Xen's ``xl`` mode.
232 - Fixed ``gnt-instance console`` with ``xl``.
233 - Fixed building with newer Haskell compiler/libraries.
234 - Fixed PID file writing in Haskell daemons (confd); this prevents
235 restart issues if confd was launched manually (outside of
236 ``daemon-util``) while another copy of it was running
237 - Fixed a type error when doing live migrations with KVM (issue 297) and
238 the error messages for failing migrations have been improved.
239 - Fixed opcode validation for the out-of-band commands (``gnt-node
241 - Fixed a type error when unsetting OS hypervisor parameters (issue
242 311); now it's possible to unset all OS-specific hypervisor
244 - Fixed the ``dry-run`` mode for many operations: verification of
245 results was over-zealous but didn't take into account the ``dry-run``
246 operation, resulting in "wrong" failures.
247 - Fixed bash completion in ``gnt-job list`` when the job queue has
248 hundreds of entries; especially with older ``bash`` versions, this
249 results in significant CPU usage.
251 And lastly, a few other improvements have been made:
253 - Added option to force master-failover without voting (issue 282).
254 - Clarified error message on lock conflict (issue 287).
255 - Logging of newly submitted jobs has been improved (issue 290).
256 - Hostname checks have been made uniform between instance rename and
258 - The ``--submit`` option is now supported by ``gnt-debug delay``.
259 - Shutting down the master daemon by sending SIGTERM now stops it from
260 processing jobs waiting for locks; instead, those jobs will be started
261 once again after the master daemon is started the next time (issue
263 - Support for Xen's ``xl`` program has been improved (besides the fixes
265 - Reduced logging noise in the Haskell confd daemon (only show one log
266 entry for each config reload, instead of two).
267 - Several man page updates and typo fixes.
273 *(Released Fri, 12 Oct 2012)*
275 A small bugfix release. Among the bugs fixed:
277 - Fixed double use of ``PRIORITY_OPT`` in ``gnt-node migrate``, that
278 made the command unusable.
279 - Commands that issue many jobs don't fail anymore just because some jobs
280 take so long that other jobs are archived.
281 - Failures during ``gnt-instance reinstall`` are reflected by the exit
283 - Issue 190 fixed. Check for DRBD in cluster verify is enabled only when
285 - When ``always_failover`` is set, ``--allow-failover`` is not required
286 in migrate commands anymore.
287 - ``bash_completion`` works even if extglob is disabled.
288 - Fixed bug with locks that made failover for RDB-based instances fail.
289 - Fixed bug in non-mirrored instance allocation that made Ganeti choose
290 a random node instead of one based on the allocator metric.
291 - Support for newer versions of pylint and pep8.
292 - Hail doesn't fail anymore when trying to add an instance of type
293 ``file``, ``sharedfile`` or ``rbd``.
294 - Added new Makefile target to rebuild the whole distribution, so that
295 all files are included.
301 *(Released Fri, 27 Jul 2012)*
304 .. attention:: The ``LUXI`` protocol has been made more consistent
305 regarding its handling of command arguments. This, however, leads to
306 incompatibility issues with previous versions. Please ensure that you
307 restart Ganeti daemons soon after the upgrade, otherwise most
308 ``LUXI`` calls (job submission, setting/resetting the drain flag,
309 pausing/resuming the watcher, cancelling and archiving jobs, querying
310 the cluster configuration) will fail.
319 The current ``admin_up`` field, which used to denote whether an instance
320 should be running or not, has been removed. Instead, ``admin_state`` is
321 introduced, with 3 possible values -- ``up``, ``down`` and ``offline``.
323 The rational behind this is that an instance being “down” can have
326 - it could be down during a reboot
327 - it could be temporarily be down for a reinstall
328 - or it could be down because it is deprecated and kept just for its
331 The previous Boolean state was making it difficult to do capacity
332 calculations: should Ganeti reserve memory for a down instance? Now, the
333 tri-state field makes it clear:
335 - in ``up`` and ``down`` state, all resources are reserved for the
336 instance, and it can be at any time brought up if it is down
337 - in ``offline`` state, only disk space is reserved for it, but not
340 The field can have an extra use: since the transition between ``up`` and
341 ``down`` and vice-versus is done via ``gnt-instance start/stop``, but
342 transition between ``offline`` and ``down`` is done via ``gnt-instance
343 modify``, it is possible to given different rights to users. For
344 example, owners of an instance could be allowed to start/stop it, but
345 not transition it out of the offline state.
347 Instance policies and specs
348 +++++++++++++++++++++++++++
350 In previous Ganeti versions, an instance creation request was not
351 limited on the minimum size and on the maximum size just by the cluster
352 resources. As such, any policy could be implemented only in third-party
353 clients (RAPI clients, or shell wrappers over ``gnt-*``
354 tools). Furthermore, calculating cluster capacity via ``hspace`` again
355 required external input with regards to instance sizes.
357 In order to improve these workflows and to allow for example better
358 per-node group differentiation, we introduced instance specs, which
361 - minimum instance disk size, disk count, memory size, cpu count
362 - maximum values for the above metrics
363 - and “standard” values (used in ``hspace`` to calculate the standard
366 The minimum/maximum values can be also customised at node-group level,
367 for example allowing more powerful hardware to support bigger instance
370 Beside the instance specs, there are a few other settings belonging to
371 the instance policy framework. It is possible now to customise, per
372 cluster and node-group:
374 - the list of allowed disk templates
375 - the maximum ratio of VCPUs per PCPUs (to control CPU oversubscription)
376 - the maximum ratio of instance to spindles (see below for more
377 information) for local storage
379 All these together should allow all tools that talk to Ganeti to know
380 what are the ranges of allowed values for instances and the
381 over-subscription that is allowed.
383 For the VCPU/PCPU ratio, we already have the VCPU configuration from the
384 instance configuration, and the physical CPU configuration from the
385 node. For the spindle ratios however, we didn't track before these
386 values, so new parameters have been added:
388 - a new node parameter ``spindle_count``, defaults to 1, customisable at
389 node group or node level
390 - at new backend parameter (for instances), ``spindle_use`` defaults to 1
392 Note that spindles in this context doesn't need to mean actual
393 mechanical hard-drives; it's just a relative number for both the node
394 I/O capacity and instance I/O consumption.
396 Instance migration behaviour
397 ++++++++++++++++++++++++++++
399 While live-migration is in general desirable over failover, it is
400 possible that for some workloads it is actually worse, due to the
401 variable time of the “suspend” phase during live migration.
403 To allow the tools to work consistently over such instances (without
404 having to hard-code instance names), a new backend parameter
405 ``always_failover`` has been added to control the migration/failover
406 behaviour. When set to True, all migration requests for an instance will
407 instead fall-back to failover.
409 Instance memory ballooning
410 ++++++++++++++++++++++++++
412 Initial support for memory ballooning has been added. The memory for an
413 instance is no longer fixed (backend parameter ``memory``), but instead
414 can vary between minimum and maximum values (backend parameters
415 ``minmem`` and ``maxmem``). Currently we only change an instance's
418 - live migrating or failing over and instance and the target node
419 doesn't have enough memory
420 - user requests changing the memory via ``gnt-instance modify
426 In order to control the use of specific CPUs by instance, support for
427 controlling CPU pinning has been added for the Xen, HVM and LXC
428 hypervisors. This is controlled by a new hypervisor parameter
429 ``cpu_mask``; details about possible values for this are in the
430 :manpage:`gnt-instance(8)`. Note that use of the most specific (precise
431 VCPU-to-CPU mapping) form will work well only when all nodes in your
432 cluster have the same amount of CPUs.
437 Another area in which Ganeti was not customisable were the parameters
438 used for storage configuration, e.g. how many stripes to use for LVM,
439 DRBD resync configuration, etc.
441 To improve this area, we've added disks parameters, which are
442 customisable at cluster and node group level, and which allow to
443 specify various parameters for disks (DRBD has the most parameters
444 currently), for example:
446 - DRBD resync algorithm and parameters (e.g. speed)
447 - the default VG for meta-data volumes for DRBD
448 - number of stripes for LVM (plain disk template)
451 These parameters can be modified via ``gnt-cluster modify -D …`` and
452 ``gnt-group modify -D …``, and are used at either instance creation (in
453 case of LVM stripes, for example) or at disk “activation” time
456 Rados block device support
457 ++++++++++++++++++++++++++
459 A Rados (http://ceph.com/wiki/Rbd) storage backend has been added,
460 denoted by the ``rbd`` disk template type. This is considered
461 experimental, feedback is welcome. For details on configuring it, see
462 the :doc:`install` document and the :manpage:`gnt-cluster(8)` man page.
467 The existing master IP functionality works well only in simple setups (a
468 single network shared by all nodes); however, if nodes belong to
469 different networks, then the ``/32`` setup and lack of routing
470 information is not enough.
472 To allow the master IP to function well in more complex cases, the
473 system was reworked as follows:
475 - a master IP netmask setting has been added
476 - the master IP activation/turn-down code was moved from the node daemon
478 - whether to run the Ganeti-supplied master IP script or a user-supplied
479 on is a ``gnt-cluster init`` setting
481 Details about the location of the standard and custom setup scripts are
482 in the man page :manpage:`gnt-cluster(8)`; for information about the
483 setup script protocol, look at the Ganeti-supplied script.
488 The `SPICE <http://www.linux-kvm.org/page/SPICE>`_ support has been
491 It is now possible to use TLS-protected connections, and when renewing
492 or changing the cluster certificates (via ``gnt-cluster renew-crypto``,
493 it is now possible to specify spice or spice CA certificates. Also, it
494 is possible to configure a password for SPICE sessions via the
495 hypervisor parameter ``spice_password_file``.
497 There are also new parameters to control the compression and streaming
498 options (e.g. ``spice_image_compression``, ``spice_streaming_video``,
499 etc.). For details, see the man page :manpage:`gnt-instance(8)` and look
500 for the spice parameters.
502 Lastly, it is now possible to see the SPICE connection information via
503 ``gnt-instance console``.
