7 *(Released Mon, 22 Aug 2011)*
9 Incompatible/important changes and bugfixes
10 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
12 - The default of the ``/2/instances/[instance_name]/rename`` RAPI
13 resource's ``ip_check`` parameter changed from ``True`` to ``False``
14 to match the underlying LUXI interface.
15 - The ``/2/nodes/[node_name]/evacuate`` RAPI resource was changed to use
16 body parameters, see :doc:`RAPI documentation <rapi>`. The server does
17 not maintain backwards-compatibility as the underlying operation
18 changed in an incompatible way. The RAPI client can talk to old
19 servers, but it needs to be told so as the return value changed.
20 - When creating file-based instances via RAPI, the ``file_driver``
21 parameter no longer defaults to ``loop`` and must be specified.
22 - The deprecated ``bridge`` NIC parameter is no longer supported. Use
24 - Support for the undocumented and deprecated RAPI instance creation
25 request format version 0 has been dropped. Use version 1, supported
26 since Ganeti 2.1.3 and :doc:`documented <rapi>`, instead.
27 - Pyparsing 1.4.6 or above is required, see :doc:`installation
28 documentation <install>`.
29 - The "cluster-verify" hooks are now executed per group by the
30 ``OP_CLUSTER_VERIFY_GROUP`` opcode. This maintains the same behavior
31 if you just run ``gnt-cluster verify``, which generates one opcode per
33 - The environment as passed to the OS scripts is cleared, and thus no
34 environment variables defined in the node daemon's environment will be
35 inherited by the scripts.
36 - The :doc:`iallocator <iallocator>` mode ``multi-evacuate`` has been
38 - :doc:`New iallocator modes <design-multi-reloc>` have been added to
39 support operations involving multiple node groups.
40 - Offline nodes are ignored when failing over an instance.
45 - The ganeti-htools project has been merged into the ganeti-core source
46 tree and will be built as part of Ganeti (see :doc:`install-quick`).
47 - Implemented support for :doc:`shared storage <design-shared-storage>`.
48 - Add support for disks larger than 2 TB in ``lvmstrap`` by supporting
49 GPT-style partition tables (requires `parted
50 <http://www.gnu.org/s/parted/>`_).
51 - Added support for floppy drive and 2nd CD-ROM drive in KVM hypervisor.
52 - Allowed adding tags on instance creation.
53 - Export instance tags to hooks (``INSTANCE_TAGS``, see :doc:`hooks`)
54 - Allow instances to be started in a paused state, enabling the user to
55 see the complete console output on boot using the console.
56 - Added new hypervisor flag to control default reboot behaviour
57 (``reboot_behavior``).
58 - Added support for KVM keymaps (hypervisor parameter ``keymap``).
59 - Improved out-of-band management support:
61 - Added ``gnt-node health`` command reporting the health status of
63 - Added ``gnt-node power`` command to manage power status of nodes.
64 - Added command for emergency power-off (EPO), ``gnt-cluster epo``.
66 - Instance migration can fall back to failover if instance is not
68 - Filters can be used when listing nodes, instances, groups and locks;
69 see *ganeti(7)* manpage.
70 - Added post-execution status as variables to :doc:`hooks <hooks>`
72 - Instance tags are exported/imported together with the instance.
73 - When given an explicit job ID, ``gnt-job info`` will work for archived
75 - Jobs can define dependencies on other jobs (not yet supported via
76 RAPI or command line, but used by internal commands and usable via
79 - Lock monitor (``gnt-debug locks``) shows jobs waiting for
82 - Instance failover is now available as a RAPI resource
83 (``/2/instances/[instance_name]/failover``).
84 - ``gnt-instance info`` defaults to static information if primary node
86 - Opcodes have a new ``comment`` attribute.
87 - Added basic SPICE support to KVM hypervisor.
89 Node group improvements
90 ~~~~~~~~~~~~~~~~~~~~~~~
92 - ``gnt-cluster verify`` has been modified to check groups separately,
93 thereby improving performance.
94 - Node group support has been added to ``gnt-cluster verify-disks``,
95 which now operates per node group.
96 - Watcher has been changed to work better with node groups.
98 - One process and state file per node group.
99 - Slow watcher in one group doesn't block other group's watcher.
101 - Added new command, ``gnt-group evacuate``, to move all instances in a
102 node group to other groups.
103 - Added ``gnt-instance change-group`` to move an instance to another
105 - ``gnt-cluster command`` and ``gnt-cluster copyfile`` now support
106 per-group operations.
107 - Node groups can be tagged.
108 - Some operations switch from an exclusive to a shared lock as soon as
110 - Instance's primary and secondary nodes' groups are now available as
111 query fields (``pnode.group``, ``pnode.group.uuid``, ``snodes.group``
112 and ``snodes.group.uuid``).
117 - Numerous updates to documentation and manpages.
