8 *(Released Thu, 19 Jul 2012)*
16 The current ``admin_up`` field, which used to denote whether an instance
17 should be running or not, has been removed. Instead, ``admin_state`` is
18 introduced, with 3 possible values -- ``up``, ``down`` and ``offline``.
20 The rational behind this is that an instance being “down” can have
23 - it could be down during a reboot
24 - it could be temporarily be down for a reinstall
25 - or it could be down because it is deprecated and kept just for its
28 The previous Boolean state was making it difficult to do capacity
29 calculations: should Ganeti reserve memory for a down instance? Now, the
30 tri-state field makes it clear:
32 - in ``up`` and ``down`` state, all resources are reserved for the
33 instance, and it can be at any time brought up if it is down
34 - in ``offline`` state, only disk space is reserved for it, but not
37 The field can have an extra use: since the transition between ``up`` and
38 ``down`` and vice-versus is done via ``gnt-instance start/stop``, but
39 transition between ``offline`` and ``down`` is done via ``gnt-instance
40 modify``, it is possible to given different rights to users. For
41 example, owners of an instance could be allowed to start/stop it, but
42 not transition it out of the offline state.
44 Instance policies and specs
45 +++++++++++++++++++++++++++
47 In previous Ganeti versions, an instance creation request was not
48 limited on the minimum size and on the maximum size just by the cluster
49 resources. As such, any policy could be implemented only in third-party
50 clients (RAPI clients, or shell wrappers over ``gnt-*``
51 tools). Furthermore, calculating cluster capacity via ``hspace`` again
52 required external input with regards to instance sizes.
54 In order to improve these workflows and to allow for example better
55 per-node group differentiation, we introduced instance specs, which
58 - minimum instance disk size, disk count, memory size, cpu count
59 - maximum values for the above metrics
60 - and “standard” values (used in ``hspace`` to calculate the standard
63 The minimum/maximum values can be also customised at node-group level,
64 for example allowing more powerful hardware to support bigger instance
67 Beside the instance specs, there are a few other settings belonging to
68 the instance policy framework. It is possible now to customise, per
69 cluster and node-group:
71 - the list of allowed disk templates
72 - the maximum ratio of VCPUs per PCPUs (to control CPU oversubscription)
73 - the maximum ratio of instance to spindles (see below for more
74 information) for local storage
76 All these together should allow all tools that talk to Ganeti to know
77 what are the ranges of allowed values for instances and the
78 over-subscription that is allowed.
80 For the VCPU/PCPU ratio, we already have the VCPU configuration from the
81 instance configuration, and the physical CPU configuration from the
82 node. For the spindle ratios however, we didn't track before these
83 values, so new parameters have been added:
85 - a new node parameter ``spindle_count``, defaults to 1, customisable at
86 node group or node level
87 - at new backend parameter (for instances), ``spindle_use`` defaults to 1
89 Note that spindles in this context doesn't need to mean actual
90 mechanical hard-drives; it's just a relative number for both the node
91 I/O capacity and instance I/O consumption.
93 Instance migration behaviour
94 ++++++++++++++++++++++++++++
96 While live-migration is in general desirable over failover, it is
97 possible that for some workloads it is actually worse, due to the
98 variable time of the “suspend” phase during live migration.
100 To allow the tools to work consistently over such instances (without
101 having to hard-code instance names), a new backend parameter
102 ``always_failover`` has been added to control the migration/failover
103 behaviour. When set to True, all migration requests for an instance will
104 instead fall-back to failover.
106 Instance memory ballooning
107 ++++++++++++++++++++++++++
109 Initial support for memory ballooning has been added. The memory for an
110 instance is no longer fixed (backend parameter ``memory``), but instead
111 can vary between minimum and maximum values (backend parameters
112 ``minmem`` and ``maxmem``). Currently we only change an instance's
115 - live migrating or failing over and instance and the target node
116 doesn't have enough memory
117 - user requests changing the memory via ``gnt-instance modify
123 In order to control the use of specific CPUs by instance, support for
124 controlling CPU pinning has been added for the Xen, HVM and LXC
125 hypervisors. This is controlled by a new hypervisor parameter
126 ``cpu_mask``; details about possible values for this are in the
127 :manpage:`gnt-instance(8)`. Note that use of the most specific (precise
128 VCPU-to-CPU mapping) form will work well only when all nodes in your
129 cluster have the same amount of CPUs.
134 Another area in which Ganeti was not customisable were the parameters
135 used for storage configuration, e.g. how many stripes to use for LVM,
136 DRBD resync configuration, etc.
138 To improve this area, we've added disks parameters, which are
139 customisable at cluster and node group level, and which allow to
140 specify various parameters for disks (DRBD has the most parameters
141 currently), for example:
143 - DRBD resync algorithm and parameters (e.g. speed)
144 - the default VG for meta-data volumes for DRBD
145 - number of stripes for LVM (plain disk template)
148 These parameters can be modified via ``gnt-cluster modify -D …`` and
149 ``gnt-group modify -D …``, and are used at either instance creation (in
150 case of LVM stripes, for example) or at disk “activation” time
153 Rados block device support
154 ++++++++++++++++++++++++++
156 A Rados (http://ceph.com/wiki/Rbd) storage backend has been added,
157 denoted by the ``rbd`` disk template type. This is considered
158 experimental, feedback is welcome. For details on configuring it, see
159 the :doc:`install` document and the :manpage:`gnt-cluster(8)` man page.
164 The existing master IP functionality works well only in simple setups (a
165 single network shared by all nodes); however, if nodes belong to
166 different networks, then the ``/32`` setup and lack of routing
167 information is not enough.
169 To allow the master IP to function well in more complex cases, the
170 system was reworked as follows:
172 - a master IP netmask setting has been added
173 - the master IP activation/turn-down code was moved from the node daemon
175 - whether to run the Ganeti-supplied master IP script or a user-supplied
176 on is a ``gnt-cluster init`` setting
178 Details about the location of the standard and custom setup scripts are
179 in the man page :manpage:`gnt-cluster(8)`; for information about the
180 setup script protocol, look at the Ganeti-supplied script.
185 The `SPICE <http://www.linux-kvm.org/page/SPICE>`_ support has been
188 It is now possible to use TLS-protected connections, and when renewing
189 or changing the cluster certificates (via ``gnt-cluster renew-crypto``,
190 it is now possible to specify spice or spice CA certificates. Also, it
191 is possible to configure a password for SPICE sessions via the
192 hypervisor parameter ``spice_password_file``.
194 There are also new parameters to control the compression and streaming
195 options (e.g. ``spice_image_compression``, ``spice_streaming_video``,
196 etc.). For details, see the man page :manpage:`gnt-instance(8)` and look
197 for the spice parameters.
199 Lastly, it is now possible to see the SPICE connection information via
200 ``gnt-instance console``.
205 A new tool (``tools/ovfconverter``) has been added that supports
206 conversion between Ganeti and the `Open Virtualization Format
207 <http://en.wikipedia.org/wiki/Open_Virtualization_Format>`_ (both to and
210 This relies on the ``qemu-img`` tool to convert the disk formats, so the
211 actual compatibility with other virtualization solutions depends on it.
216 The configuration query daemon (``ganeti-confd``) is now optional, and
217 has been rewritten in Haskell; whether to use the daemon at all, use the
218 Python (default) or the Haskell version is selectable at configure time
219 via the ``--enable-confd`` parameter, which can take one of the
220 ``haskell``, ``python`` or ``no`` values. If not used, disabling the
221 daemon will result in a smaller footprint; for larger systems, we
222 welcome feedback on the Haskell version which might become the default
225 If you want to use ``gnt-node list-drbd`` you need to have the Haskell
226 daemon running. The Python version doesn't implement the new call.
229 User interface changes
230 ~~~~~~~~~~~~~~~~~~~~~~
232 We have replaced the ``--disks`` option of ``gnt-instance
233 replace-disks`` with a more flexible ``--disk`` option, which allows
234 adding and removing disks at arbitrary indices (Issue 188). Furthermore,
235 disk size and mode can be changed upon recreation (via ``gnt-instance
236 recreate-disks``, which accepts the same ``--disk`` option).
238 As many people are used to a ``show`` command, we have added that as an
239 alias to ``info`` on all ``gnt-*`` commands.
241 The ``gnt-instance grow-disk`` command has a new mode in which it can
242 accept the target size of the disk, instead of the delta; this can be
243 more safe since two runs in absolute mode will be idempotent, and
244 sometimes it's also easier to specify the desired size directly.
246 Also the handling of instances with regard to offline secondaries has
247 been improved. Instance operations should not fail because one of it's
248 secondary nodes is offline, even though it's safe to proceed.
250 A new command ``list-drbd`` has been added to the ``gnt-node`` script to
251 support debugging of DRBD issues on nodes. It provides a mapping of DRBD
252 minors to instance name.
257 RAPI coverage has improved, with (for example) new resources for
258 recreate-disks, node power-cycle, etc.
263 There is partial support for ``xl`` in the Xen hypervisor; feedback is
266 Python 2.7 is better supported, and after Ganeti 2.6 we will investigate
267 whether to still support Python 2.4 or move to Python 2.6 as minimum
270 Support for Fedora has been slightly improved; the provided example
271 init.d script should work better on it and the INSTALL file should
272 document the needed dependencies.
277 The deprecated ``QueryLocks`` LUXI request has been removed. Use
278 ``Query(what=QR_LOCK, ...)`` instead.