508 A new tool (``tools/ovfconverter``) has been added that supports
509 conversion between Ganeti and the `Open Virtualization Format
510 <http://en.wikipedia.org/wiki/Open_Virtualization_Format>`_ (both to and
513 This relies on the ``qemu-img`` tool to convert the disk formats, so the
514 actual compatibility with other virtualization solutions depends on it.
519 The configuration query daemon (``ganeti-confd``) is now optional, and
520 has been rewritten in Haskell; whether to use the daemon at all, use the
521 Python (default) or the Haskell version is selectable at configure time
522 via the ``--enable-confd`` parameter, which can take one of the
523 ``haskell``, ``python`` or ``no`` values. If not used, disabling the
524 daemon will result in a smaller footprint; for larger systems, we
525 welcome feedback on the Haskell version which might become the default
528 If you want to use ``gnt-node list-drbd`` you need to have the Haskell
529 daemon running. The Python version doesn't implement the new call.
532 User interface changes
533 ~~~~~~~~~~~~~~~~~~~~~~
535 We have replaced the ``--disks`` option of ``gnt-instance
536 replace-disks`` with a more flexible ``--disk`` option, which allows
537 adding and removing disks at arbitrary indices (Issue 188). Furthermore,
538 disk size and mode can be changed upon recreation (via ``gnt-instance
539 recreate-disks``, which accepts the same ``--disk`` option).
541 As many people are used to a ``show`` command, we have added that as an
542 alias to ``info`` on all ``gnt-*`` commands.
544 The ``gnt-instance grow-disk`` command has a new mode in which it can
545 accept the target size of the disk, instead of the delta; this can be
546 more safe since two runs in absolute mode will be idempotent, and
547 sometimes it's also easier to specify the desired size directly.
549 Also the handling of instances with regard to offline secondaries has
550 been improved. Instance operations should not fail because one of it's
551 secondary nodes is offline, even though it's safe to proceed.
553 A new command ``list-drbd`` has been added to the ``gnt-node`` script to
554 support debugging of DRBD issues on nodes. It provides a mapping of DRBD
555 minors to instance name.
560 RAPI coverage has improved, with (for example) new resources for
561 recreate-disks, node power-cycle, etc.
566 There is partial support for ``xl`` in the Xen hypervisor; feedback is
569 Python 2.7 is better supported, and after Ganeti 2.6 we will investigate
570 whether to still support Python 2.4 or move to Python 2.6 as minimum
573 Support for Fedora has been slightly improved; the provided example
574 init.d script should work better on it and the INSTALL file should
575 document the needed dependencies.
580 The deprecated ``QueryLocks`` LUXI request has been removed. Use
581 ``Query(what=QR_LOCK, ...)`` instead.
583 The LUXI requests :pyeval:`luxi.REQ_QUERY_JOBS`,
584 :pyeval:`luxi.REQ_QUERY_INSTANCES`, :pyeval:`luxi.REQ_QUERY_NODES`,
585 :pyeval:`luxi.REQ_QUERY_GROUPS`, :pyeval:`luxi.REQ_QUERY_EXPORTS` and
586 :pyeval:`luxi.REQ_QUERY_TAGS` are deprecated and will be removed in a
587 future version. :pyeval:`luxi.REQ_QUERY` should be used instead.
589 RAPI client: ``CertificateError`` now derives from
590 ``GanetiApiError``. This should make it more easy to handle Ganeti
593 Deprecation warnings due to PyCrypto/paramiko import in
594 ``tools/setup-ssh`` have been silenced, as usually they are safe; please
595 make sure to run an up-to-date paramiko version, if you use this tool.
597 The QA scripts now depend on Python 2.5 or above (the main code base
598 still works with Python 2.4).
600 The configuration file (``config.data``) is now written without
601 indentation for performance reasons; if you want to edit it, it can be
602 re-formatted via ``tools/fmtjson``.
604 A number of bugs has been fixed in the cluster merge tool.
606 ``x509`` certification verification (used in import-export) has been
607 changed to allow the same clock skew as permitted by the cluster
608 verification. This will remove some rare but hard to diagnose errors in
615 *(Released Thu, 19 Jul 2012)*
617 Very few changes from rc4 to the final release, only bugfixes:
619 - integrated fixes from release 2.5.2 (fix general boot flag for KVM
620 instance, fix CDROM booting for KVM instances)
621 - fixed node group modification of node parameters
622 - fixed issue in LUClusterVerifyGroup with multi-group clusters
623 - fixed generation of bash completion to ensure a stable ordering
630 *(Released Fri, 13 Jul 2012)*
632 Third release candidate for 2.6. The following changes were done from
635 - Fixed ``UpgradeConfig`` w.r.t. to disk parameters on disk objects.
636 - Fixed an inconsistency in the LUXI protocol with the provided
637 arguments (NOT backwards compatible)
638 - Fixed a bug with node groups ipolicy where ``min`` was greater than
639 the cluster ``std`` value
640 - Implemented a new ``gnt-node list-drbd`` call to list DRBD minors for
641 easier instance debugging on nodes (requires ``hconfd`` to work)
647 *(Released Tue, 03 Jul 2012)*
649 Second release candidate for 2.6. The following changes were done from
652 - Fixed ``gnt-cluster verify`` regarding ``master-ip-script`` on non
654 - Fixed a RAPI regression on missing beparams/memory
655 - Fixed redistribution of files on offline nodes
656 - Added possibility to run activate-disks even though secondaries are
657 offline. With this change it relaxes also the strictness on some other
658 commands which use activate disks internally:
659 * ``gnt-instance start|reboot|rename|backup|export``
660 - Made it possible to remove safely an instance if its secondaries are
662 - Made it possible to reinstall even though secondaries are offline
668 *(Released Mon, 25 Jun 2012)*
670 First release candidate for 2.6. The following changes were done from
673 - Fixed bugs with disk parameters and ``rbd`` templates as well as
675 - Made ``gnt-instance modify`` more consistent regarding new NIC/Disk
676 behaviour. It supports now the modify operation
677 - ``hcheck`` implemented to analyze cluster health and possibility of
678 improving health by rebalance
679 - ``hbal`` has been improved in dealing with split instances
685 *(Released Mon, 11 Jun 2012)*
687 Second beta release of 2.6. The following changes were done from beta2
690 - Fixed ``daemon-util`` with non-root user models
691 - Fixed creation of plain instances with ``--no-wait-for-sync``
692 - Fix wrong iv_names when running ``cfgupgrade``
693 - Export more information in RAPI group queries
694 - Fixed bug when changing instance network interfaces
695 - Extended burnin to do NIC changes
696 - query: Added ``<``, ``>``, ``<=``, ``>=`` comparison operators
697 - Changed default for DRBD barriers
698 - Fixed DRBD error reporting for syncer rate
699 - Verify the options on disk parameters
701 And of course various fixes to documentation and improved unittests and
708 *(Released Wed, 23 May 2012)*
710 First beta release of 2.6. The following changes were done from beta1 to
713 - integrated patch for distributions without ``start-stop-daemon``
714 - adapted example init.d script to work on Fedora
715 - fixed log handling in Haskell daemons
716 - adapted checks in the watcher for pycurl linked against libnss
717 - add partial support for ``xl`` instead of ``xm`` for Xen
718 - fixed a type issue in cluster verification
719 - fixed ssconf handling in the Haskell code (was breaking confd in IPv6
722 Plus integrated fixes from the 2.5 branch:
724 - fixed ``kvm-ifup`` to use ``/bin/bash``
725 - fixed parallel build failures
726 - KVM live migration when using a custom keymap
732 *(Released Tue, 24 Jul 2012)*
734 A small bugfix release, with no new features:
736 - fixed bash-isms in kvm-ifup, for compatibility with systems which use a
737 different default shell (e.g. Debian, Ubuntu)
738 - fixed KVM startup and live migration with a custom keymap (fixes Issue
739 243 and Debian bug #650664)
740 - fixed compatibility with KVM versions that don't support multiple boot
741 devices (fixes Issue 230 and Debian bug #624256)
743 Additionally, a few fixes were done to the build system (fixed parallel
744 build failures) and to the unittests (fixed race condition in test for
745 FileID functions, and the default enable/disable mode for QA test is now
752 *(Released Fri, 11 May 2012)*
754 A small bugfix release.
756 The main issues solved are on the topic of compatibility with newer LVM
759 - fixed parsing of ``lv_attr`` field
760 - adapted to new ``vgreduce --removemissing`` behaviour where sometimes
761 the ``--force`` flag is needed
763 Also on the topic of compatibility, ``tools/lvmstrap`` has been changed
764 to accept kernel 3.x too (was hardcoded to 2.6.*).
766 A regression present in 2.5.0 that broke handling (in the gnt-* scripts)
767 of hook results and that also made display of other errors suboptimal
768 was fixed; the code behaves now like 2.4 and earlier.
770 Another change in 2.5, the cleanup of the OS scripts environment, is too
771 aggressive: it removed even the ``PATH`` variable, which requires the OS
772 scripts to *always* need to export it. Since this is a bit too strict,
773 we now export a minimal PATH, the same that we export for hooks.
775 The fix for issue 201 (Preserve bridge MTU in KVM ifup script) was
776 integrated into this release.
778 Finally, a few other miscellaneous changes were done (no new features,
779 just small improvements):
781 - Fix ``gnt-group --help`` display
782 - Fix hardcoded Xen kernel path
783 - Fix grow-disk handling of invalid units
784 - Update synopsis for ``gnt-cluster repair-disk-sizes``
785 - Accept both PUT and POST in noded (makes future upgrade to 2.6 easier)
791 *(Released Thu, 12 Apr 2012)*
793 Incompatible/important changes and bugfixes
794 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
796 - The default of the ``/2/instances/[instance_name]/rename`` RAPI
797 resource's ``ip_check`` parameter changed from ``True`` to ``False``
798 to match the underlying LUXI interface.
799 - The ``/2/nodes/[node_name]/evacuate`` RAPI resource was changed to use
800 body parameters, see :doc:`RAPI documentation <rapi>`. The server does
801 not maintain backwards-compatibility as the underlying operation
802 changed in an incompatible way. The RAPI client can talk to old
803 servers, but it needs to be told so as the return value changed.
804 - When creating file-based instances via RAPI, the ``file_driver``
805 parameter no longer defaults to ``loop`` and must be specified.
