8 *(Released Wed, 6 Mar 2013)*
10 - ``gnt-instance batch-create`` has been changed to use the bulk create
11 opcode from Ganeti. This lead to incompatible changes in the format of
12 the JSON file. It's now not a custom dict anymore but a dict
13 compatible with the ``OpInstanceCreate`` opcode.
14 - Parent directories for file storage need to be listed in
15 ``$sysconfdir/ganeti/file-storage-paths`` now. ``cfgupgrade`` will
16 write the file automatically based on old configuration values, but it
17 can not distribute it across all nodes and the file contents should be
18 verified. Use ``gnt-cluster copyfile
19 $sysconfdir/ganeti/file-storage-paths`` once the cluster has been
20 upgraded. The reason for requiring this list of paths now is that
21 before it would have been possible to inject new paths via RPC,
22 allowing files to be created in arbitrary locations. The RPC protocol
23 is protected using SSL/X.509 certificates, but as a design principle
24 Ganeti does not permit arbitrary paths to be passed.
25 - The parsing of the variants file for OSes (see
26 :manpage:`ganeti-os-interface(7)`) has been slightly changed: now empty
27 lines and comment lines (starting with ``#``) are ignored for better
29 - The ``setup-ssh`` tool added in Ganeti 2.2 has been replaced and is no
30 longer available. ``gnt-node add`` now invokes a new tool on the
31 destination node, named ``prepare-node-join``, to configure the SSH
32 daemon. Paramiko is no longer necessary to configure nodes' SSH
33 daemons via ``gnt-node add``.
34 - A new user option, :pyeval:`rapi.RAPI_ACCESS_READ`, has been added
35 for RAPI users. It allows granting permissions to query for
36 information to a specific user without giving
37 :pyeval:`rapi.RAPI_ACCESS_WRITE` permissions.
38 - A new tool named ``node-cleanup`` has been added. It cleans remains of
39 a cluster from a machine by stopping all daemons, removing
40 certificates and ssconf files. Unless the ``--no-backup`` option is
41 given, copies of the certificates are made.
42 - Draining (``gnt-cluster queue drain``) and un-draining the job queue
43 (``gnt-cluster queue undrain``) now affects all nodes in a cluster and
44 the flag is not reset after a master failover.
45 - Instance creations now support the use of opportunistic locking,
46 potentially speeding up the (parallel) creation of multiple instances.
47 This feature is currently only available via the :doc:`RAPI
48 <rapi>` interface and when an instance allocator is used. If the
49 ``opportunistic_locking`` parameter is set the opcode will try to
50 acquire as many locks as possible, but will not wait for any locks
51 held by other opcodes. If the not enough resources can be found to
52 allocate the instance, the temporary error code
53 :pyeval:`errors.ECODE_TEMP_NORES` is returned. The operation can be
54 retried thereafter, with or without opportunistic locking.
55 - Man pages can now be included when the documentation is built, in
56 which case the output is in ``doc/man-html``. The configure-time
57 option is ``--enable-manpages-in-doc``. Sphinx 1.0 or higher is
59 - A new htool called ``hroller`` helps scheduling parallel shutdown of
60 nodes on clusters where all instances are shut down, but without ever
61 rebooting both primary and secondary node for an instance at the same
63 - Queries not needing locks or RPC access to the node can now be
64 performed by the confd daemon, making them independent from jobs, and
65 thus faster to execute. This is selectable at configure time.
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)`.
69 - A new tool, called ``mon-collector``, is the stand-alone executor of
70 the data collectors for a monitoring system. As of this version, it
71 just includes the DRBD data collector, that can be executed by calling
72 ``mon-collector`` using the ``drbd`` parameter. See
73 :manpage:`mon-collector(7)`.
74 - New ``exclusive-storage`` node parameter added. When it's set to true,
75 physical disks are assigned in an exclusive fashion to instances, as
76 documented in :doc:`Partitioned Ganeti <design-partitioned>`.
77 Currently, only instances using the ``plain`` disk template are
79 - New network management functionality to support automatic allocation
80 of IP addresses and managing of network parameters. See
81 :manpage:`gnt-network(8)` for more details.
82 - New external storage backend, to allow managing arbitrary storage
83 systems external to the cluster. See
84 :manpage:`ganeti-extstorage-interface(7)`.
85 - Instance renames of LVM-based instances will now update the LV tags
86 (which can be used to recover the instance-to-LV mapping in case of
93 *(Released Fri, 21 Dec 2012)*
95 Important behaviour change: hbal won't rebalance anymore instances which
96 have the ``auto_balance`` attribute set to false. This was the intention
97 all along, but until now it only skipped those from the N+1 memory
98 reservation (DRBD-specific).
100 A significant number of bug fixes in this release:
102 - Fixed disk adoption interaction with ipolicy checks.
103 - Fixed networking issues when instances are started, stopped or
104 migrated, by forcing the tap device's MAC prefix to "fe" (issue 217).
105 - Fixed the warning in cluster verify for shared storage instances not
107 - Fixed removal of storage directory on shared file storage (issue 262).
108 - Fixed validation of LVM volume group name in OpClusterSetParams
109 (``gnt-cluster modify``) (issue 285).
110 - Fixed runtime memory increases (``gnt-instance modify -m``).
111 - Fixed live migration under Xen's ``xl`` mode.
112 - Fixed ``gnt-instance console`` with ``xl``.
113 - Fixed building with newer Haskell compiler/libraries.
114 - Fixed PID file writing in Haskell daemons (confd); this prevents
115 restart issues if confd was launched manually (outside of
116 ``daemon-util``) while another copy of it was running
117 - Fixed a type error when doing live migrations with KVM (issue 297) and
118 the error messages for failing migrations have been improved.
119 - Fixed opcode validation for the out-of-band commands (``gnt-node
121 - Fixed a type error when unsetting OS hypervisor parameters (issue
122 311); now it's possible to unset all OS-specific hypervisor
124 - Fixed the ``dry-run`` mode for many operations: verification of
125 results was over-zealous but didn't take into account the ``dry-run``
126 operation, resulting in "wrong" failures.
127 - Fixed bash completion in ``gnt-job list`` when the job queue has
128 hundreds of entries; especially with older ``bash`` versions, this
129 results in significant CPU usage.
131 And lastly, a few other improvements have been made:
133 - Added option to force master-failover without voting (issue 282).
134 - Clarified error message on lock conflict (issue 287).
135 - Logging of newly submitted jobs has been improved (issue 290).
136 - Hostname checks have been made uniform between instance rename and
138 - The ``--submit`` option is now supported by ``gnt-debug delay``.
139 - Shutting down the master daemon by sending SIGTERM now stops it from
140 processing jobs waiting for locks; instead, those jobs will be started
141 once again after the master daemon is started the next time (issue
143 - Support for Xen's ``xl`` program has been improved (besides the fixes
145 - Reduced logging noise in the Haskell confd daemon (only show one log
146 entry for each config reload, instead of two).
147 - Several man page updates and typo fixes.
153 *(Released Fri, 12 Oct 2012)*
155 A small bugfix release. Among the bugs fixed:
157 - Fixed double use of ``PRIORITY_OPT`` in ``gnt-node migrate``, that
158 made the command unusable.
159 - Commands that issue many jobs don't fail anymore just because some jobs
160 take so long that other jobs are archived.
161 - Failures during ``gnt-instance reinstall`` are reflected by the exit
163 - Issue 190 fixed. Check for DRBD in cluster verify is enabled only when
165 - When ``always_failover`` is set, ``--allow-failover`` is not required
166 in migrate commands anymore.
167 - ``bash_completion`` works even if extglob is disabled.
168 - Fixed bug with locks that made failover for RDB-based instances fail.
169 - Fixed bug in non-mirrored instance allocation that made Ganeti choose
170 a random node instead of one based on the allocator metric.
171 - Support for newer versions of pylint and pep8.
172 - Hail doesn't fail anymore when trying to add an instance of type
173 ``file``, ``sharedfile`` or ``rbd``.
174 - Added new Makefile target to rebuild the whole distribution, so that
175 all files are included.
181 *(Released Fri, 27 Jul 2012)*
184 .. attention:: The ``LUXI`` protocol has been made more consistent
185 regarding its handling of command arguments. This, however, leads to
186 incompatibility issues with previous versions. Please ensure that you
187 restart Ganeti daemons soon after the upgrade, otherwise most
188 ``LUXI`` calls (job submission, setting/resetting the drain flag,
189 pausing/resuming the watcher, cancelling and archiving jobs, querying
190 the cluster configuration) will fail.
199 The current ``admin_up`` field, which used to denote whether an instance
200 should be running or not, has been removed. Instead, ``admin_state`` is
201 introduced, with 3 possible values -- ``up``, ``down`` and ``offline``.
203 The rational behind this is that an instance being “down” can have
206 - it could be down during a reboot
207 - it could be temporarily be down for a reinstall
208 - or it could be down because it is deprecated and kept just for its
211 The previous Boolean state was making it difficult to do capacity
212 calculations: should Ganeti reserve memory for a down instance? Now, the
213 tri-state field makes it clear:
215 - in ``up`` and ``down`` state, all resources are reserved for the
216 instance, and it can be at any time brought up if it is down
217 - in ``offline`` state, only disk space is reserved for it, but not
220 The field can have an extra use: since the transition between ``up`` and
221 ``down`` and vice-versus is done via ``gnt-instance start/stop``, but
222 transition between ``offline`` and ``down`` is done via ``gnt-instance
223 modify``, it is possible to given different rights to users. For
224 example, owners of an instance could be allowed to start/stop it, but
225 not transition it out of the offline state.
227 Instance policies and specs
228 +++++++++++++++++++++++++++
230 In previous Ganeti versions, an instance creation request was not
231 limited on the minimum size and on the maximum size just by the cluster
232 resources. As such, any policy could be implemented only in third-party
233 clients (RAPI clients, or shell wrappers over ``gnt-*``
234 tools). Furthermore, calculating cluster capacity via ``hspace`` again
235 required external input with regards to instance sizes.
237 In order to improve these workflows and to allow for example better
238 per-node group differentiation, we introduced instance specs, which
241 - minimum instance disk size, disk count, memory size, cpu count
242 - maximum values for the above metrics
243 - and “standard” values (used in ``hspace`` to calculate the standard
246 The minimum/maximum values can be also customised at node-group level,
247 for example allowing more powerful hardware to support bigger instance
250 Beside the instance specs, there are a few other settings belonging to
251 the instance policy framework. It is possible now to customise, per
252 cluster and node-group:
254 - the list of allowed disk templates
255 - the maximum ratio of VCPUs per PCPUs (to control CPU oversubscription)
256 - the maximum ratio of instance to spindles (see below for more
257 information) for local storage
259 All these together should allow all tools that talk to Ganeti to know
260 what are the ranges of allowed values for instances and the
261 over-subscription that is allowed.