280 The LUXI requests :pyeval:`luxi.REQ_QUERY_JOBS`,
281 :pyeval:`luxi.REQ_QUERY_INSTANCES`, :pyeval:`luxi.REQ_QUERY_NODES`,
282 :pyeval:`luxi.REQ_QUERY_GROUPS`, :pyeval:`luxi.REQ_QUERY_EXPORTS` and
283 :pyeval:`luxi.REQ_QUERY_TAGS` are deprecated and will be removed in a
284 future version. :pyeval:`luxi.REQ_QUERY` should be used instead.
286 RAPI client: ``CertificateError`` now derives from
287 ``GanetiApiError``. This should make it more easy to handle Ganeti
290 Deprecation warnings due to PyCrypto/paramiko import in
291 ``tools/setup-ssh`` have been silenced, as usually they are safe; please
292 make sure to run an up-to-date paramiko version, if you use this tool.
294 The QA scripts now depend on Python 2.5 or above (the main code base
295 still works with Python 2.4).
297 The configuration file (``config.data``) is now written without
298 indentation for performance reasons; if you want to edit it, it can be
299 re-formatted via ``tools/fmtjson``.
301 A number of bugs has been fixed in the cluster merge tool.
303 ``x509`` certification verification (used in import-export) has been
304 changed to allow the same clock skew as permitted by the cluster
305 verification. This will remove some rare but hard to diagnose errors in
308 The ``LUXI`` protocol has been made more consistent regarding its
309 handling of command arguments. This, however, leads to incompatibility
310 issues with previous versions. Please ensure that you restart Ganeti
311 daemons after the upgrade, otherwise job submission will fail.
317 *(Released Fri, 13 Jul 2012)*
319 Third release candidate for 2.6. The following changes were done from
322 - Fixed ``UpgradeConfig`` w.r.t. to disk parameters on disk objects.
323 - Fixed an inconsistency in the LUXI protocol with the provided
324 arguments (NOT backwards compatible)
325 - Fixed a bug with node groups ipolicy where ``min`` was greater than
326 the cluster ``std`` value
327 - Implemented a new ``gnt-node list-drbd`` call to list DRBD minors for
328 easier instance debugging on nodes (requires ``hconfd`` to work)
334 *(Released Tue, 03 Jul 2012)*
336 Second release candidate for 2.6. The following changes were done from
339 - Fixed ``gnt-cluster verify`` regarding ``master-ip-script`` on non
341 - Fixed a RAPI regression on missing beparams/memory
342 - Fixed redistribution of files on offline nodes
343 - Added possibility to run activate-disks even though secondaries are
344 offline. With this change it relaxes also the strictness on some other
345 commands which use activate disks internally:
346 * ``gnt-instance start|reboot|rename|backup|export``
347 - Made it possible to remove safely an instance if its secondaries are
349 - Made it possible to reinstall even though secondaries are offline
355 *(Released Mon, 25 Jun 2012)*
357 First release candidate for 2.6. The following changes were done from
360 - Fixed bugs with disk parameters and ``rbd`` templates as well as
362 - Made ``gnt-instance modify`` more consistent regarding new NIC/Disk
363 behaviour. It supports now the modify operation
364 - ``hcheck`` implemented to analyze cluster health and possibility of
365 improving health by rebalance
366 - ``hbal`` has been improved in dealing with split instances
372 *(Released Mon, 11 Jun 2012)*
374 Second beta release of 2.6. The following changes were done from beta2
377 - Fixed ``daemon-util`` with non-root user models
378 - Fixed creation of plain instances with ``--no-wait-for-sync``
379 - Fix wrong iv_names when running ``cfgupgrade``
380 - Export more information in RAPI group queries
381 - Fixed bug when changing instance network interfaces
382 - Extended burnin to do NIC changes
383 - query: Added ``<``, ``>``, ``<=``, ``>=`` comparison operators
384 - Changed default for DRBD barriers
385 - Fixed DRBD error reporting for syncer rate
386 - Verify the options on disk parameters
388 And of course various fixes to documentation and improved unittests and
395 *(Released Wed, 23 May 2012)*
397 First beta release of 2.6. The following changes were done from beta1 to
400 - integrated patch for distributions without ``start-stop-daemon``
401 - adapted example init.d script to work on Fedora
402 - fixed log handling in Haskell daemons
403 - adapted checks in the watcher for pycurl linked against libnss
404 - add partial support for ``xl`` instead of ``xm`` for Xen
405 - fixed a type issue in cluster verification
406 - fixed ssconf handling in the Haskell code (was breaking confd in IPv6
409 Plus integrated fixes from the 2.5 branch:
411 - fixed ``kvm-ifup`` to use ``/bin/bash``
412 - fixed parallel build failures
413 - KVM live migration when using a custom keymap
419 *(Released Tue, 24 Jul 2012)*
421 A small bugfix release, with no new features:
423 - fixed bash-isms in kvm-ifup, for compatibility with systems which use a
424 different default shell (e.g. Debian, Ubuntu)
425 - fixed KVM startup and live migration with a custom keymap (fixes Issue
426 243 and Debian bug #650664)
427 - fixed compatibility with KVM versions that don't support multiple boot
428 devices (fixes Issue 230 and Debian bug #624256)
430 Additionally, a few fixes were done to the build system (fixed parallel
431 build failures) and to the unittests (fixed race condition in test for
432 FileID functions, and the default enable/disable mode for QA test is now
439 *(Released Fri, 11 May 2012)*
441 A small bugfix release.
443 The main issues solved are on the topic of compatibility with newer LVM
446 - fixed parsing of ``lv_attr`` field
447 - adapted to new ``vgreduce --removemissing`` behaviour where sometimes
448 the ``--force`` flag is needed
450 Also on the topic of compatibility, ``tools/lvmstrap`` has been changed
451 to accept kernel 3.x too (was hardcoded to 2.6.*).
453 A regression present in 2.5.0 that broke handling (in the gnt-* scripts)
454 of hook results and that also made display of other errors suboptimal
455 was fixed; the code behaves now like 2.4 and earlier.
457 Another change in 2.5, the cleanup of the OS scripts environment, is too
458 aggressive: it removed even the ``PATH`` variable, which requires the OS
459 scripts to *always* need to export it. Since this is a bit too strict,
460 we now export a minimal PATH, the same that we export for hooks.
462 The fix for issue 201 (Preserve bridge MTU in KVM ifup script) was
463 integrated into this release.
465 Finally, a few other miscellaneous changes were done (no new features,
466 just small improvements):
468 - Fix ``gnt-group --help`` display
469 - Fix hardcoded Xen kernel path
470 - Fix grow-disk handling of invalid units
471 - Update synopsis for ``gnt-cluster repair-disk-sizes``
472 - Accept both PUT and POST in noded (makes future upgrade to 2.6 easier)
478 *(Released Thu, 12 Apr 2012)*
480 Incompatible/important changes and bugfixes
481 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
483 - The default of the ``/2/instances/[instance_name]/rename`` RAPI
484 resource's ``ip_check`` parameter changed from ``True`` to ``False``
485 to match the underlying LUXI interface.
486 - The ``/2/nodes/[node_name]/evacuate`` RAPI resource was changed to use
487 body parameters, see :doc:`RAPI documentation <rapi>`. The server does
488 not maintain backwards-compatibility as the underlying operation
489 changed in an incompatible way. The RAPI client can talk to old
490 servers, but it needs to be told so as the return value changed.
491 - When creating file-based instances via RAPI, the ``file_driver``
492 parameter no longer defaults to ``loop`` and must be specified.
493 - The deprecated ``bridge`` NIC parameter is no longer supported. Use
495 - Support for the undocumented and deprecated RAPI instance creation
496 request format version 0 has been dropped. Use version 1, supported
497 since Ganeti 2.1.3 and :doc:`documented <rapi>`, instead.
498 - Pyparsing 1.4.6 or above is required, see :doc:`installation
499 documentation <install>`.
500 - The "cluster-verify" hooks are now executed per group by the
501 ``OP_CLUSTER_VERIFY_GROUP`` opcode. This maintains the same behavior
502 if you just run ``gnt-cluster verify``, which generates one opcode per
504 - The environment as passed to the OS scripts is cleared, and thus no
505 environment variables defined in the node daemon's environment will be
506 inherited by the scripts.
507 - The :doc:`iallocator <iallocator>` mode ``multi-evacuate`` has been
509 - :doc:`New iallocator modes <design-multi-reloc>` have been added to
510 support operations involving multiple node groups.
511 - Offline nodes are ignored when failing over an instance.
512 - Support for KVM version 1.0, which changed the version reporting format
514 - TCP/IP ports used by DRBD disks are returned to a pool upon instance
516 - ``Makefile`` is now compatible with Automake 1.11.2
517 - Includes all bugfixes made in the 2.4 series
522 - The ganeti-htools project has been merged into the ganeti-core source
523 tree and will be built as part of Ganeti (see :doc:`install-quick`).
524 - Implemented support for :doc:`shared storage <design-shared-storage>`.
525 - Add support for disks larger than 2 TB in ``lvmstrap`` by supporting
526 GPT-style partition tables (requires `parted
527 <http://www.gnu.org/s/parted/>`_).
528 - Added support for floppy drive and 2nd CD-ROM drive in KVM hypervisor.
529 - Allowed adding tags on instance creation.
530 - Export instance tags to hooks (``INSTANCE_TAGS``, see :doc:`hooks`)
531 - Allow instances to be started in a paused state, enabling the user to
532 see the complete console output on boot using the console.
533 - Added new hypervisor flag to control default reboot behaviour
534 (``reboot_behavior``).
535 - Added support for KVM keymaps (hypervisor parameter ``keymap``).
536 - Improved out-of-band management support:
538 - Added ``gnt-node health`` command reporting the health status of
540 - Added ``gnt-node power`` command to manage power status of nodes.
541 - Added command for emergency power-off (EPO), ``gnt-cluster epo``.