806 - The deprecated ``bridge`` NIC parameter is no longer supported. Use
808 - Support for the undocumented and deprecated RAPI instance creation
809 request format version 0 has been dropped. Use version 1, supported
810 since Ganeti 2.1.3 and :doc:`documented <rapi>`, instead.
811 - Pyparsing 1.4.6 or above is required, see :doc:`installation
812 documentation <install>`.
813 - The "cluster-verify" hooks are now executed per group by the
814 ``OP_CLUSTER_VERIFY_GROUP`` opcode. This maintains the same behavior
815 if you just run ``gnt-cluster verify``, which generates one opcode per
817 - The environment as passed to the OS scripts is cleared, and thus no
818 environment variables defined in the node daemon's environment will be
819 inherited by the scripts.
820 - The :doc:`iallocator <iallocator>` mode ``multi-evacuate`` has been
822 - :doc:`New iallocator modes <design-multi-reloc>` have been added to
823 support operations involving multiple node groups.
824 - Offline nodes are ignored when failing over an instance.
825 - Support for KVM version 1.0, which changed the version reporting format
827 - TCP/IP ports used by DRBD disks are returned to a pool upon instance
829 - ``Makefile`` is now compatible with Automake 1.11.2
830 - Includes all bugfixes made in the 2.4 series
835 - The ganeti-htools project has been merged into the ganeti-core source
836 tree and will be built as part of Ganeti (see :doc:`install-quick`).
837 - Implemented support for :doc:`shared storage <design-shared-storage>`.
838 - Add support for disks larger than 2 TB in ``lvmstrap`` by supporting
839 GPT-style partition tables (requires `parted
840 <http://www.gnu.org/s/parted/>`_).
841 - Added support for floppy drive and 2nd CD-ROM drive in KVM hypervisor.
842 - Allowed adding tags on instance creation.
843 - Export instance tags to hooks (``INSTANCE_TAGS``, see :doc:`hooks`)
844 - Allow instances to be started in a paused state, enabling the user to
845 see the complete console output on boot using the console.
846 - Added new hypervisor flag to control default reboot behaviour
847 (``reboot_behavior``).
848 - Added support for KVM keymaps (hypervisor parameter ``keymap``).
849 - Improved out-of-band management support:
851 - Added ``gnt-node health`` command reporting the health status of
853 - Added ``gnt-node power`` command to manage power status of nodes.
854 - Added command for emergency power-off (EPO), ``gnt-cluster epo``.
856 - Instance migration can fall back to failover if instance is not
858 - Filters can be used when listing nodes, instances, groups and locks;
859 see :manpage:`ganeti(7)` manpage.
860 - Added post-execution status as variables to :doc:`hooks <hooks>`
862 - Instance tags are exported/imported together with the instance.
863 - When given an explicit job ID, ``gnt-job info`` will work for archived
865 - Jobs can define dependencies on other jobs (not yet supported via
866 RAPI or command line, but used by internal commands and usable via
869 - Lock monitor (``gnt-debug locks``) shows jobs waiting for
872 - Instance failover is now available as a RAPI resource
873 (``/2/instances/[instance_name]/failover``).
874 - ``gnt-instance info`` defaults to static information if primary node
876 - Opcodes have a new ``comment`` attribute.
877 - Added basic SPICE support to KVM hypervisor.
878 - ``tools/ganeti-listrunner`` allows passing of arguments to executable.
880 Node group improvements
881 ~~~~~~~~~~~~~~~~~~~~~~~
883 - ``gnt-cluster verify`` has been modified to check groups separately,
884 thereby improving performance.
885 - Node group support has been added to ``gnt-cluster verify-disks``,
886 which now operates per node group.
887 - Watcher has been changed to work better with node groups.
889 - One process and state file per node group.
890 - Slow watcher in one group doesn't block other group's watcher.
892 - Added new command, ``gnt-group evacuate``, to move all instances in a
893 node group to other groups.
894 - Added ``gnt-instance change-group`` to move an instance to another
896 - ``gnt-cluster command`` and ``gnt-cluster copyfile`` now support
897 per-group operations.
898 - Node groups can be tagged.
899 - Some operations switch from an exclusive to a shared lock as soon as
901 - Instance's primary and secondary nodes' groups are now available as
902 query fields (``pnode.group``, ``pnode.group.uuid``, ``snodes.group``
903 and ``snodes.group.uuid``).
908 - Numerous updates to documentation and manpages.
910 - :doc:`RAPI <rapi>` documentation now has detailed parameter
912 - Some opcode/job results are now also documented, see :doc:`RAPI
915 - A lockset's internal lock is now also visible in lock monitor.
916 - Log messages from job queue workers now contain information about the
917 opcode they're processing.
918 - ``gnt-instance console`` no longer requires the instance lock.
919 - A short delay when waiting for job changes reduces the number of LUXI
920 requests significantly.
921 - DRBD metadata volumes are overwritten with zeros during disk creation.
922 - Out-of-band commands no longer acquire the cluster lock in exclusive
924 - ``devel/upload`` now uses correct permissions for directories.
930 *(Released Fri, 23 Mar 2012)*
932 This was the sixth release candidate of the 2.5 series.
938 *(Released Mon, 9 Jan 2012)*
940 This was the fifth release candidate of the 2.5 series.
946 *(Released Thu, 27 Oct 2011)*
948 This was the fourth release candidate of the 2.5 series.
954 *(Released Wed, 26 Oct 2011)*
956 This was the third release candidate of the 2.5 series.
962 *(Released Tue, 18 Oct 2011)*
964 This was the second release candidate of the 2.5 series.
970 *(Released Tue, 4 Oct 2011)*
972 This was the first release candidate of the 2.5 series.
978 *(Released Wed, 31 Aug 2011)*
980 This was the third beta release of the 2.5 series.
986 *(Released Mon, 22 Aug 2011)*
988 This was the second beta release of the 2.5 series.
994 *(Released Fri, 12 Aug 2011)*
996 This was the first beta release of the 2.5 series.
1002 *(Released Thu, 27 Oct 2011)*
1004 - Fixed bug when parsing command line parameter values ending in
1006 - Fixed assertion error after unclean master shutdown
1007 - Disable HTTP client pool for RPC, significantly reducing memory usage
1009 - Fixed queue archive creation with wrong permissions
1015 *(Released Tue, 23 Aug 2011)*
1019 - Fixed documentation for importing with ``--src-dir`` option
1020 - Fixed a bug in ``ensure-dirs`` with queue/archive permissions
1021 - Fixed a parsing issue with DRBD 8.3.11 in the Linux kernel
1027 *(Released Fri, 5 Aug 2011)*
1029 Many bug-fixes and a few small features:
1031 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
1032 issues when you tried to add an instance with two MAC addresses in one
1034 - KVM: fixed per-instance stored UID value
1035 - KVM: configure bridged NICs at migration start
1036 - KVM: Fix a bug where instance will not start with never KVM versions
1038 - Added OS search path to ``gnt-cluster info``
1039 - Fixed an issue with ``file_storage_dir`` where you were forced to
1040 provide an absolute path, but the documentation states it is a
1041 relative path, the documentation was right
1042 - Added a new parameter to instance stop/start called ``--no-remember``
1043 that will make the state change to not be remembered
1044 - Implemented ``no_remember`` at RAPI level
1045 - Improved the documentation
1046 - Node evacuation: don't call IAllocator if node is already empty
1047 - Fixed bug in DRBD8 replace disks on current nodes
1048 - Fixed bug in recreate-disks for DRBD instances
1049 - Moved assertion checking locks in ``gnt-instance replace-disks``
1050 causing it to abort with not owning the right locks for some situation
1051 - Job queue: Fixed potential race condition when cancelling queued jobs
1052 - Fixed off-by-one bug in job serial generation
1053 - ``gnt-node volumes``: Fix instance names
1054 - Fixed aliases in bash completion
1055 - Fixed a bug in reopening log files after being sent a SIGHUP
1056 - Added a flag to burnin to allow specifying VCPU count
1057 - Bugfixes to non-root Ganeti configuration
1063 *(Released Thu, 12 May 2011)*
1065 Many bug-fixes and a few new small features:
1067 - Fixed a bug related to log opening failures
1068 - Fixed a bug in instance listing with orphan instances
1069 - Fixed a bug which prevented resetting the cluster-level node parameter
1070 ``oob_program`` to the default
1071 - Many fixes related to the ``cluster-merge`` tool
1072 - Fixed a race condition in the lock monitor, which caused failures
1073 during (at least) creation of many instances in parallel
1074 - Improved output for gnt-job info
1075 - Removed the quiet flag on some ssh calls which prevented debugging
1077 - Improved the N+1 failure messages in cluster verify by actually
1078 showing the memory values (needed and available)
1079 - Increased lock attempt timeouts so that when executing long operations
1080 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
1081 too early and thus prevent the use of the 'fair' mechanism
1082 - Changed instance query data (``gnt-instance info``) to not acquire
1083 locks unless needed, thus allowing its use on locked instance if only
1084 static information is asked for
1085 - Improved behaviour with filesystems that do not support rename on an
1087 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
1088 kept locks on all nodes during the wipe, which is unneeded
1089 - Fixed ``gnt-watcher`` handling of errors during hooks execution
1090 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
1091 10GiB) which caused the wipe to fail right at the end in some cases
1092 - Fixed master IP activation when doing master failover with no-voting
1093 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
1094 the master node itself
1095 - Fixed potential data-loss in under disk full conditions, where Ganeti
1096 wouldn't check correctly the return code and would consider
1097 partially-written files 'correct'
1098 - Fixed bug related to multiple VGs and DRBD disk replacing
1099 - Added new disk parameter ``metavg`` that allows placement of the meta
1100 device for DRBD in a different volume group
1101 - Fixed error handling in the node daemon when the system libc doesn't
1102 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
1103 - Fixed lock release during replace-disks, which kept cluster-wide locks
1104 when doing disk replaces with an iallocator script
1105 - Added check for missing bridges in cluster verify
1106 - Handle EPIPE errors while writing to the terminal better, so that
1107 piping the output to e.g. ``less`` doesn't cause a backtrace
1108 - Fixed rare case where a ^C during Luxi calls could have been
1109 interpreted as server errors, instead of simply terminating
1110 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
1112 - Added a few more parameters to the KVM hypervisor, allowing a second
1113 CDROM, custom disk type for CDROMs and a floppy image
1114 - Removed redundant message in instance rename when the name is given
1116 - Added option to ``gnt-instance recreate-disks`` to allow creating the
1117 disks on new nodes, allowing recreation when the original instance
1118 nodes are completely gone
1119 - Added option when converting disk templates to DRBD to skip waiting
1120 for the resync, in order to make the instance available sooner
1121 - Added two new variables to the OS scripts environment (containing the
1123 - Made the root_path and optional parameter for the xen-pvm hypervisor,
1124 to allow use of ``pvgrub`` as bootloader
1125 - Changed the instance memory modifications to only check out-of-memory
1126 conditions on memory increases, and turned the secondary node warnings
1127 into errors (they can still be overridden via ``--force``)
1128 - Fixed the handling of a corner case when the Python installation gets
1129 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
1130 to execute a command that doesn't exist
1131 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
1132 node of the instance returned failures during instance shutdown; this
1133 adds the option ``--ignore-consistency`` to gnt-instance move
1135 And as usual, various improvements to the error messages, documentation
1142 *(Released Wed, 09 Mar 2011)*
1144 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
1145 the RAPI users file if run twice (even with ``--dry-run``).