263 For the VCPU/PCPU ratio, we already have the VCPU configuration from the
264 instance configuration, and the physical CPU configuration from the
265 node. For the spindle ratios however, we didn't track before these
266 values, so new parameters have been added:
268 - a new node parameter ``spindle_count``, defaults to 1, customisable at
269 node group or node level
270 - at new backend parameter (for instances), ``spindle_use`` defaults to 1
272 Note that spindles in this context doesn't need to mean actual
273 mechanical hard-drives; it's just a relative number for both the node
274 I/O capacity and instance I/O consumption.
276 Instance migration behaviour
277 ++++++++++++++++++++++++++++
279 While live-migration is in general desirable over failover, it is
280 possible that for some workloads it is actually worse, due to the
281 variable time of the “suspend” phase during live migration.
283 To allow the tools to work consistently over such instances (without
284 having to hard-code instance names), a new backend parameter
285 ``always_failover`` has been added to control the migration/failover
286 behaviour. When set to True, all migration requests for an instance will
287 instead fall-back to failover.
289 Instance memory ballooning
290 ++++++++++++++++++++++++++
292 Initial support for memory ballooning has been added. The memory for an
293 instance is no longer fixed (backend parameter ``memory``), but instead
294 can vary between minimum and maximum values (backend parameters
295 ``minmem`` and ``maxmem``). Currently we only change an instance's
298 - live migrating or failing over and instance and the target node
299 doesn't have enough memory
300 - user requests changing the memory via ``gnt-instance modify
306 In order to control the use of specific CPUs by instance, support for
307 controlling CPU pinning has been added for the Xen, HVM and LXC
308 hypervisors. This is controlled by a new hypervisor parameter
309 ``cpu_mask``; details about possible values for this are in the
310 :manpage:`gnt-instance(8)`. Note that use of the most specific (precise
311 VCPU-to-CPU mapping) form will work well only when all nodes in your
312 cluster have the same amount of CPUs.
317 Another area in which Ganeti was not customisable were the parameters
318 used for storage configuration, e.g. how many stripes to use for LVM,
319 DRBD resync configuration, etc.
321 To improve this area, we've added disks parameters, which are
322 customisable at cluster and node group level, and which allow to
323 specify various parameters for disks (DRBD has the most parameters
324 currently), for example:
326 - DRBD resync algorithm and parameters (e.g. speed)
327 - the default VG for meta-data volumes for DRBD
328 - number of stripes for LVM (plain disk template)
331 These parameters can be modified via ``gnt-cluster modify -D …`` and
332 ``gnt-group modify -D …``, and are used at either instance creation (in
333 case of LVM stripes, for example) or at disk “activation” time
336 Rados block device support
337 ++++++++++++++++++++++++++
339 A Rados (http://ceph.com/wiki/Rbd) storage backend has been added,
340 denoted by the ``rbd`` disk template type. This is considered
341 experimental, feedback is welcome. For details on configuring it, see
342 the :doc:`install` document and the :manpage:`gnt-cluster(8)` man page.
347 The existing master IP functionality works well only in simple setups (a
348 single network shared by all nodes); however, if nodes belong to
349 different networks, then the ``/32`` setup and lack of routing
350 information is not enough.
352 To allow the master IP to function well in more complex cases, the
353 system was reworked as follows:
355 - a master IP netmask setting has been added
356 - the master IP activation/turn-down code was moved from the node daemon
358 - whether to run the Ganeti-supplied master IP script or a user-supplied
359 on is a ``gnt-cluster init`` setting
361 Details about the location of the standard and custom setup scripts are
362 in the man page :manpage:`gnt-cluster(8)`; for information about the
363 setup script protocol, look at the Ganeti-supplied script.
368 The `SPICE <http://www.linux-kvm.org/page/SPICE>`_ support has been
371 It is now possible to use TLS-protected connections, and when renewing
372 or changing the cluster certificates (via ``gnt-cluster renew-crypto``,
373 it is now possible to specify spice or spice CA certificates. Also, it
374 is possible to configure a password for SPICE sessions via the
375 hypervisor parameter ``spice_password_file``.
377 There are also new parameters to control the compression and streaming
378 options (e.g. ``spice_image_compression``, ``spice_streaming_video``,
379 etc.). For details, see the man page :manpage:`gnt-instance(8)` and look
380 for the spice parameters.
382 Lastly, it is now possible to see the SPICE connection information via
383 ``gnt-instance console``.
388 A new tool (``tools/ovfconverter``) has been added that supports
389 conversion between Ganeti and the `Open Virtualization Format
390 <http://en.wikipedia.org/wiki/Open_Virtualization_Format>`_ (both to and
393 This relies on the ``qemu-img`` tool to convert the disk formats, so the
394 actual compatibility with other virtualization solutions depends on it.
399 The configuration query daemon (``ganeti-confd``) is now optional, and
400 has been rewritten in Haskell; whether to use the daemon at all, use the
401 Python (default) or the Haskell version is selectable at configure time
402 via the ``--enable-confd`` parameter, which can take one of the
403 ``haskell``, ``python`` or ``no`` values. If not used, disabling the
404 daemon will result in a smaller footprint; for larger systems, we
405 welcome feedback on the Haskell version which might become the default
408 If you want to use ``gnt-node list-drbd`` you need to have the Haskell
409 daemon running. The Python version doesn't implement the new call.
412 User interface changes
413 ~~~~~~~~~~~~~~~~~~~~~~
415 We have replaced the ``--disks`` option of ``gnt-instance
416 replace-disks`` with a more flexible ``--disk`` option, which allows
417 adding and removing disks at arbitrary indices (Issue 188). Furthermore,
418 disk size and mode can be changed upon recreation (via ``gnt-instance
419 recreate-disks``, which accepts the same ``--disk`` option).
421 As many people are used to a ``show`` command, we have added that as an
422 alias to ``info`` on all ``gnt-*`` commands.
424 The ``gnt-instance grow-disk`` command has a new mode in which it can
425 accept the target size of the disk, instead of the delta; this can be
426 more safe since two runs in absolute mode will be idempotent, and
427 sometimes it's also easier to specify the desired size directly.
429 Also the handling of instances with regard to offline secondaries has
430 been improved. Instance operations should not fail because one of it's
431 secondary nodes is offline, even though it's safe to proceed.
433 A new command ``list-drbd`` has been added to the ``gnt-node`` script to
434 support debugging of DRBD issues on nodes. It provides a mapping of DRBD
435 minors to instance name.
440 RAPI coverage has improved, with (for example) new resources for
441 recreate-disks, node power-cycle, etc.
446 There is partial support for ``xl`` in the Xen hypervisor; feedback is
449 Python 2.7 is better supported, and after Ganeti 2.6 we will investigate
450 whether to still support Python 2.4 or move to Python 2.6 as minimum
453 Support for Fedora has been slightly improved; the provided example
454 init.d script should work better on it and the INSTALL file should
455 document the needed dependencies.
460 The deprecated ``QueryLocks`` LUXI request has been removed. Use
461 ``Query(what=QR_LOCK, ...)`` instead.
463 The LUXI requests :pyeval:`luxi.REQ_QUERY_JOBS`,
464 :pyeval:`luxi.REQ_QUERY_INSTANCES`, :pyeval:`luxi.REQ_QUERY_NODES`,
465 :pyeval:`luxi.REQ_QUERY_GROUPS`, :pyeval:`luxi.REQ_QUERY_EXPORTS` and
466 :pyeval:`luxi.REQ_QUERY_TAGS` are deprecated and will be removed in a
467 future version. :pyeval:`luxi.REQ_QUERY` should be used instead.
469 RAPI client: ``CertificateError`` now derives from
470 ``GanetiApiError``. This should make it more easy to handle Ganeti
473 Deprecation warnings due to PyCrypto/paramiko import in
474 ``tools/setup-ssh`` have been silenced, as usually they are safe; please
475 make sure to run an up-to-date paramiko version, if you use this tool.
477 The QA scripts now depend on Python 2.5 or above (the main code base
478 still works with Python 2.4).
480 The configuration file (``config.data``) is now written without
481 indentation for performance reasons; if you want to edit it, it can be
482 re-formatted via ``tools/fmtjson``.
484 A number of bugs has been fixed in the cluster merge tool.
486 ``x509`` certification verification (used in import-export) has been
487 changed to allow the same clock skew as permitted by the cluster
488 verification. This will remove some rare but hard to diagnose errors in
495 *(Released Thu, 19 Jul 2012)*
497 Very few changes from rc4 to the final release, only bugfixes:
499 - integrated fixes from release 2.5.2 (fix general boot flag for KVM
500 instance, fix CDROM booting for KVM instances)
501 - fixed node group modification of node parameters
502 - fixed issue in LUClusterVerifyGroup with multi-group clusters
503 - fixed generation of bash completion to ensure a stable ordering
510 *(Released Fri, 13 Jul 2012)*
512 Third release candidate for 2.6. The following changes were done from
515 - Fixed ``UpgradeConfig`` w.r.t. to disk parameters on disk objects.