543 - Instance migration can fall back to failover if instance is not
545 - Filters can be used when listing nodes, instances, groups and locks;
546 see *ganeti(7)* manpage.
547 - Added post-execution status as variables to :doc:`hooks <hooks>`
549 - Instance tags are exported/imported together with the instance.
550 - When given an explicit job ID, ``gnt-job info`` will work for archived
552 - Jobs can define dependencies on other jobs (not yet supported via
553 RAPI or command line, but used by internal commands and usable via
556 - Lock monitor (``gnt-debug locks``) shows jobs waiting for
559 - Instance failover is now available as a RAPI resource
560 (``/2/instances/[instance_name]/failover``).
561 - ``gnt-instance info`` defaults to static information if primary node
563 - Opcodes have a new ``comment`` attribute.
564 - Added basic SPICE support to KVM hypervisor.
565 - ``tools/ganeti-listrunner`` allows passing of arguments to executable.
567 Node group improvements
568 ~~~~~~~~~~~~~~~~~~~~~~~
570 - ``gnt-cluster verify`` has been modified to check groups separately,
571 thereby improving performance.
572 - Node group support has been added to ``gnt-cluster verify-disks``,
573 which now operates per node group.
574 - Watcher has been changed to work better with node groups.
576 - One process and state file per node group.
577 - Slow watcher in one group doesn't block other group's watcher.
579 - Added new command, ``gnt-group evacuate``, to move all instances in a
580 node group to other groups.
581 - Added ``gnt-instance change-group`` to move an instance to another
583 - ``gnt-cluster command`` and ``gnt-cluster copyfile`` now support
584 per-group operations.
585 - Node groups can be tagged.
586 - Some operations switch from an exclusive to a shared lock as soon as
588 - Instance's primary and secondary nodes' groups are now available as
589 query fields (``pnode.group``, ``pnode.group.uuid``, ``snodes.group``
590 and ``snodes.group.uuid``).
595 - Numerous updates to documentation and manpages.
597 - :doc:`RAPI <rapi>` documentation now has detailed parameter
599 - Some opcode/job results are now also documented, see :doc:`RAPI
602 - A lockset's internal lock is now also visible in lock monitor.
603 - Log messages from job queue workers now contain information about the
604 opcode they're processing.
605 - ``gnt-instance console`` no longer requires the instance lock.
606 - A short delay when waiting for job changes reduces the number of LUXI
607 requests significantly.
608 - DRBD metadata volumes are overwritten with zeros during disk creation.
609 - Out-of-band commands no longer acquire the cluster lock in exclusive
611 - ``devel/upload`` now uses correct permissions for directories.
617 *(Released Fri, 23 Mar 2012)*
619 This was the sixth release candidate of the 2.5 series.
625 *(Released Mon, 9 Jan 2012)*
627 This was the fifth release candidate of the 2.5 series.
633 *(Released Thu, 27 Oct 2011)*
635 This was the fourth release candidate of the 2.5 series.
641 *(Released Wed, 26 Oct 2011)*
643 This was the third release candidate of the 2.5 series.
649 *(Released Tue, 18 Oct 2011)*
651 This was the second release candidate of the 2.5 series.
657 *(Released Tue, 4 Oct 2011)*
659 This was the first release candidate of the 2.5 series.
665 *(Released Wed, 31 Aug 2011)*
667 This was the third beta release of the 2.5 series.
673 *(Released Mon, 22 Aug 2011)*
675 This was the second beta release of the 2.5 series.
681 *(Released Fri, 12 Aug 2011)*
683 This was the first beta release of the 2.5 series.
689 *(Released Thu, 27 Oct 2011)*
691 - Fixed bug when parsing command line parameter values ending in
693 - Fixed assertion error after unclean master shutdown
694 - Disable HTTP client pool for RPC, significantly reducing memory usage
696 - Fixed queue archive creation with wrong permissions
702 *(Released Tue, 23 Aug 2011)*
706 - Fixed documentation for importing with ``--src-dir`` option
707 - Fixed a bug in ``ensure-dirs`` with queue/archive permissions
708 - Fixed a parsing issue with DRBD 8.3.11 in the Linux kernel
714 *(Released Fri, 5 Aug 2011)*
716 Many bug-fixes and a few small features:
718 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
719 issues when you tried to add an instance with two MAC addresses in one
721 - KVM: fixed per-instance stored UID value
722 - KVM: configure bridged NICs at migration start
723 - KVM: Fix a bug where instance will not start with never KVM versions
725 - Added OS search path to ``gnt-cluster info``
726 - Fixed an issue with ``file_storage_dir`` where you were forced to
727 provide an absolute path, but the documentation states it is a
728 relative path, the documentation was right
729 - Added a new parameter to instance stop/start called ``--no-remember``
730 that will make the state change to not be remembered
731 - Implemented ``no_remember`` at RAPI level
732 - Improved the documentation
733 - Node evacuation: don't call IAllocator if node is already empty
734 - Fixed bug in DRBD8 replace disks on current nodes
735 - Fixed bug in recreate-disks for DRBD instances
736 - Moved assertion checking locks in ``gnt-instance replace-disks``
737 causing it to abort with not owning the right locks for some situation
738 - Job queue: Fixed potential race condition when cancelling queued jobs
739 - Fixed off-by-one bug in job serial generation
740 - ``gnt-node volumes``: Fix instance names
741 - Fixed aliases in bash completion
742 - Fixed a bug in reopening log files after being sent a SIGHUP
743 - Added a flag to burnin to allow specifying VCPU count
744 - Bugfixes to non-root Ganeti configuration
750 *(Released Thu, 12 May 2011)*
752 Many bug-fixes and a few new small features:
754 - Fixed a bug related to log opening failures
755 - Fixed a bug in instance listing with orphan instances
756 - Fixed a bug which prevented resetting the cluster-level node parameter
757 ``oob_program`` to the default
758 - Many fixes related to the ``cluster-merge`` tool
759 - Fixed a race condition in the lock monitor, which caused failures
760 during (at least) creation of many instances in parallel
761 - Improved output for gnt-job info
762 - Removed the quiet flag on some ssh calls which prevented debugging
764 - Improved the N+1 failure messages in cluster verify by actually
765 showing the memory values (needed and available)
766 - Increased lock attempt timeouts so that when executing long operations
767 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
768 too early and thus prevent the use of the 'fair' mechanism
769 - Changed instance query data (``gnt-instance info``) to not acquire
770 locks unless needed, thus allowing its use on locked instance if only
771 static information is asked for
772 - Improved behaviour with filesystems that do not support rename on an
774 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
775 kept locks on all nodes during the wipe, which is unneeded
776 - Fixed ``gnt-watcher`` handling of errors during hooks execution
777 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
778 10GiB) which caused the wipe to fail right at the end in some cases
779 - Fixed master IP activation when doing master failover with no-voting
780 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
781 the master node itself
782 - Fixed potential data-loss in under disk full conditions, where Ganeti
783 wouldn't check correctly the return code and would consider
784 partially-written files 'correct'
785 - Fixed bug related to multiple VGs and DRBD disk replacing
786 - Added new disk parameter ``metavg`` that allows placement of the meta
787 device for DRBD in a different volume group
788 - Fixed error handling in the node daemon when the system libc doesn't
789 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
790 - Fixed lock release during replace-disks, which kept cluster-wide locks
791 when doing disk replaces with an iallocator script
792 - Added check for missing bridges in cluster verify
793 - Handle EPIPE errors while writing to the terminal better, so that
794 piping the output to e.g. ``less`` doesn't cause a backtrace
795 - Fixed rare case where a ^C during Luxi calls could have been
796 interpreted as server errors, instead of simply terminating
797 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
799 - Added a few more parameters to the KVM hypervisor, allowing a second
800 CDROM, custom disk type for CDROMs and a floppy image
801 - Removed redundant message in instance rename when the name is given
803 - Added option to ``gnt-instance recreate-disks`` to allow creating the
804 disks on new nodes, allowing recreation when the original instance
805 nodes are completely gone
806 - Added option when converting disk templates to DRBD to skip waiting
807 for the resync, in order to make the instance available sooner
808 - Added two new variables to the OS scripts environment (containing the
810 - Made the root_path and optional parameter for the xen-pvm hypervisor,
811 to allow use of ``pvgrub`` as bootloader
812 - Changed the instance memory modifications to only check out-of-memory
813 conditions on memory increases, and turned the secondary node warnings
814 into errors (they can still be overridden via ``--force``)
815 - Fixed the handling of a corner case when the Python installation gets
816 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
817 to execute a command that doesn't exist
818 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
819 node of the instance returned failures during instance shutdown; this
820 adds the option ``--ignore-consistency`` to gnt-instance move
822 And as usual, various improvements to the error messages, documentation
829 *(Released Wed, 09 Mar 2011)*
831 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
832 the RAPI users file if run twice (even with ``--dry-run``).