1147 The release fixes that bug (nothing else changed).
1153 *(Released Mon, 07 Mar 2011)*
1155 Final 2.4.0 release. Just a few small fixes:
1157 - Fixed RAPI node evacuate
1158 - Fixed the kvm-ifup script
1159 - Fixed internal error handling for special job cases
1160 - Updated man page to specify the escaping feature for options
1166 *(Released Mon, 28 Feb 2011)*
1168 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
1169 that this feature wiped the disks of the wrong instance, leading to loss
1174 - Fixed title of query field containing instance name
1175 - Expanded the glossary in the documentation
1176 - Fixed one unittest (internal issue)
1182 *(Released Mon, 21 Feb 2011)*
1184 A number of bug fixes plus just a couple functionality changes.
1186 On the user-visible side, the ``gnt-* list`` command output has changed
1187 with respect to "special" field states. The current rc1 style of display
1188 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
1189 the default output mode special fields states are displayed as follows:
1191 - Offline resource: ``*``
1192 - Unavailable/not applicable: ``-``
1193 - Data missing (RPC failure): ``?``
1194 - Unknown field: ``??``
1196 Another user-visible change is the addition of ``--force-join`` to
1201 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
1202 - Fixed regression in RAPI/instance reinstall where all parameters were
1203 required (instead of optional)
1204 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
1205 - Fixed iallocator usage (offline nodes were not considered offline)
1206 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
1207 - Fixed hypervisor and OS parameter validation with respect to
1208 non-vm_capable nodes
1209 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
1211 - Fixed ``tools/listrunner`` with respect to agent-based usage
1217 *(Released Fri, 4 Feb 2011)*
1219 Many changes and fixes since the beta1 release. While there were some
1220 internal changes, the code has been mostly stabilised for the RC
1223 Note: the dumb allocator was removed in this release, as it was not kept
1224 up-to-date with the IAllocator protocol changes. It is recommended to
1225 use the ``hail`` command from the ganeti-htools package.
1227 Note: the 2.4 and up versions of Ganeti are not compatible with the
1228 0.2.x branch of ganeti-htools. You need to upgrade to
1229 ganeti-htools-0.3.0 (or later).
1231 Regressions fixed from 2.3
1232 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1234 - Fixed the ``gnt-cluster verify-disks`` command
1235 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
1237 - Fixed disk adoption breakage
1238 - Fixed wrong headers in instance listing for field aliases
1243 - Fixed corner case in KVM handling of NICs
1244 - Fixed many cases of wrong handling of non-vm_capable nodes
1245 - Fixed a bug where a missing instance symlink was not possible to
1246 recreate with any ``gnt-*`` command (now ``gnt-instance
1247 activate-disks`` does it)
1248 - Fixed the volume group name as reported by ``gnt-cluster
1250 - Increased timeouts for the import-export code, hopefully leading to
1251 fewer aborts due network or instance timeouts
1252 - Fixed bug in ``gnt-node list-storage``
1253 - Fixed bug where not all daemons were started on cluster
1254 initialisation, but only at the first watcher run
1255 - Fixed many bugs in the OOB implementation
1256 - Fixed watcher behaviour in presence of instances with offline
1258 - Fixed instance list output for instances running on the wrong node
1259 - a few fixes to the cluster-merge tool, but it still cannot merge
1260 multi-node groups (currently it is not recommended to use this tool)
1266 - Improved network configuration for the KVM hypervisor
1267 - Added e1000 as a supported NIC for Xen-HVM
1268 - Improved the lvmstrap tool to also be able to use partitions, as
1269 opposed to full disks
1270 - Improved speed of disk wiping (the cluster parameter
1271 ``prealloc_wipe_disks``, so that it has a low impact on the total time
1272 of instance creations
1273 - Added documentation for the OS parameters
1274 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
1275 hypervisor is not responding
1276 - Added display of blacklisted and hidden OS information in
1277 ``gnt-cluster info``
1278 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
1279 NIC and node parameters, which might create problems with currently
1280 invalid (but undetected) configuration files, but prevents validation
1281 failures when unrelated parameters are modified
1282 - Changed cluster initialisation to wait for the master daemon to become
1284 - Expanded the RAPI interface:
1286 - Added config redistribution resource
1287 - Added activation/deactivation of instance disks
1288 - Added export of console information
1290 - Implemented log file reopening on SIGHUP, which allows using
1291 logrotate(8) for the Ganeti log files
1292 - Added a basic OOB helper script as an example
1298 *(Released Fri, 14 Jan 2011)*
1303 - Fixed timezone issues when formatting timestamps
1304 - Added support for node groups, available via ``gnt-group`` and other
1306 - Added out-of-band framework and management, see :doc:`design
1307 document <design-oob>`
1308 - Removed support for roman numbers from ``gnt-node list`` and
1309 ``gnt-instance list``.
1310 - Allowed modification of master network interface via ``gnt-cluster
1311 modify --master-netdev``
1312 - Accept offline secondaries while shutting down instance disks
1313 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
1314 - Added support for multiple LVM volume groups
1315 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
1317 - Added commands to list available fields:
1319 - ``gnt-node list-fields``
1320 - ``gnt-group list-fields``
1321 - ``gnt-instance list-fields``
1323 - Updated documentation and man pages
1328 - Moved ``rapi_users`` file into separate directory, now named
1329 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
1331 - Added new tool for running commands on many machines,
1332 ``tools/ganeti-listrunner``
1333 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
1334 improving the ``gnt-instance console`` output
1335 - Allowed customisation of disk index separator at ``configure`` time
1336 - Export node group allocation policy to :doc:`iallocator <iallocator>`
1337 - Added support for non-partitioned md disks in ``lvmstrap``
1338 - Added script to gracefully power off KVM instances
1339 - Split ``utils`` module into smaller parts
1340 - Changed query operations to return more detailed information, e.g.
1341 whether an information is unavailable due to an offline node. To use
1342 this new functionality, the LUXI call ``Query`` must be used. Field
1343 information is now stored by the master daemon and can be retrieved
1344 using ``QueryFields``. Instances, nodes and groups can also be queried
1345 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
1346 exposed via RAPI). The following commands make use of this
1347 infrastructure change:
1349 - ``gnt-group list``
1350 - ``gnt-group list-fields``
1352 - ``gnt-node list-fields``
1353 - ``gnt-instance list``
1354 - ``gnt-instance list-fields``
1355 - ``gnt-debug locks``
1360 - New RAPI resources (see :doc:`rapi`):
1364 - ``/2/groups/[group_name]``
1365 - ``/2/groups/[group_name]/assign-nodes``
1366 - ``/2/groups/[group_name]/modify``
1367 - ``/2/groups/[group_name]/rename``
1368 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
1372 - Implemented ``no_install`` for instance creation
1373 - Implemented OS parameters for instance reinstallation, allowing
1374 use of special settings on reinstallation (e.g. for preserving data)
1379 - Added IPv6 support in import/export
1380 - Pause DRBD synchronization while wiping disks on instance creation
1381 - Updated unittests and QA scripts
1382 - Improved network parameters passed to KVM
1383 - Converted man pages from docbook to reStructuredText
1389 *(Released Mon, 20 Dec 2010)*
1391 Released version 2.3.1~rc1 without any changes.
1397 *(Released Wed, 1 Dec 2010)*
1399 - impexpd: Disable OpenSSL compression in socat if possible (backport
1400 from master, commit e90739d625b, see :doc:`installation guide
1401 <install-quick>` for details)