516 - Fixed an inconsistency in the LUXI protocol with the provided
517 arguments (NOT backwards compatible)
518 - Fixed a bug with node groups ipolicy where ``min`` was greater than
519 the cluster ``std`` value
520 - Implemented a new ``gnt-node list-drbd`` call to list DRBD minors for
521 easier instance debugging on nodes (requires ``hconfd`` to work)
527 *(Released Tue, 03 Jul 2012)*
529 Second release candidate for 2.6. The following changes were done from
532 - Fixed ``gnt-cluster verify`` regarding ``master-ip-script`` on non
534 - Fixed a RAPI regression on missing beparams/memory
535 - Fixed redistribution of files on offline nodes
536 - Added possibility to run activate-disks even though secondaries are
537 offline. With this change it relaxes also the strictness on some other
538 commands which use activate disks internally:
539 * ``gnt-instance start|reboot|rename|backup|export``
540 - Made it possible to remove safely an instance if its secondaries are
542 - Made it possible to reinstall even though secondaries are offline
548 *(Released Mon, 25 Jun 2012)*
550 First release candidate for 2.6. The following changes were done from
553 - Fixed bugs with disk parameters and ``rbd`` templates as well as
555 - Made ``gnt-instance modify`` more consistent regarding new NIC/Disk
556 behaviour. It supports now the modify operation
557 - ``hcheck`` implemented to analyze cluster health and possibility of
558 improving health by rebalance
559 - ``hbal`` has been improved in dealing with split instances
565 *(Released Mon, 11 Jun 2012)*
567 Second beta release of 2.6. The following changes were done from beta2
570 - Fixed ``daemon-util`` with non-root user models
571 - Fixed creation of plain instances with ``--no-wait-for-sync``
572 - Fix wrong iv_names when running ``cfgupgrade``
573 - Export more information in RAPI group queries
574 - Fixed bug when changing instance network interfaces
575 - Extended burnin to do NIC changes
576 - query: Added ``<``, ``>``, ``<=``, ``>=`` comparison operators
577 - Changed default for DRBD barriers
578 - Fixed DRBD error reporting for syncer rate
579 - Verify the options on disk parameters
581 And of course various fixes to documentation and improved unittests and
588 *(Released Wed, 23 May 2012)*
590 First beta release of 2.6. The following changes were done from beta1 to
593 - integrated patch for distributions without ``start-stop-daemon``
594 - adapted example init.d script to work on Fedora
595 - fixed log handling in Haskell daemons
596 - adapted checks in the watcher for pycurl linked against libnss
597 - add partial support for ``xl`` instead of ``xm`` for Xen
598 - fixed a type issue in cluster verification
599 - fixed ssconf handling in the Haskell code (was breaking confd in IPv6
602 Plus integrated fixes from the 2.5 branch:
604 - fixed ``kvm-ifup`` to use ``/bin/bash``
605 - fixed parallel build failures
606 - KVM live migration when using a custom keymap
612 *(Released Tue, 24 Jul 2012)*
614 A small bugfix release, with no new features:
616 - fixed bash-isms in kvm-ifup, for compatibility with systems which use a
617 different default shell (e.g. Debian, Ubuntu)
618 - fixed KVM startup and live migration with a custom keymap (fixes Issue
619 243 and Debian bug #650664)
620 - fixed compatibility with KVM versions that don't support multiple boot
621 devices (fixes Issue 230 and Debian bug #624256)
623 Additionally, a few fixes were done to the build system (fixed parallel
624 build failures) and to the unittests (fixed race condition in test for
625 FileID functions, and the default enable/disable mode for QA test is now
632 *(Released Fri, 11 May 2012)*
634 A small bugfix release.
636 The main issues solved are on the topic of compatibility with newer LVM
639 - fixed parsing of ``lv_attr`` field
640 - adapted to new ``vgreduce --removemissing`` behaviour where sometimes
641 the ``--force`` flag is needed
643 Also on the topic of compatibility, ``tools/lvmstrap`` has been changed
644 to accept kernel 3.x too (was hardcoded to 2.6.*).
646 A regression present in 2.5.0 that broke handling (in the gnt-* scripts)
647 of hook results and that also made display of other errors suboptimal
648 was fixed; the code behaves now like 2.4 and earlier.
650 Another change in 2.5, the cleanup of the OS scripts environment, is too
651 aggressive: it removed even the ``PATH`` variable, which requires the OS
652 scripts to *always* need to export it. Since this is a bit too strict,
653 we now export a minimal PATH, the same that we export for hooks.
655 The fix for issue 201 (Preserve bridge MTU in KVM ifup script) was
656 integrated into this release.
658 Finally, a few other miscellaneous changes were done (no new features,
659 just small improvements):
661 - Fix ``gnt-group --help`` display
662 - Fix hardcoded Xen kernel path
663 - Fix grow-disk handling of invalid units
664 - Update synopsis for ``gnt-cluster repair-disk-sizes``
665 - Accept both PUT and POST in noded (makes future upgrade to 2.6 easier)
671 *(Released Thu, 12 Apr 2012)*
673 Incompatible/important changes and bugfixes
674 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
676 - The default of the ``/2/instances/[instance_name]/rename`` RAPI
677 resource's ``ip_check`` parameter changed from ``True`` to ``False``
678 to match the underlying LUXI interface.
679 - The ``/2/nodes/[node_name]/evacuate`` RAPI resource was changed to use
680 body parameters, see :doc:`RAPI documentation <rapi>`. The server does
681 not maintain backwards-compatibility as the underlying operation
682 changed in an incompatible way. The RAPI client can talk to old
683 servers, but it needs to be told so as the return value changed.
684 - When creating file-based instances via RAPI, the ``file_driver``
685 parameter no longer defaults to ``loop`` and must be specified.
686 - The deprecated ``bridge`` NIC parameter is no longer supported. Use
688 - Support for the undocumented and deprecated RAPI instance creation
689 request format version 0 has been dropped. Use version 1, supported
690 since Ganeti 2.1.3 and :doc:`documented <rapi>`, instead.
691 - Pyparsing 1.4.6 or above is required, see :doc:`installation
692 documentation <install>`.
693 - The "cluster-verify" hooks are now executed per group by the
694 ``OP_CLUSTER_VERIFY_GROUP`` opcode. This maintains the same behavior
695 if you just run ``gnt-cluster verify``, which generates one opcode per
697 - The environment as passed to the OS scripts is cleared, and thus no
698 environment variables defined in the node daemon's environment will be
699 inherited by the scripts.
700 - The :doc:`iallocator <iallocator>` mode ``multi-evacuate`` has been
702 - :doc:`New iallocator modes <design-multi-reloc>` have been added to
703 support operations involving multiple node groups.
704 - Offline nodes are ignored when failing over an instance.
705 - Support for KVM version 1.0, which changed the version reporting format
707 - TCP/IP ports used by DRBD disks are returned to a pool upon instance
709 - ``Makefile`` is now compatible with Automake 1.11.2
710 - Includes all bugfixes made in the 2.4 series
715 - The ganeti-htools project has been merged into the ganeti-core source
716 tree and will be built as part of Ganeti (see :doc:`install-quick`).
717 - Implemented support for :doc:`shared storage <design-shared-storage>`.
718 - Add support for disks larger than 2 TB in ``lvmstrap`` by supporting
719 GPT-style partition tables (requires `parted
720 <http://www.gnu.org/s/parted/>`_).
721 - Added support for floppy drive and 2nd CD-ROM drive in KVM hypervisor.
722 - Allowed adding tags on instance creation.
723 - Export instance tags to hooks (``INSTANCE_TAGS``, see :doc:`hooks`)
724 - Allow instances to be started in a paused state, enabling the user to
725 see the complete console output on boot using the console.
726 - Added new hypervisor flag to control default reboot behaviour
727 (``reboot_behavior``).
728 - Added support for KVM keymaps (hypervisor parameter ``keymap``).
729 - Improved out-of-band management support:
731 - Added ``gnt-node health`` command reporting the health status of
733 - Added ``gnt-node power`` command to manage power status of nodes.
734 - Added command for emergency power-off (EPO), ``gnt-cluster epo``.
736 - Instance migration can fall back to failover if instance is not
738 - Filters can be used when listing nodes, instances, groups and locks;
739 see :manpage:`ganeti(7)` manpage.
740 - Added post-execution status as variables to :doc:`hooks <hooks>`
742 - Instance tags are exported/imported together with the instance.
743 - When given an explicit job ID, ``gnt-job info`` will work for archived
745 - Jobs can define dependencies on other jobs (not yet supported via
746 RAPI or command line, but used by internal commands and usable via
749 - Lock monitor (``gnt-debug locks``) shows jobs waiting for
752 - Instance failover is now available as a RAPI resource
753 (``/2/instances/[instance_name]/failover``).
754 - ``gnt-instance info`` defaults to static information if primary node
756 - Opcodes have a new ``comment`` attribute.
757 - Added basic SPICE support to KVM hypervisor.
758 - ``tools/ganeti-listrunner`` allows passing of arguments to executable.
760 Node group improvements
761 ~~~~~~~~~~~~~~~~~~~~~~~
763 - ``gnt-cluster verify`` has been modified to check groups separately,
764 thereby improving performance.
765 - Node group support has been added to ``gnt-cluster verify-disks``,
766 which now operates per node group.
767 - Watcher has been changed to work better with node groups.
769 - One process and state file per node group.
770 - Slow watcher in one group doesn't block other group's watcher.
772 - Added new command, ``gnt-group evacuate``, to move all instances in a
773 node group to other groups.
774 - Added ``gnt-instance change-group`` to move an instance to another
776 - ``gnt-cluster command`` and ``gnt-cluster copyfile`` now support
777 per-group operations.
778 - Node groups can be tagged.
779 - Some operations switch from an exclusive to a shared lock as soon as
781 - Instance's primary and secondary nodes' groups are now available as
782 query fields (``pnode.group``, ``pnode.group.uuid``, ``snodes.group``
783 and ``snodes.group.uuid``).
788 - Numerous updates to documentation and manpages.
790 - :doc:`RAPI <rapi>` documentation now has detailed parameter
792 - Some opcode/job results are now also documented, see :doc:`RAPI
795 - A lockset's internal lock is now also visible in lock monitor.
796 - Log messages from job queue workers now contain information about the
797 opcode they're processing.
798 - ``gnt-instance console`` no longer requires the instance lock.
799 - A short delay when waiting for job changes reduces the number of LUXI
800 requests significantly.
801 - DRBD metadata volumes are overwritten with zeros during disk creation.
802 - Out-of-band commands no longer acquire the cluster lock in exclusive
804 - ``devel/upload`` now uses correct permissions for directories.