834 The release fixes that bug (nothing else changed).
840 *(Released Mon, 07 Mar 2011)*
842 Final 2.4.0 release. Just a few small fixes:
844 - Fixed RAPI node evacuate
845 - Fixed the kvm-ifup script
846 - Fixed internal error handling for special job cases
847 - Updated man page to specify the escaping feature for options
853 *(Released Mon, 28 Feb 2011)*
855 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
856 that this feature wiped the disks of the wrong instance, leading to loss
861 - Fixed title of query field containing instance name
862 - Expanded the glossary in the documentation
863 - Fixed one unittest (internal issue)
869 *(Released Mon, 21 Feb 2011)*
871 A number of bug fixes plus just a couple functionality changes.
873 On the user-visible side, the ``gnt-* list`` command output has changed
874 with respect to "special" field states. The current rc1 style of display
875 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
876 the default output mode special fields states are displayed as follows:
878 - Offline resource: ``*``
879 - Unavailable/not applicable: ``-``
880 - Data missing (RPC failure): ``?``
881 - Unknown field: ``??``
883 Another user-visible change is the addition of ``--force-join`` to
888 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
889 - Fixed regression in RAPI/instance reinstall where all parameters were
890 required (instead of optional)
891 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
892 - Fixed iallocator usage (offline nodes were not considered offline)
893 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
894 - Fixed hypervisor and OS parameter validation with respect to
896 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
898 - Fixed ``tools/listrunner`` with respect to agent-based usage
904 *(Released Fri, 4 Feb 2011)*
906 Many changes and fixes since the beta1 release. While there were some
907 internal changes, the code has been mostly stabilised for the RC
910 Note: the dumb allocator was removed in this release, as it was not kept
911 up-to-date with the IAllocator protocol changes. It is recommended to
912 use the ``hail`` command from the ganeti-htools package.
914 Note: the 2.4 and up versions of Ganeti are not compatible with the
915 0.2.x branch of ganeti-htools. You need to upgrade to
916 ganeti-htools-0.3.0 (or later).
918 Regressions fixed from 2.3
919 ~~~~~~~~~~~~~~~~~~~~~~~~~~
921 - Fixed the ``gnt-cluster verify-disks`` command
922 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
924 - Fixed disk adoption breakage
925 - Fixed wrong headers in instance listing for field aliases
930 - Fixed corner case in KVM handling of NICs
931 - Fixed many cases of wrong handling of non-vm_capable nodes
932 - Fixed a bug where a missing instance symlink was not possible to
933 recreate with any ``gnt-*`` command (now ``gnt-instance
934 activate-disks`` does it)
935 - Fixed the volume group name as reported by ``gnt-cluster
937 - Increased timeouts for the import-export code, hopefully leading to
938 fewer aborts due network or instance timeouts
939 - Fixed bug in ``gnt-node list-storage``
940 - Fixed bug where not all daemons were started on cluster
941 initialisation, but only at the first watcher run
942 - Fixed many bugs in the OOB implementation
943 - Fixed watcher behaviour in presence of instances with offline
945 - Fixed instance list output for instances running on the wrong node
946 - a few fixes to the cluster-merge tool, but it still cannot merge
947 multi-node groups (currently it is not recommended to use this tool)
953 - Improved network configuration for the KVM hypervisor
954 - Added e1000 as a supported NIC for Xen-HVM
955 - Improved the lvmstrap tool to also be able to use partitions, as
956 opposed to full disks
957 - Improved speed of disk wiping (the cluster parameter
958 ``prealloc_wipe_disks``, so that it has a low impact on the total time
959 of instance creations
960 - Added documentation for the OS parameters
961 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
962 hypervisor is not responding
963 - Added display of blacklisted and hidden OS information in
965 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
966 NIC and node parameters, which might create problems with currently
967 invalid (but undetected) configuration files, but prevents validation
968 failures when unrelated parameters are modified
969 - Changed cluster initialisation to wait for the master daemon to become
971 - Expanded the RAPI interface:
973 - Added config redistribution resource
974 - Added activation/deactivation of instance disks
975 - Added export of console information
977 - Implemented log file reopening on SIGHUP, which allows using
978 logrotate(8) for the Ganeti log files
979 - Added a basic OOB helper script as an example
985 *(Released Fri, 14 Jan 2011)*
990 - Fixed timezone issues when formatting timestamps
991 - Added support for node groups, available via ``gnt-group`` and other
993 - Added out-of-band framework and management, see :doc:`design
994 document <design-oob>`
995 - Removed support for roman numbers from ``gnt-node list`` and
996 ``gnt-instance list``.
997 - Allowed modification of master network interface via ``gnt-cluster
998 modify --master-netdev``
999 - Accept offline secondaries while shutting down instance disks
1000 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
1001 - Added support for multiple LVM volume groups
1002 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
1004 - Added commands to list available fields:
1006 - ``gnt-node list-fields``
1007 - ``gnt-group list-fields``
1008 - ``gnt-instance list-fields``
1010 - Updated documentation and man pages
1015 - Moved ``rapi_users`` file into separate directory, now named
1016 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
1018 - Added new tool for running commands on many machines,
1019 ``tools/ganeti-listrunner``
1020 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
1021 improving the ``gnt-instance console`` output
1022 - Allowed customisation of disk index separator at ``configure`` time
1023 - Export node group allocation policy to :doc:`iallocator <iallocator>`
1024 - Added support for non-partitioned md disks in ``lvmstrap``
1025 - Added script to gracefully power off KVM instances
1026 - Split ``utils`` module into smaller parts
1027 - Changed query operations to return more detailed information, e.g.
1028 whether an information is unavailable due to an offline node. To use
1029 this new functionality, the LUXI call ``Query`` must be used. Field
1030 information is now stored by the master daemon and can be retrieved
1031 using ``QueryFields``. Instances, nodes and groups can also be queried
1032 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
1033 exposed via RAPI). The following commands make use of this
1034 infrastructure change:
1036 - ``gnt-group list``
1037 - ``gnt-group list-fields``
1039 - ``gnt-node list-fields``
1040 - ``gnt-instance list``
1041 - ``gnt-instance list-fields``
1042 - ``gnt-debug locks``
1047 - New RAPI resources (see :doc:`rapi`):
1051 - ``/2/groups/[group_name]``
1052 - ``/2/groups/[group_name]/assign-nodes``
1053 - ``/2/groups/[group_name]/modify``
1054 - ``/2/groups/[group_name]/rename``
1055 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
1059 - Implemented ``no_install`` for instance creation
1060 - Implemented OS parameters for instance reinstallation, allowing
1061 use of special settings on reinstallation (e.g. for preserving data)
1066 - Added IPv6 support in import/export
1067 - Pause DRBD synchronization while wiping disks on instance creation
1068 - Updated unittests and QA scripts
1069 - Improved network parameters passed to KVM
1070 - Converted man pages from docbook to reStructuredText
1076 *(Released Mon, 20 Dec 2010)*
1078 Released version 2.3.1~rc1 without any changes.
1084 *(Released Wed, 1 Dec 2010)*
1086 - impexpd: Disable OpenSSL compression in socat if possible (backport
1087 from master, commit e90739d625b, see :doc:`installation guide
1088 <install-quick>` for details)
1089 - Changed unittest coverage report to exclude test scripts
1090 - Added script to check version format
1096 *(Released Wed, 1 Dec 2010)*
1098 Released version 2.3.0~rc1 without any changes.
1104 *(Released Fri, 19 Nov 2010)*
1106 A number of bugfixes and documentation updates:
1108 - Update ganeti-os-interface documentation
1109 - Fixed a bug related to duplicate MACs or similar items which should be
1111 - Fix breakage in OS state modify
1112 - Reinstall instance: disallow offline secondaries (fixes bug related to
1113 OS changing but reinstall failing)
1114 - plus all the other fixes between 2.2.1 and 2.2.2
1120 *(Released Tue, 2 Nov 2010)*
1122 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
1123 - Fixed master failover race with watcher
1124 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
1126 - Accept previously stopped instance for export with instance removal
1127 - Simplify and extend the environment variables for instance OS scripts
1128 - Added new node flags, ``master_capable`` and ``vm_capable``
1129 - Added optional instance disk wiping prior during allocation. This is a
1130 cluster-wide option and can be set/modified using
1131 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
1132 - Added IPv6 support, see :doc:`design document <design-2.3>` and
1133 :doc:`install-quick`
1134 - Added a new watcher option (``--ignore-pause``)
1135 - Added option to ignore offline node on instance start/stop
1136 (``--ignore-offline``)
1137 - Allow overriding OS parameters with ``gnt-instance reinstall``
1138 - Added ability to change node's secondary IP address using ``gnt-node
1140 - Implemented privilege separation for all daemons except
1141 ``ganeti-noded``, see ``configure`` options
1142 - Complain if an instance's disk is marked faulty in ``gnt-cluster
1144 - Implemented job priorities (see ``ganeti(7)`` manpage)
1145 - Ignore failures while shutting down instances during failover from
1147 - Exit daemon's bootstrap process only once daemon is ready
1148 - Export more information via ``LUInstanceQuery``/remote API
1149 - Improved documentation, QA and unittests
1150 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
1151 restart if the file was created or changed
1152 - Added LUXI protocol version sent with each request and response,
1153 allowing detection of server/client mismatches
1154 - Moved the Python scripts among gnt-* and ganeti-* into modules
1155 - Moved all code related to setting up SSH to an external script,
1157 - Infrastructure changes for node group support in future versions
1163 *(Released Fri, 19 Nov 2010)*
1165 A few small bugs fixed, and some improvements to the build system:
1167 - Fix documentation regarding conversion to drbd
1168 - Fix validation of parameters in cluster modify (``gnt-cluster modify
1170 - Fix error handling in node modify with multiple changes
1171 - Allow remote imports without checked names
1177 *(Released Tue, 19 Oct 2010)*
1179 - Disable SSL session ID cache in RPC client
1185 *(Released Thu, 14 Oct 2010)*
1187 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
1188 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