1402 - Changed unittest coverage report to exclude test scripts
1403 - Added script to check version format
1409 *(Released Wed, 1 Dec 2010)*
1411 Released version 2.3.0~rc1 without any changes.
1417 *(Released Fri, 19 Nov 2010)*
1419 A number of bugfixes and documentation updates:
1421 - Update ganeti-os-interface documentation
1422 - Fixed a bug related to duplicate MACs or similar items which should be
1424 - Fix breakage in OS state modify
1425 - Reinstall instance: disallow offline secondaries (fixes bug related to
1426 OS changing but reinstall failing)
1427 - plus all the other fixes between 2.2.1 and 2.2.2
1433 *(Released Tue, 2 Nov 2010)*
1435 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
1436 - Fixed master failover race with watcher
1437 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
1439 - Accept previously stopped instance for export with instance removal
1440 - Simplify and extend the environment variables for instance OS scripts
1441 - Added new node flags, ``master_capable`` and ``vm_capable``
1442 - Added optional instance disk wiping prior during allocation. This is a
1443 cluster-wide option and can be set/modified using
1444 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
1445 - Added IPv6 support, see :doc:`design document <design-2.3>` and
1446 :doc:`install-quick`
1447 - Added a new watcher option (``--ignore-pause``)
1448 - Added option to ignore offline node on instance start/stop
1449 (``--ignore-offline``)
1450 - Allow overriding OS parameters with ``gnt-instance reinstall``
1451 - Added ability to change node's secondary IP address using ``gnt-node
1453 - Implemented privilege separation for all daemons except
1454 ``ganeti-noded``, see ``configure`` options
1455 - Complain if an instance's disk is marked faulty in ``gnt-cluster
1457 - Implemented job priorities (see ``ganeti(7)`` manpage)
1458 - Ignore failures while shutting down instances during failover from
1460 - Exit daemon's bootstrap process only once daemon is ready
1461 - Export more information via ``LUInstanceQuery``/remote API
1462 - Improved documentation, QA and unittests
1463 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
1464 restart if the file was created or changed
1465 - Added LUXI protocol version sent with each request and response,
1466 allowing detection of server/client mismatches
1467 - Moved the Python scripts among gnt-* and ganeti-* into modules
1468 - Moved all code related to setting up SSH to an external script,
1470 - Infrastructure changes for node group support in future versions
1476 *(Released Fri, 19 Nov 2010)*
1478 A few small bugs fixed, and some improvements to the build system:
1480 - Fix documentation regarding conversion to drbd
1481 - Fix validation of parameters in cluster modify (``gnt-cluster modify
1483 - Fix error handling in node modify with multiple changes
1484 - Allow remote imports without checked names
1490 *(Released Tue, 19 Oct 2010)*
1492 - Disable SSL session ID cache in RPC client
1498 *(Released Thu, 14 Oct 2010)*
1500 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
1501 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
1503 - Fix problems with interaction between Curl and Python's HTTP server,
1504 resulting in increased speed in many RPC calls
1505 - Improve our release script to prevent breakage with older aclocal and
1512 *(Released Thu, 7 Oct 2010)*
1514 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
1515 value retrieved from master
1516 - Added support for blacklisted or hidden OS definitions
1517 - Added simple lock monitor (accessible via (``gnt-debug locks``)
1518 - Added support for -mem-path in KVM hypervisor abstraction layer
1519 - Allow overriding instance parameters in tool for inter-cluster
1520 instance moves (``tools/move-instance``)
1521 - Improved opcode summaries (e.g. in ``gnt-job list``)
1522 - Improve consistency of OS listing by sorting it
1523 - Documentation updates
1529 *(Released Fri, 8 Oct 2010)*
1531 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
1537 *(Released Mon, 4 Oct 2010)*
1539 - Fixed regression in ``gnt-instance rename``
1545 *(Released Wed, 22 Sep 2010)*
1547 - Fixed OS_VARIANT variable for OS scripts
1548 - Fixed cluster tag operations via RAPI
1549 - Made ``setup-ssh`` exit with non-zero code if an error occurred
1550 - Disabled RAPI CA checks in watcher
1556 *(Released Mon, 23 Aug 2010)*
1558 - Support DRBD versions of the format "a.b.c.d"
1560 - Re-introduce support for usage from multiple threads in RAPI client
1561 - Instance renames and modify via RAPI
1562 - Work around race condition between processing and archival in job
1564 - Mark opcodes following failed one as failed, too
1565 - Job field ``lock_status`` was removed due to difficulties making it
1566 work with the changed job queue in Ganeti 2.2; a better way to monitor
1567 locks is expected for a later 2.2.x release
1568 - Fixed dry-run behaviour with many commands
1569 - Support ``ssh-agent`` again when adding nodes
1570 - Many additional bugfixes
1576 *(Released Fri, 30 Jul 2010)*
1578 Important change: the internal RPC mechanism between Ganeti nodes has
1579 changed from using a home-grown http library (based on the Python base
1580 libraries) to use the PycURL library. This requires that PycURL is
1581 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
1582 against GnuTLS by default. cURL's support for GnuTLS had known issues
1583 before cURL 7.21.0 and we recommend using the latest cURL release or
1584 linking against OpenSSL. Most other distributions already link PycURL
1585 and cURL against OpenSSL. The command::
1587 python -c 'import pycurl; print pycurl.version'
1589 can be used to determine the libraries PycURL and cURL are linked
1592 Other significant changes:
1594 - Rewrote much of the internals of the job queue, in order to achieve
1595 better parallelism; this decouples job query operations from the job
1596 processing, and it should allow much nicer behaviour of the master
1597 daemon under load, and it also has uncovered some long-standing bugs
1598 related to the job serialisation (now fixed)
1599 - Added a default iallocator setting to the cluster parameters,
1600 eliminating the need to always pass nodes or an iallocator for
1601 operations that require selection of new node(s)
1602 - Added experimental support for the LXC virtualization method
1603 - Added support for OS parameters, which allows the installation of
1604 instances to pass parameter to OS scripts in order to customise the
1606 - Added a hypervisor parameter controlling the migration type (live or
1607 non-live), since hypervisors have various levels of reliability; this
1608 has renamed the 'live' parameter to 'mode'
1609 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
1610 logical volumes, meaning that cluster verify will ignore them and not
1611 flag their presence as errors
1612 - The watcher will now reset the error count for failed instances after
1613 8 hours, thus allowing self-healing if the problem that caused the
1614 instances to be down/fail to start has cleared in the meantime
1615 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
1616 check for, and warn, if the drbd module parameter ``usermode_helper``
1617 is not consistent with the cluster-wide setting; this is needed to
1618 make diagnose easier of failed drbd creations
1619 - Started adding base IPv6 support, but this is not yet
1620 enabled/available for use
1621 - Rename operations (cluster, instance) will now return the new name,
1622 which is especially useful if a short name was passed in
1623 - Added support for instance migration in RAPI
1624 - Added a tool to pre-configure nodes for the SSH setup, before joining
1625 them to the cluster; this will allow in the future a simplified model
1626 for node joining (but not yet fully enabled in 2.2); this needs the
1627 paramiko python library
1628 - Fixed handling of name-resolving errors
1629 - Fixed consistency of job results on the error path
1630 - Fixed master-failover race condition when executed multiple times in
1632 - Fixed many bugs related to the job queue (mostly introduced during the
1633 2.2 development cycle, so not all are impacting 2.1)
1634 - Fixed instance migration with missing disk symlinks
1635 - Fixed handling of unknown jobs in ``gnt-job archive``
1636 - And many other small fixes/improvements
1640 - Enhanced both the unittest and the QA coverage
1641 - Switched the opcode validation to a generic model, and extended the
1642 validation to all opcode parameters
1643 - Changed more parts of the code that write shell scripts to use the
1645 - Switched the master daemon to use the asyncore library for the Luxi
1652 *(Released Thu, 17 Jun 2010)*
1654 - Added tool (``move-instance``) and infrastructure to move instances
1655 between separate clusters (see :doc:`separate documentation
1656 <move-instance>` and :doc:`design document <design-2.2>`)
1657 - Added per-request RPC timeout
1658 - RAPI now requires a Content-Type header for requests with a body (e.g.
1659 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
1660 :rfc:`2616` (HTTP/1.1), section 7.2.1)
1661 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
1663 - Implemented initial support for running Ganeti daemons as separate
1664 users, see configure-time flags ``--with-user-prefix`` and
1665 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
1667 - Instances can be removed after export (``gnt-backup export
1668 --remove-instance``)
1669 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
1670 key (instead of 1024 bit)
1671 - Added new cluster configuration file for cluster domain secret
1672 - Import/export now use SSL instead of SSH
1673 - Added support for showing estimated time when exporting an instance,
1674 see the ``ganeti-os-interface(7)`` manpage and look for
1681 *(Released Tue, 16 Nov 2010)*
1683 Some more bugfixes. Unless critical bugs occur, this will be the last
1686 - Fix case of MAC special-values
1687 - Fix mac checker regex
1688 - backend: Fix typo causing "out of range" error
1689 - Add missing --units in gnt-instance list man page
1695 *(Released Tue, 24 Aug 2010)*
1698 - Don't ignore secondary node silently on non-mirrored disk templates
1700 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
1701 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
1702 - Properly document the usb_mouse parameter
1703 - Fix path in ganeti-rapi(8) (issue 116)
1704 - Adjust error message when the ganeti user's .ssh directory is
1706 - Add same-node-check when changing the disk template to drbd
1712 *(Released Fri, 16 Jul 2010)*
1715 - Add an option to only select some reboot types during qa/burnin.
1716 (on some hypervisors consequent reboots are not supported)
1717 - Fix infrequent race condition in master failover. Sometimes the old
1718 master ip address would be still detected as up for a short time
1719 after it was removed, causing failover to fail.
1720 - Decrease mlockall warnings when the ctypes module is missing. On
1721 Python 2.4 we support running even if no ctypes module is installed,
1722 but we were too verbose about this issue.
1723 - Fix building on old distributions, on which man doesn't have a
1725 - Fix RAPI not to ignore the MAC address on instance creation
1726 - Implement the old instance creation format in the RAPI client.
1732 *(Released Thu, 01 Jul 2010)*
1734 A small bugfix release:
1735 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
1736 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
1737 a non-existing option
1738 - Fix instance create: the --force-variant option was ignored
1739 - Improve pylint 0.21 compatibility and warnings with Python 2.6
1740 - Fix modify node storage with non-FQDN arguments
1741 - Fix RAPI client to authenticate under Python 2.6 when used
1742 for more than 5 requests needing authentication
1743 - Fix gnt-instance modify -t (storage) giving a wrong error message
1744 when converting a non-shutdown drbd instance to plain
1750 *(Released Fri, 18 Jun 2010)*
1752 A small bugfix release:
1754 - Fix live migration of KVM instances started with older Ganeti
1755 versions which had fewer hypervisor parameters
1756 - Fix gnt-instance grow-disk on down instances
1757 - Fix an error-reporting bug during instance migration
1758 - Better checking of the ``--net`` and ``--disk`` values, to avoid
1759 silently ignoring broken ones
1760 - Fix an RPC error reporting bug affecting, for example, RAPI client
1762 - Fix bug triggered by different API version os-es on different nodes
1763 - Fix a bug in instance startup with custom hvparams: OS level
1764 parameters would fail to be applied.