810 *(Released Fri, 23 Mar 2012)*
812 This was the sixth release candidate of the 2.5 series.
818 *(Released Mon, 9 Jan 2012)*
820 This was the fifth release candidate of the 2.5 series.
826 *(Released Thu, 27 Oct 2011)*
828 This was the fourth release candidate of the 2.5 series.
834 *(Released Wed, 26 Oct 2011)*
836 This was the third release candidate of the 2.5 series.
842 *(Released Tue, 18 Oct 2011)*
844 This was the second release candidate of the 2.5 series.
850 *(Released Tue, 4 Oct 2011)*
852 This was the first release candidate of the 2.5 series.
858 *(Released Wed, 31 Aug 2011)*
860 This was the third beta release of the 2.5 series.
866 *(Released Mon, 22 Aug 2011)*
868 This was the second beta release of the 2.5 series.
874 *(Released Fri, 12 Aug 2011)*
876 This was the first beta release of the 2.5 series.
882 *(Released Thu, 27 Oct 2011)*
884 - Fixed bug when parsing command line parameter values ending in
886 - Fixed assertion error after unclean master shutdown
887 - Disable HTTP client pool for RPC, significantly reducing memory usage
889 - Fixed queue archive creation with wrong permissions
895 *(Released Tue, 23 Aug 2011)*
899 - Fixed documentation for importing with ``--src-dir`` option
900 - Fixed a bug in ``ensure-dirs`` with queue/archive permissions
901 - Fixed a parsing issue with DRBD 8.3.11 in the Linux kernel
907 *(Released Fri, 5 Aug 2011)*
909 Many bug-fixes and a few small features:
911 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
912 issues when you tried to add an instance with two MAC addresses in one
914 - KVM: fixed per-instance stored UID value
915 - KVM: configure bridged NICs at migration start
916 - KVM: Fix a bug where instance will not start with never KVM versions
918 - Added OS search path to ``gnt-cluster info``
919 - Fixed an issue with ``file_storage_dir`` where you were forced to
920 provide an absolute path, but the documentation states it is a
921 relative path, the documentation was right
922 - Added a new parameter to instance stop/start called ``--no-remember``
923 that will make the state change to not be remembered
924 - Implemented ``no_remember`` at RAPI level
925 - Improved the documentation
926 - Node evacuation: don't call IAllocator if node is already empty
927 - Fixed bug in DRBD8 replace disks on current nodes
928 - Fixed bug in recreate-disks for DRBD instances
929 - Moved assertion checking locks in ``gnt-instance replace-disks``
930 causing it to abort with not owning the right locks for some situation
931 - Job queue: Fixed potential race condition when cancelling queued jobs
932 - Fixed off-by-one bug in job serial generation
933 - ``gnt-node volumes``: Fix instance names
934 - Fixed aliases in bash completion
935 - Fixed a bug in reopening log files after being sent a SIGHUP
936 - Added a flag to burnin to allow specifying VCPU count
937 - Bugfixes to non-root Ganeti configuration
943 *(Released Thu, 12 May 2011)*
945 Many bug-fixes and a few new small features:
947 - Fixed a bug related to log opening failures
948 - Fixed a bug in instance listing with orphan instances
949 - Fixed a bug which prevented resetting the cluster-level node parameter
950 ``oob_program`` to the default
951 - Many fixes related to the ``cluster-merge`` tool
952 - Fixed a race condition in the lock monitor, which caused failures
953 during (at least) creation of many instances in parallel
954 - Improved output for gnt-job info
955 - Removed the quiet flag on some ssh calls which prevented debugging
957 - Improved the N+1 failure messages in cluster verify by actually
958 showing the memory values (needed and available)
959 - Increased lock attempt timeouts so that when executing long operations
960 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
961 too early and thus prevent the use of the 'fair' mechanism
962 - Changed instance query data (``gnt-instance info``) to not acquire
963 locks unless needed, thus allowing its use on locked instance if only
964 static information is asked for
965 - Improved behaviour with filesystems that do not support rename on an
967 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
968 kept locks on all nodes during the wipe, which is unneeded
969 - Fixed ``gnt-watcher`` handling of errors during hooks execution
970 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
971 10GiB) which caused the wipe to fail right at the end in some cases
972 - Fixed master IP activation when doing master failover with no-voting
973 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
974 the master node itself
975 - Fixed potential data-loss in under disk full conditions, where Ganeti
976 wouldn't check correctly the return code and would consider
977 partially-written files 'correct'
978 - Fixed bug related to multiple VGs and DRBD disk replacing
979 - Added new disk parameter ``metavg`` that allows placement of the meta
980 device for DRBD in a different volume group
981 - Fixed error handling in the node daemon when the system libc doesn't
982 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
983 - Fixed lock release during replace-disks, which kept cluster-wide locks
984 when doing disk replaces with an iallocator script
985 - Added check for missing bridges in cluster verify
986 - Handle EPIPE errors while writing to the terminal better, so that
987 piping the output to e.g. ``less`` doesn't cause a backtrace
988 - Fixed rare case where a ^C during Luxi calls could have been
989 interpreted as server errors, instead of simply terminating
990 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
992 - Added a few more parameters to the KVM hypervisor, allowing a second
993 CDROM, custom disk type for CDROMs and a floppy image
994 - Removed redundant message in instance rename when the name is given
996 - Added option to ``gnt-instance recreate-disks`` to allow creating the
997 disks on new nodes, allowing recreation when the original instance
998 nodes are completely gone
999 - Added option when converting disk templates to DRBD to skip waiting
1000 for the resync, in order to make the instance available sooner
1001 - Added two new variables to the OS scripts environment (containing the
1003 - Made the root_path and optional parameter for the xen-pvm hypervisor,
1004 to allow use of ``pvgrub`` as bootloader
1005 - Changed the instance memory modifications to only check out-of-memory
1006 conditions on memory increases, and turned the secondary node warnings
1007 into errors (they can still be overridden via ``--force``)
1008 - Fixed the handling of a corner case when the Python installation gets
1009 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
1010 to execute a command that doesn't exist
1011 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
1012 node of the instance returned failures during instance shutdown; this
1013 adds the option ``--ignore-consistency`` to gnt-instance move
1015 And as usual, various improvements to the error messages, documentation
1022 *(Released Wed, 09 Mar 2011)*
1024 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
1025 the RAPI users file if run twice (even with ``--dry-run``).
1027 The release fixes that bug (nothing else changed).
1033 *(Released Mon, 07 Mar 2011)*
1035 Final 2.4.0 release. Just a few small fixes:
1037 - Fixed RAPI node evacuate
1038 - Fixed the kvm-ifup script
1039 - Fixed internal error handling for special job cases
1040 - Updated man page to specify the escaping feature for options
1046 *(Released Mon, 28 Feb 2011)*
1048 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
1049 that this feature wiped the disks of the wrong instance, leading to loss
1054 - Fixed title of query field containing instance name
1055 - Expanded the glossary in the documentation
1056 - Fixed one unittest (internal issue)
1062 *(Released Mon, 21 Feb 2011)*
1064 A number of bug fixes plus just a couple functionality changes.
1066 On the user-visible side, the ``gnt-* list`` command output has changed
1067 with respect to "special" field states. The current rc1 style of display
1068 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
1069 the default output mode special fields states are displayed as follows:
1071 - Offline resource: ``*``
1072 - Unavailable/not applicable: ``-``
1073 - Data missing (RPC failure): ``?``
1074 - Unknown field: ``??``
1076 Another user-visible change is the addition of ``--force-join`` to
1081 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
1082 - Fixed regression in RAPI/instance reinstall where all parameters were
1083 required (instead of optional)
1084 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
1085 - Fixed iallocator usage (offline nodes were not considered offline)
1086 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
1087 - Fixed hypervisor and OS parameter validation with respect to
1088 non-vm_capable nodes
1089 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
1091 - Fixed ``tools/listrunner`` with respect to agent-based usage
1097 *(Released Fri, 4 Feb 2011)*
1099 Many changes and fixes since the beta1 release. While there were some
1100 internal changes, the code has been mostly stabilised for the RC
1103 Note: the dumb allocator was removed in this release, as it was not kept
1104 up-to-date with the IAllocator protocol changes. It is recommended to
1105 use the ``hail`` command from the ganeti-htools package.
1107 Note: the 2.4 and up versions of Ganeti are not compatible with the
1108 0.2.x branch of ganeti-htools. You need to upgrade to
1109 ganeti-htools-0.3.0 (or later).
1111 Regressions fixed from 2.3
1112 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1114 - Fixed the ``gnt-cluster verify-disks`` command
1115 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
1117 - Fixed disk adoption breakage
1118 - Fixed wrong headers in instance listing for field aliases
1123 - Fixed corner case in KVM handling of NICs
1124 - Fixed many cases of wrong handling of non-vm_capable nodes
1125 - Fixed a bug where a missing instance symlink was not possible to
1126 recreate with any ``gnt-*`` command (now ``gnt-instance
1127 activate-disks`` does it)
1128 - Fixed the volume group name as reported by ``gnt-cluster
1130 - Increased timeouts for the import-export code, hopefully leading to
1131 fewer aborts due network or instance timeouts
1132 - Fixed bug in ``gnt-node list-storage``
1133 - Fixed bug where not all daemons were started on cluster
1134 initialisation, but only at the first watcher run
1135 - Fixed many bugs in the OOB implementation
1136 - Fixed watcher behaviour in presence of instances with offline
1138 - Fixed instance list output for instances running on the wrong node
1139 - a few fixes to the cluster-merge tool, but it still cannot merge
1140 multi-node groups (currently it is not recommended to use this tool)
1146 - Improved network configuration for the KVM hypervisor
1147 - Added e1000 as a supported NIC for Xen-HVM
1148 - Improved the lvmstrap tool to also be able to use partitions, as
1149 opposed to full disks
1150 - Improved speed of disk wiping (the cluster parameter
1151 ``prealloc_wipe_disks``, so that it has a low impact on the total time
1152 of instance creations
1153 - Added documentation for the OS parameters
1154 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
1155 hypervisor is not responding
1156 - Added display of blacklisted and hidden OS information in
1157 ``gnt-cluster info``
1158 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
1159 NIC and node parameters, which might create problems with currently
1160 invalid (but undetected) configuration files, but prevents validation
1161 failures when unrelated parameters are modified
1162 - Changed cluster initialisation to wait for the master daemon to become
1164 - Expanded the RAPI interface:
1166 - Added config redistribution resource
1167 - Added activation/deactivation of instance disks
1168 - Added export of console information
1170 - Implemented log file reopening on SIGHUP, which allows using
1171 logrotate(8) for the Ganeti log files
1172 - Added a basic OOB helper script as an example
1178 *(Released Fri, 14 Jan 2011)*
1183 - Fixed timezone issues when formatting timestamps
1184 - Added support for node groups, available via ``gnt-group`` and other
1186 - Added out-of-band framework and management, see :doc:`design
1187 document <design-oob>`
1188 - Removed support for roman numbers from ``gnt-node list`` and
1189 ``gnt-instance list``.
1190 - Allowed modification of master network interface via ``gnt-cluster
1191 modify --master-netdev``
1192 - Accept offline secondaries while shutting down instance disks
1193 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
1194 - Added support for multiple LVM volume groups
1195 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
1197 - Added commands to list available fields:
1199 - ``gnt-node list-fields``
1200 - ``gnt-group list-fields``
1201 - ``gnt-instance list-fields``
1203 - Updated documentation and man pages
1208 - Moved ``rapi_users`` file into separate directory, now named
1209 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
1211 - Added new tool for running commands on many machines,
1212 ``tools/ganeti-listrunner``
1213 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
1214 improving the ``gnt-instance console`` output
1215 - Allowed customisation of disk index separator at ``configure`` time
1216 - Export node group allocation policy to :doc:`iallocator <iallocator>`
1217 - Added support for non-partitioned md disks in ``lvmstrap``
1218 - Added script to gracefully power off KVM instances
1219 - Split ``utils`` module into smaller parts
1220 - Changed query operations to return more detailed information, e.g.