1190 - Fix problems with interaction between Curl and Python's HTTP server,
1191 resulting in increased speed in many RPC calls
1192 - Improve our release script to prevent breakage with older aclocal and
1199 *(Released Thu, 7 Oct 2010)*
1201 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
1202 value retrieved from master
1203 - Added support for blacklisted or hidden OS definitions
1204 - Added simple lock monitor (accessible via (``gnt-debug locks``)
1205 - Added support for -mem-path in KVM hypervisor abstraction layer
1206 - Allow overriding instance parameters in tool for inter-cluster
1207 instance moves (``tools/move-instance``)
1208 - Improved opcode summaries (e.g. in ``gnt-job list``)
1209 - Improve consistency of OS listing by sorting it
1210 - Documentation updates
1216 *(Released Fri, 8 Oct 2010)*
1218 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
1224 *(Released Mon, 4 Oct 2010)*
1226 - Fixed regression in ``gnt-instance rename``
1232 *(Released Wed, 22 Sep 2010)*
1234 - Fixed OS_VARIANT variable for OS scripts
1235 - Fixed cluster tag operations via RAPI
1236 - Made ``setup-ssh`` exit with non-zero code if an error occurred
1237 - Disabled RAPI CA checks in watcher
1243 *(Released Mon, 23 Aug 2010)*
1245 - Support DRBD versions of the format "a.b.c.d"
1247 - Re-introduce support for usage from multiple threads in RAPI client
1248 - Instance renames and modify via RAPI
1249 - Work around race condition between processing and archival in job
1251 - Mark opcodes following failed one as failed, too
1252 - Job field ``lock_status`` was removed due to difficulties making it
1253 work with the changed job queue in Ganeti 2.2; a better way to monitor
1254 locks is expected for a later 2.2.x release
1255 - Fixed dry-run behaviour with many commands
1256 - Support ``ssh-agent`` again when adding nodes
1257 - Many additional bugfixes
1263 *(Released Fri, 30 Jul 2010)*
1265 Important change: the internal RPC mechanism between Ganeti nodes has
1266 changed from using a home-grown http library (based on the Python base
1267 libraries) to use the PycURL library. This requires that PycURL is
1268 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
1269 against GnuTLS by default. cURL's support for GnuTLS had known issues
1270 before cURL 7.21.0 and we recommend using the latest cURL release or
1271 linking against OpenSSL. Most other distributions already link PycURL
1272 and cURL against OpenSSL. The command::
1274 python -c 'import pycurl; print pycurl.version'
1276 can be used to determine the libraries PycURL and cURL are linked
1279 Other significant changes:
1281 - Rewrote much of the internals of the job queue, in order to achieve
1282 better parallelism; this decouples job query operations from the job
1283 processing, and it should allow much nicer behaviour of the master
1284 daemon under load, and it also has uncovered some long-standing bugs
1285 related to the job serialisation (now fixed)
1286 - Added a default iallocator setting to the cluster parameters,
1287 eliminating the need to always pass nodes or an iallocator for
1288 operations that require selection of new node(s)
1289 - Added experimental support for the LXC virtualization method
1290 - Added support for OS parameters, which allows the installation of
1291 instances to pass parameter to OS scripts in order to customise the
1293 - Added a hypervisor parameter controlling the migration type (live or
1294 non-live), since hypervisors have various levels of reliability; this
1295 has renamed the 'live' parameter to 'mode'
1296 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
1297 logical volumes, meaning that cluster verify will ignore them and not
1298 flag their presence as errors
1299 - The watcher will now reset the error count for failed instances after
1300 8 hours, thus allowing self-healing if the problem that caused the
1301 instances to be down/fail to start has cleared in the meantime
1302 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
1303 check for, and warn, if the drbd module parameter ``usermode_helper``
1304 is not consistent with the cluster-wide setting; this is needed to
1305 make diagnose easier of failed drbd creations
1306 - Started adding base IPv6 support, but this is not yet
1307 enabled/available for use
1308 - Rename operations (cluster, instance) will now return the new name,
1309 which is especially useful if a short name was passed in
1310 - Added support for instance migration in RAPI
1311 - Added a tool to pre-configure nodes for the SSH setup, before joining
1312 them to the cluster; this will allow in the future a simplified model
1313 for node joining (but not yet fully enabled in 2.2); this needs the
1314 paramiko python library
1315 - Fixed handling of name-resolving errors
1316 - Fixed consistency of job results on the error path
1317 - Fixed master-failover race condition when executed multiple times in
1319 - Fixed many bugs related to the job queue (mostly introduced during the
1320 2.2 development cycle, so not all are impacting 2.1)
1321 - Fixed instance migration with missing disk symlinks
1322 - Fixed handling of unknown jobs in ``gnt-job archive``
1323 - And many other small fixes/improvements
1327 - Enhanced both the unittest and the QA coverage
1328 - Switched the opcode validation to a generic model, and extended the
1329 validation to all opcode parameters
1330 - Changed more parts of the code that write shell scripts to use the
1332 - Switched the master daemon to use the asyncore library for the Luxi
1339 *(Released Thu, 17 Jun 2010)*
1341 - Added tool (``move-instance``) and infrastructure to move instances
1342 between separate clusters (see :doc:`separate documentation
1343 <move-instance>` and :doc:`design document <design-2.2>`)
1344 - Added per-request RPC timeout
1345 - RAPI now requires a Content-Type header for requests with a body (e.g.
1346 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
1347 :rfc:`2616` (HTTP/1.1), section 7.2.1)
1348 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
1350 - Implemented initial support for running Ganeti daemons as separate
1351 users, see configure-time flags ``--with-user-prefix`` and
1352 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
1354 - Instances can be removed after export (``gnt-backup export
1355 --remove-instance``)
1356 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
1357 key (instead of 1024 bit)
1358 - Added new cluster configuration file for cluster domain secret
1359 - Import/export now use SSL instead of SSH
1360 - Added support for showing estimated time when exporting an instance,
1361 see the ``ganeti-os-interface(7)`` manpage and look for
1368 *(Released Tue, 16 Nov 2010)*
1370 Some more bugfixes. Unless critical bugs occur, this will be the last
1373 - Fix case of MAC special-values
1374 - Fix mac checker regex
1375 - backend: Fix typo causing "out of range" error
1376 - Add missing --units in gnt-instance list man page
1382 *(Released Tue, 24 Aug 2010)*
1385 - Don't ignore secondary node silently on non-mirrored disk templates
1387 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
1388 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
1389 - Properly document the usb_mouse parameter
1390 - Fix path in ganeti-rapi(8) (issue 116)
1391 - Adjust error message when the ganeti user's .ssh directory is
1393 - Add same-node-check when changing the disk template to drbd
1399 *(Released Fri, 16 Jul 2010)*
1402 - Add an option to only select some reboot types during qa/burnin.
1403 (on some hypervisors consequent reboots are not supported)
1404 - Fix infrequent race condition in master failover. Sometimes the old
1405 master ip address would be still detected as up for a short time
1406 after it was removed, causing failover to fail.
1407 - Decrease mlockall warnings when the ctypes module is missing. On
1408 Python 2.4 we support running even if no ctypes module is installed,
1409 but we were too verbose about this issue.
1410 - Fix building on old distributions, on which man doesn't have a
1412 - Fix RAPI not to ignore the MAC address on instance creation
1413 - Implement the old instance creation format in the RAPI client.
1419 *(Released Thu, 01 Jul 2010)*
1421 A small bugfix release:
1422 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
1423 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
1424 a non-existing option
1425 - Fix instance create: the --force-variant option was ignored
1426 - Improve pylint 0.21 compatibility and warnings with Python 2.6
1427 - Fix modify node storage with non-FQDN arguments
1428 - Fix RAPI client to authenticate under Python 2.6 when used
1429 for more than 5 requests needing authentication
1430 - Fix gnt-instance modify -t (storage) giving a wrong error message
1431 when converting a non-shutdown drbd instance to plain
1437 *(Released Fri, 18 Jun 2010)*
1439 A small bugfix release:
1441 - Fix live migration of KVM instances started with older Ganeti
1442 versions which had fewer hypervisor parameters
1443 - Fix gnt-instance grow-disk on down instances
1444 - Fix an error-reporting bug during instance migration
1445 - Better checking of the ``--net`` and ``--disk`` values, to avoid
1446 silently ignoring broken ones
1447 - Fix an RPC error reporting bug affecting, for example, RAPI client
1449 - Fix bug triggered by different API version os-es on different nodes
1450 - Fix a bug in instance startup with custom hvparams: OS level
1451 parameters would fail to be applied.
1452 - Fix the RAPI client under Python 2.6 (but more work is needed to
1453 make it work completely well with OpenSSL)
1454 - Fix handling of errors when resolving names from DNS
1460 *(Released Thu, 3 Jun 2010)*
1462 A medium sized development cycle. Some new features, and some
1463 fixes/small improvements/cleanups.
1465 Significant features
1466 ~~~~~~~~~~~~~~~~~~~~
1468 The node deamon now tries to mlock itself into memory, unless the
1469 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
1470 its logs, and falls back to console logging. This allows emergency
1471 features such as ``gnt-node powercycle`` to work even in the event of a
1472 broken node disk (tested offlining the disk hosting the node's
1473 filesystem and dropping its memory caches; don't try this at home)
1475 KVM: add vhost-net acceleration support. It can be tested with a new
1476 enough version of the kernel and of qemu-kvm.
1478 KVM: Add instance chrooting feature. If you use privilege dropping for
1479 your VMs you can also now force them to chroot to an empty directory,
1480 before starting the emulated guest.
1482 KVM: Add maximum migration bandwith and maximum downtime tweaking
1483 support (requires a new-enough version of qemu-kvm).
1485 Cluster verify will now warn if the master node doesn't have the master
1486 ip configured on it.
1488 Add a new (incompatible) instance creation request format to RAPI which
1489 supports all parameters (previously only a subset was supported, and it
1490 wasn't possible to extend the old format to accomodate all the new
1491 features. The old format is still supported, and a client can check for
1492 this feature, before using it, by checking for its presence in the
1493 ``features`` RAPI resource.