1765 - Fix the RAPI client under Python 2.6 (but more work is needed to
1766 make it work completely well with OpenSSL)
1767 - Fix handling of errors when resolving names from DNS
1773 *(Released Thu, 3 Jun 2010)*
1775 A medium sized development cycle. Some new features, and some
1776 fixes/small improvements/cleanups.
1778 Significant features
1779 ~~~~~~~~~~~~~~~~~~~~
1781 The node deamon now tries to mlock itself into memory, unless the
1782 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
1783 its logs, and falls back to console logging. This allows emergency
1784 features such as ``gnt-node powercycle`` to work even in the event of a
1785 broken node disk (tested offlining the disk hosting the node's
1786 filesystem and dropping its memory caches; don't try this at home)
1788 KVM: add vhost-net acceleration support. It can be tested with a new
1789 enough version of the kernel and of qemu-kvm.
1791 KVM: Add instance chrooting feature. If you use privilege dropping for
1792 your VMs you can also now force them to chroot to an empty directory,
1793 before starting the emulated guest.
1795 KVM: Add maximum migration bandwith and maximum downtime tweaking
1796 support (requires a new-enough version of qemu-kvm).
1798 Cluster verify will now warn if the master node doesn't have the master
1799 ip configured on it.
1801 Add a new (incompatible) instance creation request format to RAPI which
1802 supports all parameters (previously only a subset was supported, and it
1803 wasn't possible to extend the old format to accomodate all the new
1804 features. The old format is still supported, and a client can check for
1805 this feature, before using it, by checking for its presence in the
1806 ``features`` RAPI resource.
1808 Now with ancient latin support. Try it passing the ``--roman`` option to
1809 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
1810 (requires the python-roman module to be installed, in order to work).
1815 As usual many internal code refactorings, documentation updates, and
1818 - Lots of improvements and cleanups to the experimental Remote API
1819 (RAPI) client library.
1820 - A new unit test suite for the core daemon libraries.
1821 - A fix to creating missing directories makes sure the umask is not
1822 applied anymore. This enforces the same directory permissions
1824 - Better handling terminating daemons with ctrl+c (used when running
1825 them in debugging mode).
1826 - Fix a race condition in live migrating a KVM instance, when stat()
1827 on the old proc status file returned EINVAL, which is an unexpected
1829 - Fixed manpage checking with newer man and utf-8 charachters. But now
1830 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
1836 *(Released Fri, 7 May 2010)*
1838 Fix a bug which prevented untagged KVM instances from starting.
1844 *(Released Fri, 7 May 2010)*
1846 Another release with a long development cycle, during which many
1847 different features were added.
1849 Significant features
1850 ~~~~~~~~~~~~~~~~~~~~
1852 The KVM hypervisor now can run the individual instances as non-root, to
1853 reduce the impact of a VM being hijacked due to bugs in the
1854 hypervisor. It is possible to run all instances as a single (non-root)
1855 user, to manually specify a user for each instance, or to dynamically
1856 allocate a user out of a cluster-wide pool to each instance, with the
1857 guarantee that no two instances will run under the same user ID on any
1860 An experimental RAPI client library, that can be used standalone
1861 (without the other Ganeti libraries), is provided in the source tree as
1862 ``lib/rapi/client.py``. Note this client might change its interface in
1863 the future, as we iterate on its capabilities.
1865 A new command, ``gnt-cluster renew-crypto`` has been added to easily
1866 replace the cluster's certificates and crypto keys. This might help in
1867 case they have been compromised, or have simply expired.
1869 A new disk option for instance creation has been added that allows one
1870 to "adopt" currently existing logical volumes, with data
1871 preservation. This should allow easier migration to Ganeti from
1872 unmanaged (or managed via other software) instances.
1874 Another disk improvement is the possibility to convert between redundant
1875 (DRBD) and plain (LVM) disk configuration for an instance. This should
1876 allow better scalability (starting with one node and growing the
1877 cluster, or shrinking a two-node cluster to one node).
1879 A new feature that could help with automated node failovers has been
1880 implemented: if a node sees itself as offline (by querying the master
1881 candidates), it will try to shutdown (hard) all instances and any active
1882 DRBD devices. This reduces the risk of duplicate instances if an
1883 external script automatically failovers the instances on such nodes. To
1884 enable this, the cluster parameter ``maintain_node_health`` should be
1885 enabled; in the future this option (per the name) will enable other
1886 automatic maintenance features.
1888 Instance export/import now will reuse the original instance
1889 specifications for all parameters; that means exporting an instance,
1890 deleting it and the importing it back should give an almost identical
1891 instance. Note that the default import behaviour has changed from
1892 before, where it created only one NIC; now it recreates the original
1895 Cluster verify has added a few new checks: SSL certificates validity,
1896 /etc/hosts consistency across the cluster, etc.
1901 As usual, many internal changes were done, documentation fixes,
1904 - Fixed cluster initialization with disabled cluster storage (regression
1905 introduced in 2.1.1)
1906 - File-based storage supports growing the disks
1907 - Fixed behaviour of node role changes
1908 - Fixed cluster verify for some corner cases, plus a general rewrite of
1909 cluster verify to allow future extension with more checks
1910 - Fixed log spamming by watcher and node daemon (regression introduced
1912 - Fixed possible validation issues when changing the list of enabled
1914 - Fixed cleanup of /etc/hosts during node removal
1915 - Fixed RAPI response for invalid methods
1916 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
1917 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
1918 from ide disks, etc.)
1919 - Allow OS changes without re-installation (to record a changed OS
1920 outside of Ganeti, or to allow OS renames)
1921 - Allow instance creation without OS installation (useful for example if
1922 the OS will be installed manually, or restored from a backup not in
1924 - Implemented option to make cluster ``copyfile`` use the replication
1926 - Added list of enabled hypervisors to ssconf (possibly useful for
1928 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
1930 - A partial form of node re-IP is possible via node readd, which now
1931 allows changed node primary IP
1932 - Command line utilities now show an informational message if the job is
1934 - The logs of the master daemon now show the PID/UID/GID of the
1941 *(Released Fri, 12 Mar 2010)*
1943 During the 2.1.0 long release candidate cycle, a lot of improvements and
1944 changes have accumulated with were released later as 2.1.1.
1949 The node evacuate command (``gnt-node evacuate``) was significantly
1950 rewritten, and as such the IAllocator protocol was changed - a new
1951 request type has been added. This unfortunate change during a stable
1952 series is designed to improve performance of node evacuations; on
1953 clusters with more than about five nodes and which are well-balanced,
1954 evacuation should proceed in parallel for all instances of the node
1955 being evacuated. As such, any existing IAllocator scripts need to be
1956 updated, otherwise the above command will fail due to the unknown
1957 request. The provided "dumb" allocator has not been updated; but the
1958 ganeti-htools package supports the new protocol since version 0.2.4.
1960 Another important change is increased validation of node and instance
1961 names. This might create problems in special cases, if invalid host
1962 names are being used.
1964 Also, a new layer of hypervisor parameters has been added, that sits at
1965 OS level between the cluster defaults and the instance ones. This allows
1966 customisation of virtualization parameters depending on the installed
1967 OS. For example instances with OS 'X' may have a different KVM kernel
1968 (or any other parameter) than the cluster defaults. This is intended to
1969 help managing a multiple OSes on the same cluster, without manual
1970 modification of each instance's parameters.
1972 A tool for merging clusters, ``cluster-merge``, has been added in the
1973 tools sub-directory.