1221 whether an information is unavailable due to an offline node. To use
1222 this new functionality, the LUXI call ``Query`` must be used. Field
1223 information is now stored by the master daemon and can be retrieved
1224 using ``QueryFields``. Instances, nodes and groups can also be queried
1225 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
1226 exposed via RAPI). The following commands make use of this
1227 infrastructure change:
1229 - ``gnt-group list``
1230 - ``gnt-group list-fields``
1232 - ``gnt-node list-fields``
1233 - ``gnt-instance list``
1234 - ``gnt-instance list-fields``
1235 - ``gnt-debug locks``
1240 - New RAPI resources (see :doc:`rapi`):
1244 - ``/2/groups/[group_name]``
1245 - ``/2/groups/[group_name]/assign-nodes``
1246 - ``/2/groups/[group_name]/modify``
1247 - ``/2/groups/[group_name]/rename``
1248 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
1252 - Implemented ``no_install`` for instance creation
1253 - Implemented OS parameters for instance reinstallation, allowing
1254 use of special settings on reinstallation (e.g. for preserving data)
1259 - Added IPv6 support in import/export
1260 - Pause DRBD synchronization while wiping disks on instance creation
1261 - Updated unittests and QA scripts
1262 - Improved network parameters passed to KVM
1263 - Converted man pages from docbook to reStructuredText
1269 *(Released Mon, 20 Dec 2010)*
1271 Released version 2.3.1~rc1 without any changes.
1277 *(Released Wed, 1 Dec 2010)*
1279 - impexpd: Disable OpenSSL compression in socat if possible (backport
1280 from master, commit e90739d625b, see :doc:`installation guide
1281 <install-quick>` for details)
1282 - Changed unittest coverage report to exclude test scripts
1283 - Added script to check version format
1289 *(Released Wed, 1 Dec 2010)*
1291 Released version 2.3.0~rc1 without any changes.
1297 *(Released Fri, 19 Nov 2010)*
1299 A number of bugfixes and documentation updates:
1301 - Update ganeti-os-interface documentation
1302 - Fixed a bug related to duplicate MACs or similar items which should be
1304 - Fix breakage in OS state modify
1305 - Reinstall instance: disallow offline secondaries (fixes bug related to
1306 OS changing but reinstall failing)
1307 - plus all the other fixes between 2.2.1 and 2.2.2
1313 *(Released Tue, 2 Nov 2010)*
1315 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
1316 - Fixed master failover race with watcher
1317 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
1319 - Accept previously stopped instance for export with instance removal
1320 - Simplify and extend the environment variables for instance OS scripts
1321 - Added new node flags, ``master_capable`` and ``vm_capable``
1322 - Added optional instance disk wiping prior during allocation. This is a
1323 cluster-wide option and can be set/modified using
1324 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
1325 - Added IPv6 support, see :doc:`design document <design-2.3>` and
1326 :doc:`install-quick`
1327 - Added a new watcher option (``--ignore-pause``)
1328 - Added option to ignore offline node on instance start/stop
1329 (``--ignore-offline``)
1330 - Allow overriding OS parameters with ``gnt-instance reinstall``
1331 - Added ability to change node's secondary IP address using ``gnt-node
1333 - Implemented privilege separation for all daemons except
1334 ``ganeti-noded``, see ``configure`` options
1335 - Complain if an instance's disk is marked faulty in ``gnt-cluster
1337 - Implemented job priorities (see ``ganeti(7)`` manpage)
1338 - Ignore failures while shutting down instances during failover from
1340 - Exit daemon's bootstrap process only once daemon is ready
1341 - Export more information via ``LUInstanceQuery``/remote API
1342 - Improved documentation, QA and unittests
1343 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
1344 restart if the file was created or changed
1345 - Added LUXI protocol version sent with each request and response,
1346 allowing detection of server/client mismatches
1347 - Moved the Python scripts among gnt-* and ganeti-* into modules
1348 - Moved all code related to setting up SSH to an external script,
1350 - Infrastructure changes for node group support in future versions
1356 *(Released Fri, 19 Nov 2010)*
1358 A few small bugs fixed, and some improvements to the build system:
1360 - Fix documentation regarding conversion to drbd
1361 - Fix validation of parameters in cluster modify (``gnt-cluster modify
1363 - Fix error handling in node modify with multiple changes
1364 - Allow remote imports without checked names
1370 *(Released Tue, 19 Oct 2010)*
1372 - Disable SSL session ID cache in RPC client
1378 *(Released Thu, 14 Oct 2010)*
1380 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
1381 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
1383 - Fix problems with interaction between Curl and Python's HTTP server,
1384 resulting in increased speed in many RPC calls
1385 - Improve our release script to prevent breakage with older aclocal and
1392 *(Released Thu, 7 Oct 2010)*
1394 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
1395 value retrieved from master
1396 - Added support for blacklisted or hidden OS definitions
1397 - Added simple lock monitor (accessible via (``gnt-debug locks``)
1398 - Added support for -mem-path in KVM hypervisor abstraction layer
1399 - Allow overriding instance parameters in tool for inter-cluster
1400 instance moves (``tools/move-instance``)
1401 - Improved opcode summaries (e.g. in ``gnt-job list``)
1402 - Improve consistency of OS listing by sorting it
1403 - Documentation updates
1409 *(Released Fri, 8 Oct 2010)*
1411 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
1417 *(Released Mon, 4 Oct 2010)*
1419 - Fixed regression in ``gnt-instance rename``
1425 *(Released Wed, 22 Sep 2010)*
1427 - Fixed OS_VARIANT variable for OS scripts
1428 - Fixed cluster tag operations via RAPI
1429 - Made ``setup-ssh`` exit with non-zero code if an error occurred
1430 - Disabled RAPI CA checks in watcher
1436 *(Released Mon, 23 Aug 2010)*
1438 - Support DRBD versions of the format "a.b.c.d"
1440 - Re-introduce support for usage from multiple threads in RAPI client
1441 - Instance renames and modify via RAPI
1442 - Work around race condition between processing and archival in job
1444 - Mark opcodes following failed one as failed, too
1445 - Job field ``lock_status`` was removed due to difficulties making it
1446 work with the changed job queue in Ganeti 2.2; a better way to monitor
1447 locks is expected for a later 2.2.x release
1448 - Fixed dry-run behaviour with many commands
1449 - Support ``ssh-agent`` again when adding nodes
1450 - Many additional bugfixes
1456 *(Released Fri, 30 Jul 2010)*
1458 Important change: the internal RPC mechanism between Ganeti nodes has
1459 changed from using a home-grown http library (based on the Python base
1460 libraries) to use the PycURL library. This requires that PycURL is
1461 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
1462 against GnuTLS by default. cURL's support for GnuTLS had known issues
1463 before cURL 7.21.0 and we recommend using the latest cURL release or
1464 linking against OpenSSL. Most other distributions already link PycURL
1465 and cURL against OpenSSL. The command::
1467 python -c 'import pycurl; print pycurl.version'
1469 can be used to determine the libraries PycURL and cURL are linked
1472 Other significant changes:
1474 - Rewrote much of the internals of the job queue, in order to achieve
1475 better parallelism; this decouples job query operations from the job
1476 processing, and it should allow much nicer behaviour of the master
1477 daemon under load, and it also has uncovered some long-standing bugs
1478 related to the job serialisation (now fixed)
1479 - Added a default iallocator setting to the cluster parameters,
1480 eliminating the need to always pass nodes or an iallocator for
1481 operations that require selection of new node(s)
1482 - Added experimental support for the LXC virtualization method
1483 - Added support for OS parameters, which allows the installation of
1484 instances to pass parameter to OS scripts in order to customise the
1486 - Added a hypervisor parameter controlling the migration type (live or
1487 non-live), since hypervisors have various levels of reliability; this
1488 has renamed the 'live' parameter to 'mode'
1489 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
1490 logical volumes, meaning that cluster verify will ignore them and not
1491 flag their presence as errors
1492 - The watcher will now reset the error count for failed instances after
1493 8 hours, thus allowing self-healing if the problem that caused the
1494 instances to be down/fail to start has cleared in the meantime
1495 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
1496 check for, and warn, if the drbd module parameter ``usermode_helper``
1497 is not consistent with the cluster-wide setting; this is needed to
1498 make diagnose easier of failed drbd creations
1499 - Started adding base IPv6 support, but this is not yet
1500 enabled/available for use
1501 - Rename operations (cluster, instance) will now return the new name,
1502 which is especially useful if a short name was passed in
1503 - Added support for instance migration in RAPI
1504 - Added a tool to pre-configure nodes for the SSH setup, before joining
1505 them to the cluster; this will allow in the future a simplified model
1506 for node joining (but not yet fully enabled in 2.2); this needs the
1507 paramiko python library
1508 - Fixed handling of name-resolving errors
1509 - Fixed consistency of job results on the error path
1510 - Fixed master-failover race condition when executed multiple times in
1512 - Fixed many bugs related to the job queue (mostly introduced during the
1513 2.2 development cycle, so not all are impacting 2.1)
1514 - Fixed instance migration with missing disk symlinks
1515 - Fixed handling of unknown jobs in ``gnt-job archive``
1516 - And many other small fixes/improvements
1520 - Enhanced both the unittest and the QA coverage
1521 - Switched the opcode validation to a generic model, and extended the
1522 validation to all opcode parameters
1523 - Changed more parts of the code that write shell scripts to use the
1525 - Switched the master daemon to use the asyncore library for the Luxi
1532 *(Released Thu, 17 Jun 2010)*
1534 - Added tool (``move-instance``) and infrastructure to move instances
1535 between separate clusters (see :doc:`separate documentation
1536 <move-instance>` and :doc:`design document <design-2.2>`)
1537 - Added per-request RPC timeout
1538 - RAPI now requires a Content-Type header for requests with a body (e.g.
1539 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
1540 :rfc:`2616` (HTTP/1.1), section 7.2.1)
1541 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
1543 - Implemented initial support for running Ganeti daemons as separate
1544 users, see configure-time flags ``--with-user-prefix`` and
1545 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
1547 - Instances can be removed after export (``gnt-backup export
1548 --remove-instance``)
1549 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
1550 key (instead of 1024 bit)
1551 - Added new cluster configuration file for cluster domain secret
1552 - Import/export now use SSL instead of SSH
1553 - Added support for showing estimated time when exporting an instance,
1554 see the ``ganeti-os-interface(7)`` manpage and look for
1561 *(Released Tue, 16 Nov 2010)*
1563 Some more bugfixes. Unless critical bugs occur, this will be the last
1566 - Fix case of MAC special-values
1567 - Fix mac checker regex
1568 - backend: Fix typo causing "out of range" error
1569 - Add missing --units in gnt-instance list man page
1575 *(Released Tue, 24 Aug 2010)*
1578 - Don't ignore secondary node silently on non-mirrored disk templates
1580 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
1581 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
1582 - Properly document the usb_mouse parameter
1583 - Fix path in ganeti-rapi(8) (issue 116)
1584 - Adjust error message when the ganeti user's .ssh directory is
1586 - Add same-node-check when changing the disk template to drbd
1592 *(Released Fri, 16 Jul 2010)*
1595 - Add an option to only select some reboot types during qa/burnin.