1495 Now with ancient latin support. Try it passing the ``--roman`` option to
1496 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
1497 (requires the python-roman module to be installed, in order to work).
1502 As usual many internal code refactorings, documentation updates, and
1505 - Lots of improvements and cleanups to the experimental Remote API
1506 (RAPI) client library.
1507 - A new unit test suite for the core daemon libraries.
1508 - A fix to creating missing directories makes sure the umask is not
1509 applied anymore. This enforces the same directory permissions
1511 - Better handling terminating daemons with ctrl+c (used when running
1512 them in debugging mode).
1513 - Fix a race condition in live migrating a KVM instance, when stat()
1514 on the old proc status file returned EINVAL, which is an unexpected
1516 - Fixed manpage checking with newer man and utf-8 charachters. But now
1517 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
1523 *(Released Fri, 7 May 2010)*
1525 Fix a bug which prevented untagged KVM instances from starting.
1531 *(Released Fri, 7 May 2010)*
1533 Another release with a long development cycle, during which many
1534 different features were added.
1536 Significant features
1537 ~~~~~~~~~~~~~~~~~~~~
1539 The KVM hypervisor now can run the individual instances as non-root, to
1540 reduce the impact of a VM being hijacked due to bugs in the
1541 hypervisor. It is possible to run all instances as a single (non-root)
1542 user, to manually specify a user for each instance, or to dynamically
1543 allocate a user out of a cluster-wide pool to each instance, with the
1544 guarantee that no two instances will run under the same user ID on any
1547 An experimental RAPI client library, that can be used standalone
1548 (without the other Ganeti libraries), is provided in the source tree as
1549 ``lib/rapi/client.py``. Note this client might change its interface in
1550 the future, as we iterate on its capabilities.
1552 A new command, ``gnt-cluster renew-crypto`` has been added to easily
1553 replace the cluster's certificates and crypto keys. This might help in
1554 case they have been compromised, or have simply expired.
1556 A new disk option for instance creation has been added that allows one
1557 to "adopt" currently existing logical volumes, with data
1558 preservation. This should allow easier migration to Ganeti from
1559 unmanaged (or managed via other software) instances.
1561 Another disk improvement is the possibility to convert between redundant
1562 (DRBD) and plain (LVM) disk configuration for an instance. This should
1563 allow better scalability (starting with one node and growing the
1564 cluster, or shrinking a two-node cluster to one node).
1566 A new feature that could help with automated node failovers has been
1567 implemented: if a node sees itself as offline (by querying the master
1568 candidates), it will try to shutdown (hard) all instances and any active
1569 DRBD devices. This reduces the risk of duplicate instances if an
1570 external script automatically failovers the instances on such nodes. To
1571 enable this, the cluster parameter ``maintain_node_health`` should be
1572 enabled; in the future this option (per the name) will enable other
1573 automatic maintenance features.
1575 Instance export/import now will reuse the original instance
1576 specifications for all parameters; that means exporting an instance,
1577 deleting it and the importing it back should give an almost identical
1578 instance. Note that the default import behaviour has changed from
1579 before, where it created only one NIC; now it recreates the original
1582 Cluster verify has added a few new checks: SSL certificates validity,
1583 /etc/hosts consistency across the cluster, etc.
1588 As usual, many internal changes were done, documentation fixes,
1591 - Fixed cluster initialization with disabled cluster storage (regression
1592 introduced in 2.1.1)
1593 - File-based storage supports growing the disks
1594 - Fixed behaviour of node role changes
1595 - Fixed cluster verify for some corner cases, plus a general rewrite of
1596 cluster verify to allow future extension with more checks
1597 - Fixed log spamming by watcher and node daemon (regression introduced
1599 - Fixed possible validation issues when changing the list of enabled
1601 - Fixed cleanup of /etc/hosts during node removal
1602 - Fixed RAPI response for invalid methods
1603 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
1604 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
1605 from ide disks, etc.)
1606 - Allow OS changes without re-installation (to record a changed OS
1607 outside of Ganeti, or to allow OS renames)
1608 - Allow instance creation without OS installation (useful for example if
1609 the OS will be installed manually, or restored from a backup not in
1611 - Implemented option to make cluster ``copyfile`` use the replication
1613 - Added list of enabled hypervisors to ssconf (possibly useful for
1615 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
1617 - A partial form of node re-IP is possible via node readd, which now
1618 allows changed node primary IP
1619 - Command line utilities now show an informational message if the job is
1621 - The logs of the master daemon now show the PID/UID/GID of the
1628 *(Released Fri, 12 Mar 2010)*
1630 During the 2.1.0 long release candidate cycle, a lot of improvements and
1631 changes have accumulated with were released later as 2.1.1.
1636 The node evacuate command (``gnt-node evacuate``) was significantly
1637 rewritten, and as such the IAllocator protocol was changed - a new
1638 request type has been added. This unfortunate change during a stable
1639 series is designed to improve performance of node evacuations; on
1640 clusters with more than about five nodes and which are well-balanced,
1641 evacuation should proceed in parallel for all instances of the node
1642 being evacuated. As such, any existing IAllocator scripts need to be
1643 updated, otherwise the above command will fail due to the unknown
1644 request. The provided "dumb" allocator has not been updated; but the
1645 ganeti-htools package supports the new protocol since version 0.2.4.
1647 Another important change is increased validation of node and instance
1648 names. This might create problems in special cases, if invalid host
1649 names are being used.
1651 Also, a new layer of hypervisor parameters has been added, that sits at
1652 OS level between the cluster defaults and the instance ones. This allows
1653 customisation of virtualization parameters depending on the installed
1654 OS. For example instances with OS 'X' may have a different KVM kernel
1655 (or any other parameter) than the cluster defaults. This is intended to
1656 help managing a multiple OSes on the same cluster, without manual
1657 modification of each instance's parameters.
1659 A tool for merging clusters, ``cluster-merge``, has been added in the
1660 tools sub-directory.
1665 - Improved the int/float conversions that should make the code more
1666 robust in face of errors from the node daemons
1667 - Fixed the remove node code in case of internal configuration errors
1668 - Fixed the node daemon behaviour in face of inconsistent queue
1669 directory (e.g. read-only file-system where we can't open the files
1671 - Fixed the behaviour of gnt-node modify for master candidate demotion;
1672 now it either aborts cleanly or, if given the new "auto_promote"
1673 parameter, will automatically promote other nodes as needed
1674 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
1675 completely prevent Ganeti from working
1676 - Fixed bug for instance export when not all disks were successfully
1678 - Fixed behaviour of node add when the new node is slow in starting up
1680 - Fixed handling of signals in the LUXI client, which should improve
1681 behaviour of command-line scripts
1682 - Added checks for invalid node/instance names in the configuration (now
1683 flagged during cluster verify)
1684 - Fixed watcher behaviour for disk activation errors
1685 - Fixed two potentially endless loops in http library, which led to the
1686 RAPI daemon hanging and consuming 100% CPU in some cases
1687 - Fixed bug in RAPI daemon related to hashed passwords
1688 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
1690 - Enhanced compatibility with non-Debian OSes, but not using absolute
1691 path in some commands and allowing customisation of the ssh
1692 configuration directory
1693 - Fixed possible future issue with new Python versions by abiding to the
1694 proper use of ``__slots__`` attribute on classes
1695 - Added checks that should prevent directory traversal attacks
1696 - Many documentation fixes based on feedback from users
1701 - Added an "early_release" more for instance replace disks and node
1702 evacuate, where we release locks earlier and thus allow higher
1703 parallelism within the cluster
1704 - Added watcher hooks, intended to allow the watcher to restart other
1705 daemons (e.g. from the ganeti-nbma project), but they can be used of
1706 course for any other purpose
1707 - Added a compile-time disable for DRBD barriers, to increase
1708 performance if the administrator trusts the power supply or the
1709 storage system to not lose writes
1710 - Added the option of using syslog for logging instead of, or in
1711 addition to, Ganeti's own log files
1712 - Removed boot restriction for paravirtual NICs for KVM, recent versions
1713 can indeed boot from a paravirtual NIC
1714 - Added a generic debug level for many operations; while this is not
1715 used widely yet, it allows one to pass the debug value all the way to
1717 - Enhanced the hooks environment for instance moves (failovers,
1718 migrations) where the primary/secondary nodes changed during the
1719 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
1720 - Enhanced data validations for many user-supplied values; one important
1721 item is the restrictions imposed on instance and node names, which
1722 might reject some (invalid) host names
1723 - Add a configure-time option to disable file-based storage, if it's not
1724 needed; this allows greater security separation between the master
1725 node and the other nodes from the point of view of the inter-node RPC
1727 - Added user notification in interactive tools if job is waiting in the
1728 job queue or trying to acquire locks
1729 - Added log messages when a job is waiting for locks
1730 - Added filtering by node tags in instance operations which admit
1731 multiple instances (start, stop, reboot, reinstall)
1732 - Added a new tool for cluster mergers, ``cluster-merge``
1733 - Parameters from command line which are of the form ``a=b,c=d`` can now
1734 use backslash escapes to pass in values which contain commas,
1735 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
1737 - For KVM, the instance name is the first parameter passed to KVM, so
1738 that it's more visible in the process list
1744 *(Released Tue, 2 Mar 2010)*
1746 Ganeti 2.1 brings many improvements with it. Major changes:
1748 - Added infrastructure to ease automated disk repairs
1749 - Added new daemon to export configuration data in a cheaper way than
1750 using the remote API
1751 - Instance NICs can now be routed instead of being associated with a
1753 - Improved job locking logic to reduce impact of jobs acquiring multiple
1754 locks waiting for other long-running jobs
1756 In-depth implementation details can be found in the Ganeti 2.1 design
1762 - Added chroot hypervisor
1763 - Added more options to xen-hvm hypervisor (``kernel_path`` and
1765 - Added more options to xen-pvm hypervisor (``use_bootloader``,
1766 ``bootloader_path`` and ``bootloader_args``)
1767 - Added the ``use_localtime`` option for the xen-hvm and kvm
1768 hypervisors, and the default value for this has changed to false (in
1769 2.0 xen-hvm always enabled it)
1770 - Added luxi call to submit multiple jobs in one go
1771 - Added cluster initialization option to not modify ``/etc/hosts``
1773 - Added network interface parameters
1774 - Added dry run mode to some LUs
1775 - Added RAPI resources:
1777 - ``/2/instances/[instance_name]/info``
1778 - ``/2/instances/[instance_name]/replace-disks``
1779 - ``/2/nodes/[node_name]/evacuate``
1780 - ``/2/nodes/[node_name]/migrate``
1781 - ``/2/nodes/[node_name]/role``
1782 - ``/2/nodes/[node_name]/storage``
1783 - ``/2/nodes/[node_name]/storage/modify``
1784 - ``/2/nodes/[node_name]/storage/repair``
1786 - Added OpCodes to evacuate or migrate all instances on a node
1787 - Added new command to list storage elements on nodes (``gnt-node
1788 list-storage``) and modify them (``gnt-node modify-storage``)
1789 - Added new ssconf files with master candidate IP address