1978 - Improved the int/float conversions that should make the code more
1979 robust in face of errors from the node daemons
1980 - Fixed the remove node code in case of internal configuration errors
1981 - Fixed the node daemon behaviour in face of inconsistent queue
1982 directory (e.g. read-only file-system where we can't open the files
1984 - Fixed the behaviour of gnt-node modify for master candidate demotion;
1985 now it either aborts cleanly or, if given the new "auto_promote"
1986 parameter, will automatically promote other nodes as needed
1987 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
1988 completely prevent Ganeti from working
1989 - Fixed bug for instance export when not all disks were successfully
1991 - Fixed behaviour of node add when the new node is slow in starting up
1993 - Fixed handling of signals in the LUXI client, which should improve
1994 behaviour of command-line scripts
1995 - Added checks for invalid node/instance names in the configuration (now
1996 flagged during cluster verify)
1997 - Fixed watcher behaviour for disk activation errors
1998 - Fixed two potentially endless loops in http library, which led to the
1999 RAPI daemon hanging and consuming 100% CPU in some cases
2000 - Fixed bug in RAPI daemon related to hashed passwords
2001 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
2003 - Enhanced compatibility with non-Debian OSes, but not using absolute
2004 path in some commands and allowing customisation of the ssh
2005 configuration directory
2006 - Fixed possible future issue with new Python versions by abiding to the
2007 proper use of ``__slots__`` attribute on classes
2008 - Added checks that should prevent directory traversal attacks
2009 - Many documentation fixes based on feedback from users
2014 - Added an "early_release" more for instance replace disks and node
2015 evacuate, where we release locks earlier and thus allow higher
2016 parallelism within the cluster
2017 - Added watcher hooks, intended to allow the watcher to restart other
2018 daemons (e.g. from the ganeti-nbma project), but they can be used of
2019 course for any other purpose
2020 - Added a compile-time disable for DRBD barriers, to increase
2021 performance if the administrator trusts the power supply or the
2022 storage system to not lose writes
2023 - Added the option of using syslog for logging instead of, or in
2024 addition to, Ganeti's own log files
2025 - Removed boot restriction for paravirtual NICs for KVM, recent versions
2026 can indeed boot from a paravirtual NIC
2027 - Added a generic debug level for many operations; while this is not
2028 used widely yet, it allows one to pass the debug value all the way to
2030 - Enhanced the hooks environment for instance moves (failovers,
2031 migrations) where the primary/secondary nodes changed during the
2032 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
2033 - Enhanced data validations for many user-supplied values; one important
2034 item is the restrictions imposed on instance and node names, which
2035 might reject some (invalid) host names
2036 - Add a configure-time option to disable file-based storage, if it's not
2037 needed; this allows greater security separation between the master
2038 node and the other nodes from the point of view of the inter-node RPC
2040 - Added user notification in interactive tools if job is waiting in the
2041 job queue or trying to acquire locks
2042 - Added log messages when a job is waiting for locks
2043 - Added filtering by node tags in instance operations which admit
2044 multiple instances (start, stop, reboot, reinstall)
2045 - Added a new tool for cluster mergers, ``cluster-merge``
2046 - Parameters from command line which are of the form ``a=b,c=d`` can now
2047 use backslash escapes to pass in values which contain commas,
2048 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
2050 - For KVM, the instance name is the first parameter passed to KVM, so
2051 that it's more visible in the process list
2057 *(Released Tue, 2 Mar 2010)*
2059 Ganeti 2.1 brings many improvements with it. Major changes:
2061 - Added infrastructure to ease automated disk repairs
2062 - Added new daemon to export configuration data in a cheaper way than
2063 using the remote API
2064 - Instance NICs can now be routed instead of being associated with a
2066 - Improved job locking logic to reduce impact of jobs acquiring multiple
2067 locks waiting for other long-running jobs
2069 In-depth implementation details can be found in the Ganeti 2.1 design
2075 - Added chroot hypervisor
2076 - Added more options to xen-hvm hypervisor (``kernel_path`` and
2078 - Added more options to xen-pvm hypervisor (``use_bootloader``,
2079 ``bootloader_path`` and ``bootloader_args``)
2080 - Added the ``use_localtime`` option for the xen-hvm and kvm
2081 hypervisors, and the default value for this has changed to false (in
2082 2.0 xen-hvm always enabled it)
2083 - Added luxi call to submit multiple jobs in one go
2084 - Added cluster initialization option to not modify ``/etc/hosts``
2086 - Added network interface parameters
2087 - Added dry run mode to some LUs
2088 - Added RAPI resources:
2090 - ``/2/instances/[instance_name]/info``
2091 - ``/2/instances/[instance_name]/replace-disks``
2092 - ``/2/nodes/[node_name]/evacuate``
2093 - ``/2/nodes/[node_name]/migrate``
2094 - ``/2/nodes/[node_name]/role``
2095 - ``/2/nodes/[node_name]/storage``
2096 - ``/2/nodes/[node_name]/storage/modify``
2097 - ``/2/nodes/[node_name]/storage/repair``
2099 - Added OpCodes to evacuate or migrate all instances on a node
2100 - Added new command to list storage elements on nodes (``gnt-node
2101 list-storage``) and modify them (``gnt-node modify-storage``)
2102 - Added new ssconf files with master candidate IP address
2103 (``ssconf_master_candidates_ips``), node primary IP address
2104 (``ssconf_node_primary_ips``) and node secondary IP address
2105 (``ssconf_node_secondary_ips``)
2106 - Added ``ganeti-confd`` and a client library to query the Ganeti
2107 configuration via UDP
2108 - Added ability to run hooks after cluster initialization and before
2110 - Added automatic mode for disk replace (``gnt-instance replace-disks
2112 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
2113 after catastrophic data-loss
2114 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
2116 - Added ``gnt-instance move`` command to move instances
2117 - Added ``gnt-cluster watcher`` command to control watcher
2118 - Added ``gnt-node powercycle`` command to powercycle nodes
2119 - Added new job status field ``lock_status``
2120 - Added parseable error codes to cluster verification (``gnt-cluster
2121 verify --error-codes``) and made output less verbose (use
2122 ``--verbose`` to restore previous behaviour)
2123 - Added UUIDs to the main config entities (cluster, nodes, instances)
2124 - Added support for OS variants
2125 - Added support for hashed passwords in the Ganeti remote API users file
2127 - Added option to specify maximum timeout on instance shutdown
2128 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
2129 - Added new helper script to start and stop Ganeti daemons
2130 (``daemon-util``), with the intent to reduce the work necessary to
2131 adjust Ganeti for non-Debian distributions and to start/stop daemons
2133 - Added more unittests
2134 - Fixed critical bug in ganeti-masterd startup
2135 - Removed the configure-time ``kvm-migration-port`` parameter, this is
2136 now customisable at the cluster level for both the KVM and Xen
2137 hypervisors using the new ``migration_port`` parameter
2138 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
2139 reinstalling an instance
2140 - Allowed ``@`` in tag names
2141 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
2142 - Many documentation updates
2143 - Distribute hypervisor files on ``gnt-cluster redist-conf``
2144 - ``gnt-instance reinstall`` can now reinstall multiple instances
2145 - Updated many command line parameters
2146 - Introduced new OS API version 15
2147 - No longer support a default hypervisor
2148 - Treat virtual LVs as inexistent
2149 - Improved job locking logic to reduce lock contention
2150 - Match instance and node names case insensitively
2151 - Reimplemented bash completion script to be more complete
2158 *(Released Thu, 4 Feb 2010)*
2160 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
2161 - Fix a string formatting bug
2162 - Improve safety of the code in some error paths
2163 - Improve data validation in the master of values returned from nodes
2169 *(Released Thu, 17 Dec 2009)*
2171 - Fix security issue due to missing validation of iallocator names; this
2172 allows local and remote execution of arbitrary executables
2173 - Fix failure of gnt-node list during instance removal
2174 - Ship the RAPI documentation in the archive
2180 *(Released Wed, 30 Sep 2009)*
2182 - Fixed many wrong messages
2183 - Fixed a few bugs related to the locking library
2184 - Fixed MAC checking at instance creation time
2185 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
2186 - Fixed a few issues related to signal handling in both daemons and
2188 - Fixed the example startup script provided
2189 - Fixed insserv dependencies in the example startup script (patch from
2191 - Fixed handling of drained nodes in the iallocator framework
2192 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
2194 - Fixed error related to invalid job IDs in job polling
2195 - Fixed job/opcode persistence on unclean master shutdown
2196 - Fixed handling of partial job processing after unclean master
2198 - Fixed error reporting from LUs, previously all errors were converted
2199 into execution errors
2200 - Fixed error reporting from burnin
2201 - Decreased significantly the memory usage of the job queue
2202 - Optimised slightly multi-job submission
2203 - Optimised slightly opcode loading
2204 - Backported the multi-job submit framework from the development
2205 branch; multi-instance start and stop should be faster
2206 - Added script to clean archived jobs after 21 days; this will reduce
2207 the size of the queue directory
2208 - Added some extra checks in disk size tracking
2209 - Added an example ethers hook script
2210 - Added a cluster parameter that prevents Ganeti from modifying of
2212 - Added more node information to RAPI responses
2213 - Added a ``gnt-job watch`` command that allows following the ouput of a
2215 - Added a bind-address option to ganeti-rapi
2216 - Added more checks to the configuration verify
2217 - Enhanced the burnin script such that some operations can be retried
2219 - Converted instance reinstall to multi-instance model
2225 *(Released Fri, 7 Aug 2009)*
2227 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
2228 to allow using the pre-2.0.2 behaviour in activation, if any existing
2229 instances have mismatched disk sizes in the configuration
2230 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
2231 any configuration mismatches for disk sizes
2232 - Added ``gnt-master cluste-failover --no-voting`` to allow master
2233 failover to work on two-node clusters
2234 - Fixed the ``--net`` option of ``gnt-backup import``, which was
2236 - Fixed detection of OS script errors in ``gnt-backup export``
2237 - Fixed exit code of ``gnt-backup export``
2243 *(Released Fri, 17 Jul 2009)*
2245 - Added experimental support for stripped logical volumes; this should
2246 enhance performance but comes with a higher complexity in the block
2247 device handling; stripping is only enabled when passing
2248 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
2249 affected even in the non-stripped mode
2250 - Improved resiliency against transient failures at the end of DRBD
2251 resyncs, and in general of DRBD resync checks
2252 - Fixed a couple of issues with exports and snapshot errors
2253 - Fixed a couple of issues in instance listing
2254 - Added display of the disk size in ``gnt-instance info``
2255 - Fixed checking for valid OSes in instance creation
2256 - Fixed handling of the "vcpus" parameter in instance listing and in
2257 general of invalid parameters
2258 - Fixed http server library, and thus RAPI, to handle invalid
2259 username/password combinations correctly; this means that now they
2260 report unauthorized for queries too, not only for modifications,
2261 allowing earlier detect of configuration problems
2262 - Added a new "role" node list field, equivalent to the master/master
2263 candidate/drained/offline flags combinations
2264 - Fixed cluster modify and changes of candidate pool size
2265 - Fixed cluster verify error messages for wrong files on regular nodes
2266 - Fixed a couple of issues with node demotion from master candidate role
2267 - Fixed node readd issues
2268 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
2269 - Added a new ``--no-voting`` option for masterfailover to fix failover
2270 on two-nodes clusters when the former master node is unreachable
2271 - Added instance reinstall over RAPI
2277 *(Released Tue, 16 Jun 2009)*
2279 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
2280 allow re-adding the start in single-user option (regression from 1.2)
2281 - the watcher writes the instance status to a file, to allow monitoring
2282 to report the instance status (from the master) based on cached
2283 results of the watcher's queries; while this can get stale if the
2284 watcher is being locked due to other work on the cluster, this is
2285 still an improvement