1596 (on some hypervisors consequent reboots are not supported)
1597 - Fix infrequent race condition in master failover. Sometimes the old
1598 master ip address would be still detected as up for a short time
1599 after it was removed, causing failover to fail.
1600 - Decrease mlockall warnings when the ctypes module is missing. On
1601 Python 2.4 we support running even if no ctypes module is installed,
1602 but we were too verbose about this issue.
1603 - Fix building on old distributions, on which man doesn't have a
1605 - Fix RAPI not to ignore the MAC address on instance creation
1606 - Implement the old instance creation format in the RAPI client.
1612 *(Released Thu, 01 Jul 2010)*
1614 A small bugfix release:
1615 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
1616 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
1617 a non-existing option
1618 - Fix instance create: the --force-variant option was ignored
1619 - Improve pylint 0.21 compatibility and warnings with Python 2.6
1620 - Fix modify node storage with non-FQDN arguments
1621 - Fix RAPI client to authenticate under Python 2.6 when used
1622 for more than 5 requests needing authentication
1623 - Fix gnt-instance modify -t (storage) giving a wrong error message
1624 when converting a non-shutdown drbd instance to plain
1630 *(Released Fri, 18 Jun 2010)*
1632 A small bugfix release:
1634 - Fix live migration of KVM instances started with older Ganeti
1635 versions which had fewer hypervisor parameters
1636 - Fix gnt-instance grow-disk on down instances
1637 - Fix an error-reporting bug during instance migration
1638 - Better checking of the ``--net`` and ``--disk`` values, to avoid
1639 silently ignoring broken ones
1640 - Fix an RPC error reporting bug affecting, for example, RAPI client
1642 - Fix bug triggered by different API version os-es on different nodes
1643 - Fix a bug in instance startup with custom hvparams: OS level
1644 parameters would fail to be applied.
1645 - Fix the RAPI client under Python 2.6 (but more work is needed to
1646 make it work completely well with OpenSSL)
1647 - Fix handling of errors when resolving names from DNS
1653 *(Released Thu, 3 Jun 2010)*
1655 A medium sized development cycle. Some new features, and some
1656 fixes/small improvements/cleanups.
1658 Significant features
1659 ~~~~~~~~~~~~~~~~~~~~
1661 The node deamon now tries to mlock itself into memory, unless the
1662 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
1663 its logs, and falls back to console logging. This allows emergency
1664 features such as ``gnt-node powercycle`` to work even in the event of a
1665 broken node disk (tested offlining the disk hosting the node's
1666 filesystem and dropping its memory caches; don't try this at home)
1668 KVM: add vhost-net acceleration support. It can be tested with a new
1669 enough version of the kernel and of qemu-kvm.
1671 KVM: Add instance chrooting feature. If you use privilege dropping for
1672 your VMs you can also now force them to chroot to an empty directory,
1673 before starting the emulated guest.
1675 KVM: Add maximum migration bandwith and maximum downtime tweaking
1676 support (requires a new-enough version of qemu-kvm).
1678 Cluster verify will now warn if the master node doesn't have the master
1679 ip configured on it.
1681 Add a new (incompatible) instance creation request format to RAPI which
1682 supports all parameters (previously only a subset was supported, and it
1683 wasn't possible to extend the old format to accomodate all the new
1684 features. The old format is still supported, and a client can check for
1685 this feature, before using it, by checking for its presence in the
1686 ``features`` RAPI resource.
1688 Now with ancient latin support. Try it passing the ``--roman`` option to
1689 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
1690 (requires the python-roman module to be installed, in order to work).
1695 As usual many internal code refactorings, documentation updates, and
1698 - Lots of improvements and cleanups to the experimental Remote API
1699 (RAPI) client library.
1700 - A new unit test suite for the core daemon libraries.
1701 - A fix to creating missing directories makes sure the umask is not
1702 applied anymore. This enforces the same directory permissions
1704 - Better handling terminating daemons with ctrl+c (used when running
1705 them in debugging mode).
1706 - Fix a race condition in live migrating a KVM instance, when stat()
1707 on the old proc status file returned EINVAL, which is an unexpected
1709 - Fixed manpage checking with newer man and utf-8 charachters. But now
1710 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
1716 *(Released Fri, 7 May 2010)*
1718 Fix a bug which prevented untagged KVM instances from starting.
1724 *(Released Fri, 7 May 2010)*
1726 Another release with a long development cycle, during which many
1727 different features were added.
1729 Significant features
1730 ~~~~~~~~~~~~~~~~~~~~
1732 The KVM hypervisor now can run the individual instances as non-root, to
1733 reduce the impact of a VM being hijacked due to bugs in the
1734 hypervisor. It is possible to run all instances as a single (non-root)
1735 user, to manually specify a user for each instance, or to dynamically
1736 allocate a user out of a cluster-wide pool to each instance, with the
1737 guarantee that no two instances will run under the same user ID on any
1740 An experimental RAPI client library, that can be used standalone
1741 (without the other Ganeti libraries), is provided in the source tree as
1742 ``lib/rapi/client.py``. Note this client might change its interface in
1743 the future, as we iterate on its capabilities.
1745 A new command, ``gnt-cluster renew-crypto`` has been added to easily
1746 replace the cluster's certificates and crypto keys. This might help in
1747 case they have been compromised, or have simply expired.
1749 A new disk option for instance creation has been added that allows one
1750 to "adopt" currently existing logical volumes, with data
1751 preservation. This should allow easier migration to Ganeti from
1752 unmanaged (or managed via other software) instances.
1754 Another disk improvement is the possibility to convert between redundant
1755 (DRBD) and plain (LVM) disk configuration for an instance. This should
1756 allow better scalability (starting with one node and growing the
1757 cluster, or shrinking a two-node cluster to one node).
1759 A new feature that could help with automated node failovers has been
1760 implemented: if a node sees itself as offline (by querying the master
1761 candidates), it will try to shutdown (hard) all instances and any active
1762 DRBD devices. This reduces the risk of duplicate instances if an
1763 external script automatically failovers the instances on such nodes. To
1764 enable this, the cluster parameter ``maintain_node_health`` should be
1765 enabled; in the future this option (per the name) will enable other
1766 automatic maintenance features.
1768 Instance export/import now will reuse the original instance
1769 specifications for all parameters; that means exporting an instance,
1770 deleting it and the importing it back should give an almost identical
1771 instance. Note that the default import behaviour has changed from
1772 before, where it created only one NIC; now it recreates the original
1775 Cluster verify has added a few new checks: SSL certificates validity,
1776 /etc/hosts consistency across the cluster, etc.
1781 As usual, many internal changes were done, documentation fixes,
1784 - Fixed cluster initialization with disabled cluster storage (regression
1785 introduced in 2.1.1)
1786 - File-based storage supports growing the disks
1787 - Fixed behaviour of node role changes
1788 - Fixed cluster verify for some corner cases, plus a general rewrite of
1789 cluster verify to allow future extension with more checks
1790 - Fixed log spamming by watcher and node daemon (regression introduced
1792 - Fixed possible validation issues when changing the list of enabled
1794 - Fixed cleanup of /etc/hosts during node removal
1795 - Fixed RAPI response for invalid methods
1796 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
1797 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
1798 from ide disks, etc.)
1799 - Allow OS changes without re-installation (to record a changed OS
1800 outside of Ganeti, or to allow OS renames)
1801 - Allow instance creation without OS installation (useful for example if
1802 the OS will be installed manually, or restored from a backup not in
1804 - Implemented option to make cluster ``copyfile`` use the replication
1806 - Added list of enabled hypervisors to ssconf (possibly useful for
1808 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
1810 - A partial form of node re-IP is possible via node readd, which now
1811 allows changed node primary IP
1812 - Command line utilities now show an informational message if the job is
1814 - The logs of the master daemon now show the PID/UID/GID of the
1821 *(Released Fri, 12 Mar 2010)*
1823 During the 2.1.0 long release candidate cycle, a lot of improvements and
1824 changes have accumulated with were released later as 2.1.1.
1829 The node evacuate command (``gnt-node evacuate``) was significantly
1830 rewritten, and as such the IAllocator protocol was changed - a new
1831 request type has been added. This unfortunate change during a stable
1832 series is designed to improve performance of node evacuations; on
1833 clusters with more than about five nodes and which are well-balanced,
1834 evacuation should proceed in parallel for all instances of the node
1835 being evacuated. As such, any existing IAllocator scripts need to be
1836 updated, otherwise the above command will fail due to the unknown
1837 request. The provided "dumb" allocator has not been updated; but the
1838 ganeti-htools package supports the new protocol since version 0.2.4.
1840 Another important change is increased validation of node and instance
1841 names. This might create problems in special cases, if invalid host
1842 names are being used.
1844 Also, a new layer of hypervisor parameters has been added, that sits at
1845 OS level between the cluster defaults and the instance ones. This allows
1846 customisation of virtualization parameters depending on the installed
1847 OS. For example instances with OS 'X' may have a different KVM kernel
1848 (or any other parameter) than the cluster defaults. This is intended to
1849 help managing a multiple OSes on the same cluster, without manual
1850 modification of each instance's parameters.
1852 A tool for merging clusters, ``cluster-merge``, has been added in the
1853 tools sub-directory.