1790 (``ssconf_master_candidates_ips``), node primary IP address
1791 (``ssconf_node_primary_ips``) and node secondary IP address
1792 (``ssconf_node_secondary_ips``)
1793 - Added ``ganeti-confd`` and a client library to query the Ganeti
1794 configuration via UDP
1795 - Added ability to run hooks after cluster initialization and before
1797 - Added automatic mode for disk replace (``gnt-instance replace-disks
1799 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
1800 after catastrophic data-loss
1801 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
1803 - Added ``gnt-instance move`` command to move instances
1804 - Added ``gnt-cluster watcher`` command to control watcher
1805 - Added ``gnt-node powercycle`` command to powercycle nodes
1806 - Added new job status field ``lock_status``
1807 - Added parseable error codes to cluster verification (``gnt-cluster
1808 verify --error-codes``) and made output less verbose (use
1809 ``--verbose`` to restore previous behaviour)
1810 - Added UUIDs to the main config entities (cluster, nodes, instances)
1811 - Added support for OS variants
1812 - Added support for hashed passwords in the Ganeti remote API users file
1814 - Added option to specify maximum timeout on instance shutdown
1815 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
1816 - Added new helper script to start and stop Ganeti daemons
1817 (``daemon-util``), with the intent to reduce the work necessary to
1818 adjust Ganeti for non-Debian distributions and to start/stop daemons
1820 - Added more unittests
1821 - Fixed critical bug in ganeti-masterd startup
1822 - Removed the configure-time ``kvm-migration-port`` parameter, this is
1823 now customisable at the cluster level for both the KVM and Xen
1824 hypervisors using the new ``migration_port`` parameter
1825 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
1826 reinstalling an instance
1827 - Allowed ``@`` in tag names
1828 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
1829 - Many documentation updates
1830 - Distribute hypervisor files on ``gnt-cluster redist-conf``
1831 - ``gnt-instance reinstall`` can now reinstall multiple instances
1832 - Updated many command line parameters
1833 - Introduced new OS API version 15
1834 - No longer support a default hypervisor
1835 - Treat virtual LVs as inexistent
1836 - Improved job locking logic to reduce lock contention
1837 - Match instance and node names case insensitively
1838 - Reimplemented bash completion script to be more complete
1845 *(Released Thu, 4 Feb 2010)*
1847 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
1848 - Fix a string formatting bug
1849 - Improve safety of the code in some error paths
1850 - Improve data validation in the master of values returned from nodes
1856 *(Released Thu, 17 Dec 2009)*
1858 - Fix security issue due to missing validation of iallocator names; this
1859 allows local and remote execution of arbitrary executables
1860 - Fix failure of gnt-node list during instance removal
1861 - Ship the RAPI documentation in the archive
1867 *(Released Wed, 30 Sep 2009)*
1869 - Fixed many wrong messages
1870 - Fixed a few bugs related to the locking library
1871 - Fixed MAC checking at instance creation time
1872 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
1873 - Fixed a few issues related to signal handling in both daemons and
1875 - Fixed the example startup script provided
1876 - Fixed insserv dependencies in the example startup script (patch from
1878 - Fixed handling of drained nodes in the iallocator framework
1879 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
1881 - Fixed error related to invalid job IDs in job polling
1882 - Fixed job/opcode persistence on unclean master shutdown
1883 - Fixed handling of partial job processing after unclean master
1885 - Fixed error reporting from LUs, previously all errors were converted
1886 into execution errors
1887 - Fixed error reporting from burnin
1888 - Decreased significantly the memory usage of the job queue
1889 - Optimised slightly multi-job submission
1890 - Optimised slightly opcode loading
1891 - Backported the multi-job submit framework from the development
1892 branch; multi-instance start and stop should be faster
1893 - Added script to clean archived jobs after 21 days; this will reduce
1894 the size of the queue directory
1895 - Added some extra checks in disk size tracking
1896 - Added an example ethers hook script
1897 - Added a cluster parameter that prevents Ganeti from modifying of
1899 - Added more node information to RAPI responses
1900 - Added a ``gnt-job watch`` command that allows following the ouput of a
1902 - Added a bind-address option to ganeti-rapi
1903 - Added more checks to the configuration verify
1904 - Enhanced the burnin script such that some operations can be retried
1906 - Converted instance reinstall to multi-instance model
1912 *(Released Fri, 7 Aug 2009)*
1914 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
1915 to allow using the pre-2.0.2 behaviour in activation, if any existing
1916 instances have mismatched disk sizes in the configuration
1917 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
1918 any configuration mismatches for disk sizes
1919 - Added ``gnt-master cluste-failover --no-voting`` to allow master
1920 failover to work on two-node clusters
1921 - Fixed the ``--net`` option of ``gnt-backup import``, which was
1923 - Fixed detection of OS script errors in ``gnt-backup export``
1924 - Fixed exit code of ``gnt-backup export``
1930 *(Released Fri, 17 Jul 2009)*
1932 - Added experimental support for stripped logical volumes; this should
1933 enhance performance but comes with a higher complexity in the block
1934 device handling; stripping is only enabled when passing
1935 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
1936 affected even in the non-stripped mode
1937 - Improved resiliency against transient failures at the end of DRBD
1938 resyncs, and in general of DRBD resync checks
1939 - Fixed a couple of issues with exports and snapshot errors
1940 - Fixed a couple of issues in instance listing
1941 - Added display of the disk size in ``gnt-instance info``
1942 - Fixed checking for valid OSes in instance creation
1943 - Fixed handling of the "vcpus" parameter in instance listing and in
1944 general of invalid parameters
1945 - Fixed http server library, and thus RAPI, to handle invalid
1946 username/password combinations correctly; this means that now they
1947 report unauthorized for queries too, not only for modifications,
1948 allowing earlier detect of configuration problems
1949 - Added a new "role" node list field, equivalent to the master/master
1950 candidate/drained/offline flags combinations
1951 - Fixed cluster modify and changes of candidate pool size
1952 - Fixed cluster verify error messages for wrong files on regular nodes
1953 - Fixed a couple of issues with node demotion from master candidate role
1954 - Fixed node readd issues
1955 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
1956 - Added a new ``--no-voting`` option for masterfailover to fix failover
1957 on two-nodes clusters when the former master node is unreachable
1958 - Added instance reinstall over RAPI
1964 *(Released Tue, 16 Jun 2009)*
1966 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
1967 allow re-adding the start in single-user option (regression from 1.2)
1968 - the watcher writes the instance status to a file, to allow monitoring
1969 to report the instance status (from the master) based on cached
1970 results of the watcher's queries; while this can get stale if the
1971 watcher is being locked due to other work on the cluster, this is
1972 still an improvement
1973 - the watcher now also restarts the node daemon and the rapi daemon if
1975 - fixed the watcher to handle full and drained queue cases
1976 - hooks export more instance data in the environment, which helps if
1977 hook scripts need to take action based on the instance's properties
1978 (no longer need to query back into ganeti)
1979 - instance failovers when the instance is stopped do not check for free
1980 RAM, so that failing over a stopped instance is possible in low memory
1982 - rapi uses queries for tags instead of jobs (for less job traffic), and
1983 for cluster tags it won't talk to masterd at all but read them from
1985 - a couple of error handling fixes in RAPI
1986 - drbd handling: improved the error handling of inconsistent disks after
1987 resync to reduce the frequency of "there are some degraded disks for
1988 this instance" messages
1989 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
1990 error handling path called a wrong function name)
1996 *(Released Wed, 27 May 2009)*
1998 - no changes from rc5
2004 *(Released Wed, 20 May 2009)*
2006 - fix a couple of bugs (validation, argument checks)
2007 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
2008 - some small improvements to RAPI and IAllocator
2009 - make watcher automatically start the master daemon if down
2015 *(Released Mon, 27 Apr 2009)*
2017 - change the OS list to not require locks; this helps with big clusters
2018 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
2019 volume group is broken
2020 - ``gnt-instance info``, without any arguments, doesn't run for all
2021 instances anymore; either pass ``--all`` or pass the desired
2022 instances; this helps against mistakes on big clusters where listing
2023 the information for all instances takes a long time
2024 - miscellaneous doc and man pages fixes
2030 *(Released Wed, 8 Apr 2009)*
2032 - Change the internal locking model of some ``gnt-node`` commands, in
2033 order to reduce contention (and blocking of master daemon) when
2034 batching many creation/reinstall jobs
2035 - Fixes to Xen soft reboot
2036 - No longer build documentation at build time, instead distribute it in
2037 the archive, in order to reduce the need for the whole docbook/rst
2044 *(Released Fri, 27 Mar 2009)*
2046 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
2047 - Fix watcher startup sequence, improves the behaviour of busy clusters
2048 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
2049 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
2050 bind address and other places
2051 - Some documentation fixes and updates
2057 *(Released Mon, 2 Mar 2009)*
2059 - More documentation updates, now all docs should be more-or-less
2061 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
2063 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
2065 - Fix an issue related to $libdir/run/ganeti and cluster creation
2071 *(Released Thu, 19 Feb 2009)*
2073 - Xen PVM and KVM have switched the default value for the instance root
2074 disk to the first partition on the first drive, instead of the whole
2075 drive; this means that the OS installation scripts must be changed
2077 - Man pages have been updated
2078 - RAPI has been switched by default to HTTPS, and the exported functions
2079 should all work correctly
2080 - RAPI v1 has been removed
2081 - Many improvements to the KVM hypervisor
2082 - Block device errors are now better reported
2083 - Many other bugfixes and small improvements
2089 *(Released Mon, 26 Jan 2009)*
2091 - Version 2 is a general rewrite of the code and therefore the
2092 differences are too many to list, see the design document for 2.0 in
2093 the ``doc/`` subdirectory for more details
2094 - In this beta version there is not yet a migration path from 1.2 (there
2095 will be one in the final 2.0 release)
2096 - A few significant changes are:
2098 - all commands are executed by a daemon (``ganeti-masterd``) and the
2099 various ``gnt-*`` commands are just front-ends to it
2100 - all the commands are entered into, and executed from a job queue,
2101 see the ``gnt-job(8)`` manpage
2102 - the RAPI daemon supports read-write operations, secured by basic
2103 HTTP authentication on top of HTTPS
2104 - DRBD version 0.7 support has been removed, DRBD 8 is the only
2105 supported version (when migrating from Ganeti 1.2 to 2.0, you need
2106 to migrate to DRBD 8 first while still running Ganeti 1.2)
2107 - DRBD devices are using statically allocated minor numbers, which
2108 will be assigned to existing instances during the migration process
2109 - there is support for both Xen PVM and Xen HVM instances running on
2111 - KVM virtualization is supported too
2112 - file-based storage has been implemented, which means that it is
2113 possible to run the cluster without LVM and DRBD storage, for
2114 example using a shared filesystem exported from shared storage (and
2115 still have live migration)
2121 *(Released Tue, 13 Jan 2009)*
2123 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
2124 - Reuse the old instance mac address by default on instance import, if
2125 the instance name is the same.