2286 - the watcher now also restarts the node daemon and the rapi daemon if
2288 - fixed the watcher to handle full and drained queue cases
2289 - hooks export more instance data in the environment, which helps if
2290 hook scripts need to take action based on the instance's properties
2291 (no longer need to query back into ganeti)
2292 - instance failovers when the instance is stopped do not check for free
2293 RAM, so that failing over a stopped instance is possible in low memory
2295 - rapi uses queries for tags instead of jobs (for less job traffic), and
2296 for cluster tags it won't talk to masterd at all but read them from
2298 - a couple of error handling fixes in RAPI
2299 - drbd handling: improved the error handling of inconsistent disks after
2300 resync to reduce the frequency of "there are some degraded disks for
2301 this instance" messages
2302 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
2303 error handling path called a wrong function name)
2309 *(Released Wed, 27 May 2009)*
2311 - no changes from rc5
2317 *(Released Wed, 20 May 2009)*
2319 - fix a couple of bugs (validation, argument checks)
2320 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
2321 - some small improvements to RAPI and IAllocator
2322 - make watcher automatically start the master daemon if down
2328 *(Released Mon, 27 Apr 2009)*
2330 - change the OS list to not require locks; this helps with big clusters
2331 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
2332 volume group is broken
2333 - ``gnt-instance info``, without any arguments, doesn't run for all
2334 instances anymore; either pass ``--all`` or pass the desired
2335 instances; this helps against mistakes on big clusters where listing
2336 the information for all instances takes a long time
2337 - miscellaneous doc and man pages fixes
2343 *(Released Wed, 8 Apr 2009)*
2345 - Change the internal locking model of some ``gnt-node`` commands, in
2346 order to reduce contention (and blocking of master daemon) when
2347 batching many creation/reinstall jobs
2348 - Fixes to Xen soft reboot
2349 - No longer build documentation at build time, instead distribute it in
2350 the archive, in order to reduce the need for the whole docbook/rst
2357 *(Released Fri, 27 Mar 2009)*
2359 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
2360 - Fix watcher startup sequence, improves the behaviour of busy clusters
2361 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
2362 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
2363 bind address and other places
2364 - Some documentation fixes and updates
2370 *(Released Mon, 2 Mar 2009)*
2372 - More documentation updates, now all docs should be more-or-less
2374 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
2376 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
2378 - Fix an issue related to $libdir/run/ganeti and cluster creation
2384 *(Released Thu, 19 Feb 2009)*
2386 - Xen PVM and KVM have switched the default value for the instance root
2387 disk to the first partition on the first drive, instead of the whole
2388 drive; this means that the OS installation scripts must be changed
2390 - Man pages have been updated
2391 - RAPI has been switched by default to HTTPS, and the exported functions
2392 should all work correctly
2393 - RAPI v1 has been removed
2394 - Many improvements to the KVM hypervisor
2395 - Block device errors are now better reported
2396 - Many other bugfixes and small improvements
2402 *(Released Mon, 26 Jan 2009)*
2404 - Version 2 is a general rewrite of the code and therefore the
2405 differences are too many to list, see the design document for 2.0 in
2406 the ``doc/`` subdirectory for more details
2407 - In this beta version there is not yet a migration path from 1.2 (there
2408 will be one in the final 2.0 release)
2409 - A few significant changes are:
2411 - all commands are executed by a daemon (``ganeti-masterd``) and the
2412 various ``gnt-*`` commands are just front-ends to it
2413 - all the commands are entered into, and executed from a job queue,
2414 see the ``gnt-job(8)`` manpage
2415 - the RAPI daemon supports read-write operations, secured by basic
2416 HTTP authentication on top of HTTPS
2417 - DRBD version 0.7 support has been removed, DRBD 8 is the only
2418 supported version (when migrating from Ganeti 1.2 to 2.0, you need
2419 to migrate to DRBD 8 first while still running Ganeti 1.2)
2420 - DRBD devices are using statically allocated minor numbers, which
2421 will be assigned to existing instances during the migration process
2422 - there is support for both Xen PVM and Xen HVM instances running on
2424 - KVM virtualization is supported too
2425 - file-based storage has been implemented, which means that it is
2426 possible to run the cluster without LVM and DRBD storage, for
2427 example using a shared filesystem exported from shared storage (and
2428 still have live migration)
2434 *(Released Tue, 13 Jan 2009)*
2436 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
2437 - Reuse the old instance mac address by default on instance import, if
2438 the instance name is the same.
2439 - Handle situations in which the node info rpc returns incomplete
2441 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
2442 - Improved version of batcher:
2444 - state file support
2445 - instance mac address support
2446 - support for HVM clusters/instances
2448 - Add an option to show the number of cpu sockets and nodes in
2450 - Support OSes that handle more than one version of the OS api (but do
2451 not change the current API in any other way)
2452 - Fix ``gnt-node migrate``
2453 - ``gnt-debug`` man page
2454 - Fixes various more typos and small issues
2455 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
2461 *(Released Wed, 24 Sep 2008)*
2463 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
2464 type of disk exported to fully virtualized instances.
2465 - provide access to the serial console of HVM instances
2466 - instance auto_balance flag, set by default. If turned off it will
2467 avoid warnings on cluster verify if there is not enough memory to fail
2468 over an instance. in the future it will prevent automatically failing
2469 it over when we will support that.
2470 - batcher tool for instance creation, see ``tools/README.batcher``
2471 - ``gnt-instance reinstall --select-os`` to interactively select a new
2472 operating system when reinstalling an instance.
2473 - when changing the memory amount on instance modify a check has been
2474 added that the instance will be able to start. also warnings are
2475 emitted if the instance will not be able to fail over, if auto_balance
2477 - documentation fixes
2478 - sync fields between ``gnt-instance list/modify/add/import``
2479 - fix a race condition in drbd when the sync speed was set after giving
2480 the device a remote peer.
2486 *(Released Tue, 22 Jul 2008)*
2488 - note: the allowed size and number of tags per object were reduced
2489 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
2490 - fixed twisted 8.x compatibility
2491 - fixed ``gnt-instance replace-disks`` with iallocator
2492 - add TCP keepalives on twisted connections to detect restarted nodes
2493 - disk increase support, see ``gnt-instance grow-disk``
2494 - implement bulk node/instance query for RAPI
2495 - add tags in node/instance listing (optional)
2496 - experimental migration (and live migration) support, read the man page
2497 for ``gnt-instance migrate``
2498 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
2499 has some small improvements in handling its state file
2505 *(Released Fri, 13 Jun 2008)*
2507 - Experimental readonly, REST-based remote API implementation;
2508 automatically started on master node, TCP port 5080, if enabled by
2509 ``--enable-rapi`` parameter to configure script.
2510 - Instance allocator support. Add and import instance accept a
2511 ``--iallocator`` parameter, and call that instance allocator to decide
2512 which node to use for the instance. The iallocator document describes
2513 what's expected from an allocator script.
2514 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
2515 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
2516 node is lost there is still enough memory to fail over the instances
2518 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
2519 ``gnt-cluster verify``, to check for site-specific compliance. All the
2520 hooks will run, and their output, if any, will be displayed. Any
2521 failing hook will make the verification return an error value.
2522 - ``gnt-cluster verify`` now checks that its peers are reachable on the
2523 primary and secondary interfaces
2524 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
2525 that is still declared as part of the cluster and has failed.
2526 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
2527 specifying output fields, that just adds the chosen fields to the
2529 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
2530 export from the filesystem.
2531 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
2532 have been added. Using them you can enable/disable acpi and pae
2533 support, and specify a path for a cd image to be exported to the
2534 instance. These parameters as the name suggest only work on HVM
2536 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
2537 PAE support will be set to the previously hardcoded values, but the
2538 (previously hardcoded) path to the CDROM ISO image will be unset and
2539 if required, needs to be set manually with ``gnt-instance modify``
2541 - The address to which an instance's VNC console is bound is now
2542 selectable per-instance, rather than being cluster wide. Of course
2543 this only applies to instances controlled via VNC, so currently just
2544 applies to HVM clusters.
2550 *(Released Mon, 18 Feb 2008)*
2552 - more tweaks to the disk activation code (especially helpful for DRBD)
2553 - change the default ``gnt-instance list`` output format, now there is
2554 one combined status field (see the manpage for the exact values this
2556 - some more fixes for the mac export to hooks change
2557 - make Ganeti not break with DRBD 8.2.x (which changed the version
2558 format in ``/proc/drbd``) (issue 24)
2559 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
2560 template, allowing migration from DRBD0.7+MD to DRBD8
2566 *(Released Wed, 30 Jan 2008)*
2568 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
2570 - add command aliases infrastructure and a few aliases
2571 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
2572 man pages and the ``--help`` option of ``gnt-node
2573 list``/``gnt-instance list``
2574 - fix ``gnt-backup list`` with down nodes (issue 21)
2575 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
2576 - fix the dist archive and add a check for including svn/git files in
2578 - some developer-related changes: improve the burnin and the QA suite,
2579 add an upload script for testing during development
2585 *(Released Wed, 16 Jan 2008)*
2587 - experimental HVM support, read the install document, section
2588 "Initializing the cluster"
2589 - allow for the PVM hypervisor per-instance kernel and initrd paths
2590 - add a new command ``gnt-cluster verify-disks`` which uses a new
2591 algorithm to improve the reconnection of the DRBD pairs if the device
2592 on the secondary node has gone away
2593 - make logical volume code auto-activate LVs at disk activation time
2594 - slightly improve the speed of activating disks
2595 - allow specification of the MAC address at instance creation time, and
2596 changing it later via ``gnt-instance modify``
2597 - fix handling of external commands that generate lots of output on
2599 - update documentation with regard to minimum version of DRBD8 supported
2605 *(Released Tue, 4 Dec 2007)*
2607 - Log the ``xm create`` output to the node daemon log on failure (to
2608 help diagnosing the error)
2609 - In debug mode, log all external commands output if failed to the logs
2610 - Change parsing of lvm commands to ignore stderr
2616 *(Released Wed, 28 Nov 2007)*
2618 - Another round of updates to the DRBD 8 code to deal with more failures
2619 in the replace secondary node operation
2620 - Some more logging of failures in disk operations (lvm, drbd)
2621 - A few documentation updates
2628 *(Released Tue, 13 Nov 2007)*
2630 - Change configuration file format from Python's Pickle to JSON.
2631 Upgrading is possible using the cfgupgrade utility.
2632 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
2633 faster replace disks and is more stable (DRBD 8 has many improvements
2634 compared to DRBD 0.7)
2635 - Added command line tags support (see man pages for ``gnt-instance``,
2636 ``gnt-node``, ``gnt-cluster``)
2637 - Added instance rename support
2638 - Added multi-instance startup/shutdown
2639 - Added cluster rename support
2640 - Added ``gnt-node evacuate`` to simplify some node operations
2641 - Added instance reboot operation that can speedup reboot as compared to
2643 - Soften the requirement that hostnames are in FQDN format
2644 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
2646 - Removed dependency on debian's patched fping that uses the
2647 non-standard ``-S`` option
2648 - Now the OS definitions are searched for in multiple, configurable
2649 paths (easier for distros to package)
2650 - Some changes to the hooks infrastructure (especially the new
2651 post-configuration update hook)
2652 - Other small bugfixes
2654 .. vim: set textwidth=72 syntax=rst :