1858 - Improved the int/float conversions that should make the code more
1859 robust in face of errors from the node daemons
1860 - Fixed the remove node code in case of internal configuration errors
1861 - Fixed the node daemon behaviour in face of inconsistent queue
1862 directory (e.g. read-only file-system where we can't open the files
1864 - Fixed the behaviour of gnt-node modify for master candidate demotion;
1865 now it either aborts cleanly or, if given the new "auto_promote"
1866 parameter, will automatically promote other nodes as needed
1867 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
1868 completely prevent Ganeti from working
1869 - Fixed bug for instance export when not all disks were successfully
1871 - Fixed behaviour of node add when the new node is slow in starting up
1873 - Fixed handling of signals in the LUXI client, which should improve
1874 behaviour of command-line scripts
1875 - Added checks for invalid node/instance names in the configuration (now
1876 flagged during cluster verify)
1877 - Fixed watcher behaviour for disk activation errors
1878 - Fixed two potentially endless loops in http library, which led to the
1879 RAPI daemon hanging and consuming 100% CPU in some cases
1880 - Fixed bug in RAPI daemon related to hashed passwords
1881 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
1883 - Enhanced compatibility with non-Debian OSes, but not using absolute
1884 path in some commands and allowing customisation of the ssh
1885 configuration directory
1886 - Fixed possible future issue with new Python versions by abiding to the
1887 proper use of ``__slots__`` attribute on classes
1888 - Added checks that should prevent directory traversal attacks
1889 - Many documentation fixes based on feedback from users
1894 - Added an "early_release" more for instance replace disks and node
1895 evacuate, where we release locks earlier and thus allow higher
1896 parallelism within the cluster
1897 - Added watcher hooks, intended to allow the watcher to restart other
1898 daemons (e.g. from the ganeti-nbma project), but they can be used of
1899 course for any other purpose
1900 - Added a compile-time disable for DRBD barriers, to increase
1901 performance if the administrator trusts the power supply or the
1902 storage system to not lose writes
1903 - Added the option of using syslog for logging instead of, or in
1904 addition to, Ganeti's own log files
1905 - Removed boot restriction for paravirtual NICs for KVM, recent versions
1906 can indeed boot from a paravirtual NIC
1907 - Added a generic debug level for many operations; while this is not
1908 used widely yet, it allows one to pass the debug value all the way to
1910 - Enhanced the hooks environment for instance moves (failovers,
1911 migrations) where the primary/secondary nodes changed during the
1912 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
1913 - Enhanced data validations for many user-supplied values; one important
1914 item is the restrictions imposed on instance and node names, which
1915 might reject some (invalid) host names
1916 - Add a configure-time option to disable file-based storage, if it's not
1917 needed; this allows greater security separation between the master
1918 node and the other nodes from the point of view of the inter-node RPC
1920 - Added user notification in interactive tools if job is waiting in the
1921 job queue or trying to acquire locks
1922 - Added log messages when a job is waiting for locks
1923 - Added filtering by node tags in instance operations which admit
1924 multiple instances (start, stop, reboot, reinstall)
1925 - Added a new tool for cluster mergers, ``cluster-merge``
1926 - Parameters from command line which are of the form ``a=b,c=d`` can now
1927 use backslash escapes to pass in values which contain commas,
1928 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
1930 - For KVM, the instance name is the first parameter passed to KVM, so
1931 that it's more visible in the process list
1937 *(Released Tue, 2 Mar 2010)*
1939 Ganeti 2.1 brings many improvements with it. Major changes:
1941 - Added infrastructure to ease automated disk repairs
1942 - Added new daemon to export configuration data in a cheaper way than
1943 using the remote API
1944 - Instance NICs can now be routed instead of being associated with a
1946 - Improved job locking logic to reduce impact of jobs acquiring multiple
1947 locks waiting for other long-running jobs
1949 In-depth implementation details can be found in the Ganeti 2.1 design
1955 - Added chroot hypervisor
1956 - Added more options to xen-hvm hypervisor (``kernel_path`` and
1958 - Added more options to xen-pvm hypervisor (``use_bootloader``,
1959 ``bootloader_path`` and ``bootloader_args``)
1960 - Added the ``use_localtime`` option for the xen-hvm and kvm
1961 hypervisors, and the default value for this has changed to false (in
1962 2.0 xen-hvm always enabled it)
1963 - Added luxi call to submit multiple jobs in one go
1964 - Added cluster initialization option to not modify ``/etc/hosts``
1966 - Added network interface parameters
1967 - Added dry run mode to some LUs
1968 - Added RAPI resources:
1970 - ``/2/instances/[instance_name]/info``
1971 - ``/2/instances/[instance_name]/replace-disks``
1972 - ``/2/nodes/[node_name]/evacuate``
1973 - ``/2/nodes/[node_name]/migrate``
1974 - ``/2/nodes/[node_name]/role``
1975 - ``/2/nodes/[node_name]/storage``
1976 - ``/2/nodes/[node_name]/storage/modify``
1977 - ``/2/nodes/[node_name]/storage/repair``
1979 - Added OpCodes to evacuate or migrate all instances on a node
1980 - Added new command to list storage elements on nodes (``gnt-node
1981 list-storage``) and modify them (``gnt-node modify-storage``)
1982 - Added new ssconf files with master candidate IP address
1983 (``ssconf_master_candidates_ips``), node primary IP address
1984 (``ssconf_node_primary_ips``) and node secondary IP address
1985 (``ssconf_node_secondary_ips``)
1986 - Added ``ganeti-confd`` and a client library to query the Ganeti
1987 configuration via UDP
1988 - Added ability to run hooks after cluster initialization and before
1990 - Added automatic mode for disk replace (``gnt-instance replace-disks
1992 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
1993 after catastrophic data-loss
1994 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
1996 - Added ``gnt-instance move`` command to move instances
1997 - Added ``gnt-cluster watcher`` command to control watcher
1998 - Added ``gnt-node powercycle`` command to powercycle nodes
1999 - Added new job status field ``lock_status``
2000 - Added parseable error codes to cluster verification (``gnt-cluster
2001 verify --error-codes``) and made output less verbose (use
2002 ``--verbose`` to restore previous behaviour)
2003 - Added UUIDs to the main config entities (cluster, nodes, instances)
2004 - Added support for OS variants
2005 - Added support for hashed passwords in the Ganeti remote API users file
2007 - Added option to specify maximum timeout on instance shutdown
2008 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
2009 - Added new helper script to start and stop Ganeti daemons
2010 (``daemon-util``), with the intent to reduce the work necessary to
2011 adjust Ganeti for non-Debian distributions and to start/stop daemons
2013 - Added more unittests
2014 - Fixed critical bug in ganeti-masterd startup
2015 - Removed the configure-time ``kvm-migration-port`` parameter, this is
2016 now customisable at the cluster level for both the KVM and Xen
2017 hypervisors using the new ``migration_port`` parameter
2018 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
2019 reinstalling an instance
2020 - Allowed ``@`` in tag names
2021 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
2022 - Many documentation updates
2023 - Distribute hypervisor files on ``gnt-cluster redist-conf``
2024 - ``gnt-instance reinstall`` can now reinstall multiple instances
2025 - Updated many command line parameters
2026 - Introduced new OS API version 15
2027 - No longer support a default hypervisor
2028 - Treat virtual LVs as inexistent
2029 - Improved job locking logic to reduce lock contention
2030 - Match instance and node names case insensitively
2031 - Reimplemented bash completion script to be more complete
2038 *(Released Thu, 4 Feb 2010)*
2040 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
2041 - Fix a string formatting bug
2042 - Improve safety of the code in some error paths
2043 - Improve data validation in the master of values returned from nodes
2049 *(Released Thu, 17 Dec 2009)*
2051 - Fix security issue due to missing validation of iallocator names; this
2052 allows local and remote execution of arbitrary executables
2053 - Fix failure of gnt-node list during instance removal
2054 - Ship the RAPI documentation in the archive
2060 *(Released Wed, 30 Sep 2009)*
2062 - Fixed many wrong messages
2063 - Fixed a few bugs related to the locking library
2064 - Fixed MAC checking at instance creation time
2065 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
2066 - Fixed a few issues related to signal handling in both daemons and
2068 - Fixed the example startup script provided
2069 - Fixed insserv dependencies in the example startup script (patch from
2071 - Fixed handling of drained nodes in the iallocator framework
2072 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
2074 - Fixed error related to invalid job IDs in job polling
2075 - Fixed job/opcode persistence on unclean master shutdown
2076 - Fixed handling of partial job processing after unclean master
2078 - Fixed error reporting from LUs, previously all errors were converted
2079 into execution errors
2080 - Fixed error reporting from burnin
2081 - Decreased significantly the memory usage of the job queue
2082 - Optimised slightly multi-job submission
2083 - Optimised slightly opcode loading
2084 - Backported the multi-job submit framework from the development
2085 branch; multi-instance start and stop should be faster
2086 - Added script to clean archived jobs after 21 days; this will reduce
2087 the size of the queue directory
2088 - Added some extra checks in disk size tracking
2089 - Added an example ethers hook script
2090 - Added a cluster parameter that prevents Ganeti from modifying of
2092 - Added more node information to RAPI responses
2093 - Added a ``gnt-job watch`` command that allows following the ouput of a
2095 - Added a bind-address option to ganeti-rapi
2096 - Added more checks to the configuration verify
2097 - Enhanced the burnin script such that some operations can be retried
2099 - Converted instance reinstall to multi-instance model
2105 *(Released Fri, 7 Aug 2009)*
2107 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
2108 to allow using the pre-2.0.2 behaviour in activation, if any existing
2109 instances have mismatched disk sizes in the configuration
2110 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
2111 any configuration mismatches for disk sizes
2112 - Added ``gnt-master cluste-failover --no-voting`` to allow master
2113 failover to work on two-node clusters
2114 - Fixed the ``--net`` option of ``gnt-backup import``, which was
2116 - Fixed detection of OS script errors in ``gnt-backup export``
2117 - Fixed exit code of ``gnt-backup export``
2123 *(Released Fri, 17 Jul 2009)*
2125 - Added experimental support for stripped logical volumes; this should
2126 enhance performance but comes with a higher complexity in the block
2127 device handling; stripping is only enabled when passing
2128 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
2129 affected even in the non-stripped mode
2130 - Improved resiliency against transient failures at the end of DRBD
2131 resyncs, and in general of DRBD resync checks
2132 - Fixed a couple of issues with exports and snapshot errors
2133 - Fixed a couple of issues in instance listing
2134 - Added display of the disk size in ``gnt-instance info``
2135 - Fixed checking for valid OSes in instance creation
2136 - Fixed handling of the "vcpus" parameter in instance listing and in
2137 general of invalid parameters
2138 - Fixed http server library, and thus RAPI, to handle invalid
2139 username/password combinations correctly; this means that now they
2140 report unauthorized for queries too, not only for modifications,
2141 allowing earlier detect of configuration problems
2142 - Added a new "role" node list field, equivalent to the master/master
2143 candidate/drained/offline flags combinations
2144 - Fixed cluster modify and changes of candidate pool size
2145 - Fixed cluster verify error messages for wrong files on regular nodes
2146 - Fixed a couple of issues with node demotion from master candidate role
2147 - Fixed node readd issues
2148 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
2149 - Added a new ``--no-voting`` option for masterfailover to fix failover
2150 on two-nodes clusters when the former master node is unreachable
2151 - Added instance reinstall over RAPI
2157 *(Released Tue, 16 Jun 2009)*
2159 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
2160 allow re-adding the start in single-user option (regression from 1.2)
2161 - the watcher writes the instance status to a file, to allow monitoring
2162 to report the instance status (from the master) based on cached
2163 results of the watcher's queries; while this can get stale if the
2164 watcher is being locked due to other work on the cluster, this is
2165 still an improvement