2126 - Handle situations in which the node info rpc returns incomplete
2128 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
2129 - Improved version of batcher:
2131 - state file support
2132 - instance mac address support
2133 - support for HVM clusters/instances
2135 - Add an option to show the number of cpu sockets and nodes in
2137 - Support OSes that handle more than one version of the OS api (but do
2138 not change the current API in any other way)
2139 - Fix ``gnt-node migrate``
2140 - ``gnt-debug`` man page
2141 - Fixes various more typos and small issues
2142 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
2148 *(Released Wed, 24 Sep 2008)*
2150 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
2151 type of disk exported to fully virtualized instances.
2152 - provide access to the serial console of HVM instances
2153 - instance auto_balance flag, set by default. If turned off it will
2154 avoid warnings on cluster verify if there is not enough memory to fail
2155 over an instance. in the future it will prevent automatically failing
2156 it over when we will support that.
2157 - batcher tool for instance creation, see ``tools/README.batcher``
2158 - ``gnt-instance reinstall --select-os`` to interactively select a new
2159 operating system when reinstalling an instance.
2160 - when changing the memory amount on instance modify a check has been
2161 added that the instance will be able to start. also warnings are
2162 emitted if the instance will not be able to fail over, if auto_balance
2164 - documentation fixes
2165 - sync fields between ``gnt-instance list/modify/add/import``
2166 - fix a race condition in drbd when the sync speed was set after giving
2167 the device a remote peer.
2173 *(Released Tue, 22 Jul 2008)*
2175 - note: the allowed size and number of tags per object were reduced
2176 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
2177 - fixed twisted 8.x compatibility
2178 - fixed ``gnt-instance replace-disks`` with iallocator
2179 - add TCP keepalives on twisted connections to detect restarted nodes
2180 - disk increase support, see ``gnt-instance grow-disk``
2181 - implement bulk node/instance query for RAPI
2182 - add tags in node/instance listing (optional)
2183 - experimental migration (and live migration) support, read the man page
2184 for ``gnt-instance migrate``
2185 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
2186 has some small improvements in handling its state file
2192 *(Released Fri, 13 Jun 2008)*
2194 - Experimental readonly, REST-based remote API implementation;
2195 automatically started on master node, TCP port 5080, if enabled by
2196 ``--enable-rapi`` parameter to configure script.
2197 - Instance allocator support. Add and import instance accept a
2198 ``--iallocator`` parameter, and call that instance allocator to decide
2199 which node to use for the instance. The iallocator document describes
2200 what's expected from an allocator script.
2201 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
2202 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
2203 node is lost there is still enough memory to fail over the instances
2205 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
2206 ``gnt-cluster verify``, to check for site-specific compliance. All the
2207 hooks will run, and their output, if any, will be displayed. Any
2208 failing hook will make the verification return an error value.
2209 - ``gnt-cluster verify`` now checks that its peers are reachable on the
2210 primary and secondary interfaces
2211 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
2212 that is still declared as part of the cluster and has failed.
2213 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
2214 specifying output fields, that just adds the chosen fields to the
2216 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
2217 export from the filesystem.
2218 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
2219 have been added. Using them you can enable/disable acpi and pae
2220 support, and specify a path for a cd image to be exported to the
2221 instance. These parameters as the name suggest only work on HVM
2223 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
2224 PAE support will be set to the previously hardcoded values, but the
2225 (previously hardcoded) path to the CDROM ISO image will be unset and
2226 if required, needs to be set manually with ``gnt-instance modify``
2228 - The address to which an instance's VNC console is bound is now
2229 selectable per-instance, rather than being cluster wide. Of course
2230 this only applies to instances controlled via VNC, so currently just
2231 applies to HVM clusters.
2237 *(Released Mon, 18 Feb 2008)*
2239 - more tweaks to the disk activation code (especially helpful for DRBD)
2240 - change the default ``gnt-instance list`` output format, now there is
2241 one combined status field (see the manpage for the exact values this
2243 - some more fixes for the mac export to hooks change
2244 - make Ganeti not break with DRBD 8.2.x (which changed the version
2245 format in ``/proc/drbd``) (issue 24)
2246 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
2247 template, allowing migration from DRBD0.7+MD to DRBD8
2253 *(Released Wed, 30 Jan 2008)*
2255 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
2257 - add command aliases infrastructure and a few aliases
2258 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
2259 man pages and the ``--help`` option of ``gnt-node
2260 list``/``gnt-instance list``
2261 - fix ``gnt-backup list`` with down nodes (issue 21)
2262 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
2263 - fix the dist archive and add a check for including svn/git files in
2265 - some developer-related changes: improve the burnin and the QA suite,
2266 add an upload script for testing during development
2272 *(Released Wed, 16 Jan 2008)*
2274 - experimental HVM support, read the install document, section
2275 "Initializing the cluster"
2276 - allow for the PVM hypervisor per-instance kernel and initrd paths
2277 - add a new command ``gnt-cluster verify-disks`` which uses a new
2278 algorithm to improve the reconnection of the DRBD pairs if the device
2279 on the secondary node has gone away
2280 - make logical volume code auto-activate LVs at disk activation time
2281 - slightly improve the speed of activating disks
2282 - allow specification of the MAC address at instance creation time, and
2283 changing it later via ``gnt-instance modify``
2284 - fix handling of external commands that generate lots of output on
2286 - update documentation with regard to minimum version of DRBD8 supported
2292 *(Released Tue, 4 Dec 2007)*
2294 - Log the ``xm create`` output to the node daemon log on failure (to
2295 help diagnosing the error)
2296 - In debug mode, log all external commands output if failed to the logs
2297 - Change parsing of lvm commands to ignore stderr
2303 *(Released Wed, 28 Nov 2007)*
2305 - Another round of updates to the DRBD 8 code to deal with more failures
2306 in the replace secondary node operation
2307 - Some more logging of failures in disk operations (lvm, drbd)
2308 - A few documentation updates
2315 *(Released Tue, 13 Nov 2007)*
2317 - Change configuration file format from Python's Pickle to JSON.
2318 Upgrading is possible using the cfgupgrade utility.
2319 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
2320 faster replace disks and is more stable (DRBD 8 has many improvements
2321 compared to DRBD 0.7)
2322 - Added command line tags support (see man pages for ``gnt-instance``,
2323 ``gnt-node``, ``gnt-cluster``)
2324 - Added instance rename support
2325 - Added multi-instance startup/shutdown
2326 - Added cluster rename support
2327 - Added ``gnt-node evacuate`` to simplify some node operations
2328 - Added instance reboot operation that can speedup reboot as compared to
2330 - Soften the requirement that hostnames are in FQDN format
2331 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
2333 - Removed dependency on debian's patched fping that uses the
2334 non-standard ``-S`` option
2335 - Now the OS definitions are searched for in multiple, configurable
2336 paths (easier for distros to package)
2337 - Some changes to the hooks infrastructure (especially the new
2338 post-configuration update hook)
2339 - Other small bugfixes
2341 .. vim: set textwidth=72 syntax=rst :