2166 - the watcher now also restarts the node daemon and the rapi daemon if
2168 - fixed the watcher to handle full and drained queue cases
2169 - hooks export more instance data in the environment, which helps if
2170 hook scripts need to take action based on the instance's properties
2171 (no longer need to query back into ganeti)
2172 - instance failovers when the instance is stopped do not check for free
2173 RAM, so that failing over a stopped instance is possible in low memory
2175 - rapi uses queries for tags instead of jobs (for less job traffic), and
2176 for cluster tags it won't talk to masterd at all but read them from
2178 - a couple of error handling fixes in RAPI
2179 - drbd handling: improved the error handling of inconsistent disks after
2180 resync to reduce the frequency of "there are some degraded disks for
2181 this instance" messages
2182 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
2183 error handling path called a wrong function name)
2189 *(Released Wed, 27 May 2009)*
2191 - no changes from rc5
2197 *(Released Wed, 20 May 2009)*
2199 - fix a couple of bugs (validation, argument checks)
2200 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
2201 - some small improvements to RAPI and IAllocator
2202 - make watcher automatically start the master daemon if down
2208 *(Released Mon, 27 Apr 2009)*
2210 - change the OS list to not require locks; this helps with big clusters
2211 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
2212 volume group is broken
2213 - ``gnt-instance info``, without any arguments, doesn't run for all
2214 instances anymore; either pass ``--all`` or pass the desired
2215 instances; this helps against mistakes on big clusters where listing
2216 the information for all instances takes a long time
2217 - miscellaneous doc and man pages fixes
2223 *(Released Wed, 8 Apr 2009)*
2225 - Change the internal locking model of some ``gnt-node`` commands, in
2226 order to reduce contention (and blocking of master daemon) when
2227 batching many creation/reinstall jobs
2228 - Fixes to Xen soft reboot
2229 - No longer build documentation at build time, instead distribute it in
2230 the archive, in order to reduce the need for the whole docbook/rst
2237 *(Released Fri, 27 Mar 2009)*
2239 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
2240 - Fix watcher startup sequence, improves the behaviour of busy clusters
2241 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
2242 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
2243 bind address and other places
2244 - Some documentation fixes and updates
2250 *(Released Mon, 2 Mar 2009)*
2252 - More documentation updates, now all docs should be more-or-less
2254 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
2256 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
2258 - Fix an issue related to $libdir/run/ganeti and cluster creation
2264 *(Released Thu, 19 Feb 2009)*
2266 - Xen PVM and KVM have switched the default value for the instance root
2267 disk to the first partition on the first drive, instead of the whole
2268 drive; this means that the OS installation scripts must be changed
2270 - Man pages have been updated
2271 - RAPI has been switched by default to HTTPS, and the exported functions
2272 should all work correctly
2273 - RAPI v1 has been removed
2274 - Many improvements to the KVM hypervisor
2275 - Block device errors are now better reported
2276 - Many other bugfixes and small improvements
2282 *(Released Mon, 26 Jan 2009)*
2284 - Version 2 is a general rewrite of the code and therefore the
2285 differences are too many to list, see the design document for 2.0 in
2286 the ``doc/`` subdirectory for more details
2287 - In this beta version there is not yet a migration path from 1.2 (there
2288 will be one in the final 2.0 release)
2289 - A few significant changes are:
2291 - all commands are executed by a daemon (``ganeti-masterd``) and the
2292 various ``gnt-*`` commands are just front-ends to it
2293 - all the commands are entered into, and executed from a job queue,
2294 see the ``gnt-job(8)`` manpage
2295 - the RAPI daemon supports read-write operations, secured by basic
2296 HTTP authentication on top of HTTPS
2297 - DRBD version 0.7 support has been removed, DRBD 8 is the only
2298 supported version (when migrating from Ganeti 1.2 to 2.0, you need
2299 to migrate to DRBD 8 first while still running Ganeti 1.2)
2300 - DRBD devices are using statically allocated minor numbers, which
2301 will be assigned to existing instances during the migration process
2302 - there is support for both Xen PVM and Xen HVM instances running on
2304 - KVM virtualization is supported too
2305 - file-based storage has been implemented, which means that it is
2306 possible to run the cluster without LVM and DRBD storage, for
2307 example using a shared filesystem exported from shared storage (and
2308 still have live migration)
2314 *(Released Tue, 13 Jan 2009)*
2316 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
2317 - Reuse the old instance mac address by default on instance import, if
2318 the instance name is the same.
2319 - Handle situations in which the node info rpc returns incomplete
2321 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
2322 - Improved version of batcher:
2324 - state file support
2325 - instance mac address support
2326 - support for HVM clusters/instances
2328 - Add an option to show the number of cpu sockets and nodes in
2330 - Support OSes that handle more than one version of the OS api (but do
2331 not change the current API in any other way)
2332 - Fix ``gnt-node migrate``
2333 - ``gnt-debug`` man page
2334 - Fixes various more typos and small issues
2335 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
2341 *(Released Wed, 24 Sep 2008)*
2343 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
2344 type of disk exported to fully virtualized instances.
2345 - provide access to the serial console of HVM instances
2346 - instance auto_balance flag, set by default. If turned off it will
2347 avoid warnings on cluster verify if there is not enough memory to fail
2348 over an instance. in the future it will prevent automatically failing
2349 it over when we will support that.
2350 - batcher tool for instance creation, see ``tools/README.batcher``
2351 - ``gnt-instance reinstall --select-os`` to interactively select a new
2352 operating system when reinstalling an instance.
2353 - when changing the memory amount on instance modify a check has been
2354 added that the instance will be able to start. also warnings are
2355 emitted if the instance will not be able to fail over, if auto_balance
2357 - documentation fixes
2358 - sync fields between ``gnt-instance list/modify/add/import``
2359 - fix a race condition in drbd when the sync speed was set after giving
2360 the device a remote peer.
2366 *(Released Tue, 22 Jul 2008)*
2368 - note: the allowed size and number of tags per object were reduced
2369 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
2370 - fixed twisted 8.x compatibility
2371 - fixed ``gnt-instance replace-disks`` with iallocator
2372 - add TCP keepalives on twisted connections to detect restarted nodes
2373 - disk increase support, see ``gnt-instance grow-disk``
2374 - implement bulk node/instance query for RAPI
2375 - add tags in node/instance listing (optional)
2376 - experimental migration (and live migration) support, read the man page
2377 for ``gnt-instance migrate``
2378 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
2379 has some small improvements in handling its state file
2385 *(Released Fri, 13 Jun 2008)*
2387 - Experimental readonly, REST-based remote API implementation;
2388 automatically started on master node, TCP port 5080, if enabled by
2389 ``--enable-rapi`` parameter to configure script.
2390 - Instance allocator support. Add and import instance accept a
2391 ``--iallocator`` parameter, and call that instance allocator to decide
2392 which node to use for the instance. The iallocator document describes
2393 what's expected from an allocator script.
2394 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
2395 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
2396 node is lost there is still enough memory to fail over the instances
2398 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
2399 ``gnt-cluster verify``, to check for site-specific compliance. All the
2400 hooks will run, and their output, if any, will be displayed. Any
2401 failing hook will make the verification return an error value.
2402 - ``gnt-cluster verify`` now checks that its peers are reachable on the
2403 primary and secondary interfaces
2404 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
2405 that is still declared as part of the cluster and has failed.
2406 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
2407 specifying output fields, that just adds the chosen fields to the
2409 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
2410 export from the filesystem.
2411 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
2412 have been added. Using them you can enable/disable acpi and pae
2413 support, and specify a path for a cd image to be exported to the
2414 instance. These parameters as the name suggest only work on HVM
2416 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
2417 PAE support will be set to the previously hardcoded values, but the
2418 (previously hardcoded) path to the CDROM ISO image will be unset and
2419 if required, needs to be set manually with ``gnt-instance modify``
2421 - The address to which an instance's VNC console is bound is now
2422 selectable per-instance, rather than being cluster wide. Of course
2423 this only applies to instances controlled via VNC, so currently just
2424 applies to HVM clusters.
2430 *(Released Mon, 18 Feb 2008)*
2432 - more tweaks to the disk activation code (especially helpful for DRBD)
2433 - change the default ``gnt-instance list`` output format, now there is
2434 one combined status field (see the manpage for the exact values this
2436 - some more fixes for the mac export to hooks change
2437 - make Ganeti not break with DRBD 8.2.x (which changed the version
2438 format in ``/proc/drbd``) (issue 24)
2439 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
2440 template, allowing migration from DRBD0.7+MD to DRBD8
2446 *(Released Wed, 30 Jan 2008)*
2448 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
2450 - add command aliases infrastructure and a few aliases
2451 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
2452 man pages and the ``--help`` option of ``gnt-node
2453 list``/``gnt-instance list``
2454 - fix ``gnt-backup list`` with down nodes (issue 21)
2455 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
2456 - fix the dist archive and add a check for including svn/git files in
2458 - some developer-related changes: improve the burnin and the QA suite,
2459 add an upload script for testing during development
2465 *(Released Wed, 16 Jan 2008)*
2467 - experimental HVM support, read the install document, section
2468 "Initializing the cluster"
2469 - allow for the PVM hypervisor per-instance kernel and initrd paths
2470 - add a new command ``gnt-cluster verify-disks`` which uses a new
2471 algorithm to improve the reconnection of the DRBD pairs if the device
2472 on the secondary node has gone away
2473 - make logical volume code auto-activate LVs at disk activation time
2474 - slightly improve the speed of activating disks
2475 - allow specification of the MAC address at instance creation time, and
2476 changing it later via ``gnt-instance modify``
2477 - fix handling of external commands that generate lots of output on
2479 - update documentation with regard to minimum version of DRBD8 supported
2485 *(Released Tue, 4 Dec 2007)*
2487 - Log the ``xm create`` output to the node daemon log on failure (to
2488 help diagnosing the error)
2489 - In debug mode, log all external commands output if failed to the logs
2490 - Change parsing of lvm commands to ignore stderr
2496 *(Released Wed, 28 Nov 2007)*
2498 - Another round of updates to the DRBD 8 code to deal with more failures
2499 in the replace secondary node operation
2500 - Some more logging of failures in disk operations (lvm, drbd)
2501 - A few documentation updates
2508 *(Released Tue, 13 Nov 2007)*
2510 - Change configuration file format from Python's Pickle to JSON.
2511 Upgrading is possible using the cfgupgrade utility.
2512 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
2513 faster replace disks and is more stable (DRBD 8 has many improvements
2514 compared to DRBD 0.7)
2515 - Added command line tags support (see man pages for ``gnt-instance``,
2516 ``gnt-node``, ``gnt-cluster``)
2517 - Added instance rename support
2518 - Added multi-instance startup/shutdown
2519 - Added cluster rename support
2520 - Added ``gnt-node evacuate`` to simplify some node operations
2521 - Added instance reboot operation that can speedup reboot as compared to
2523 - Soften the requirement that hostnames are in FQDN format
2524 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
2526 - Removed dependency on debian's patched fping that uses the
2527 non-standard ``-S`` option
2528 - Now the OS definitions are searched for in multiple, configurable
2529 paths (easier for distros to package)
2530 - Some changes to the hooks infrastructure (especially the new
2531 post-configuration update hook)
2532 - Other small bugfixes
2534 .. vim: set textwidth=72 syntax=rst :