119 - :doc:`RAPI <rapi>` documentation now has detailed parameter
122 - A lockset's internal lock is now also visible in lock monitor.
123 - Log messages from job queue workers now contain information about the
124 opcode they're processing.
125 - ``gnt-instance console`` no longer requires the instance lock.
126 - A short delay when waiting for job changes reduces the number of LUXI
127 requests significantly.
128 - DRBD metadata volumes are overwritten with zeros during disk creation.
134 *(Released Fri, 12 Aug 2011)*
136 This was the first beta release of the 2.5 series.
142 *(Released Tue, 23 Aug 2011)*
146 - Fixed documentation for importing with ``--src-dir`` option
147 - Fixed a bug in ``ensure-dirs`` with queue/archive permissions
148 - Fixed a parsing issue with DRBD 8.3.11 in the Linux kernel
154 *(Released Fri, 5 Aug 2011)*
156 Many bug-fixes and a few small features:
158 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
159 issues when you tried to add an instance with two MAC addresses in one
161 - KVM: fixed per-instance stored UID value
162 - KVM: configure bridged NICs at migration start
163 - KVM: Fix a bug where instance will not start with never KVM versions
165 - Added OS search path to ``gnt-cluster info``
166 - Fixed an issue with ``file_storage_dir`` where you were forced to
167 provide an absolute path, but the documentation states it is a
168 relative path, the documentation was right
169 - Added a new parameter to instance stop/start called ``--no-remember``
170 that will make the state change to not be remembered
171 - Implemented ``no_remember`` at RAPI level
172 - Improved the documentation
173 - Node evacuation: don't call IAllocator if node is already empty
174 - Fixed bug in DRBD8 replace disks on current nodes
175 - Fixed bug in recreate-disks for DRBD instances
176 - Moved assertion checking locks in ``gnt-instance replace-disks``
177 causing it to abort with not owning the right locks for some situation
178 - Job queue: Fixed potential race condition when cancelling queued jobs
179 - Fixed off-by-one bug in job serial generation
180 - ``gnt-node volumes``: Fix instance names
181 - Fixed aliases in bash completion
182 - Fixed a bug in reopening log files after being sent a SIGHUP
183 - Added a flag to burnin to allow specifying VCPU count
184 - Bugfixes to non-root Ganeti configuration
190 *(Released Thu, 12 May 2011)*
192 Many bug-fixes and a few new small features:
194 - Fixed a bug related to log opening failures
195 - Fixed a bug in instance listing with orphan instances
196 - Fixed a bug which prevented resetting the cluster-level node parameter
197 ``oob_program`` to the default
198 - Many fixes related to the ``cluster-merge`` tool
199 - Fixed a race condition in the lock monitor, which caused failures
200 during (at least) creation of many instances in parallel
201 - Improved output for gnt-job info
202 - Removed the quiet flag on some ssh calls which prevented debugging
204 - Improved the N+1 failure messages in cluster verify by actually
205 showing the memory values (needed and available)
206 - Increased lock attempt timeouts so that when executing long operations
207 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
208 too early and thus prevent the use of the 'fair' mechanism
209 - Changed instance query data (``gnt-instance info``) to not acquire
210 locks unless needed, thus allowing its use on locked instance if only
211 static information is asked for
212 - Improved behaviour with filesystems that do not support rename on an
214 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
215 kept locks on all nodes during the wipe, which is unneeded
216 - Fixed ``gnt-watcher`` handling of errors during hooks execution
217 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
218 10GiB) which caused the wipe to fail right at the end in some cases
219 - Fixed master IP activation when doing master failover with no-voting
220 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
221 the master node itself
222 - Fixed potential data-loss in under disk full conditions, where Ganeti
223 wouldn't check correctly the return code and would consider
224 partially-written files 'correct'
225 - Fixed bug related to multiple VGs and DRBD disk replacing
226 - Added new disk parameter ``metavg`` that allows placement of the meta
227 device for DRBD in a different volume group
228 - Fixed error handling in the node daemon when the system libc doesn't
229 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
230 - Fixed lock release during replace-disks, which kept cluster-wide locks
231 when doing disk replaces with an iallocator script
232 - Added check for missing bridges in cluster verify
233 - Handle EPIPE errors while writing to the terminal better, so that
234 piping the output to e.g. ``less`` doesn't cause a backtrace
235 - Fixed rare case where a ^C during Luxi calls could have been
236 interpreted as server errors, instead of simply terminating
237 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
239 - Added a few more parameters to the KVM hypervisor, allowing a second
240 CDROM, custom disk type for CDROMs and a floppy image
241 - Removed redundant message in instance rename when the name is given
243 - Added option to ``gnt-instance recreate-disks`` to allow creating the
244 disks on new nodes, allowing recreation when the original instance
245 nodes are completely gone
246 - Added option when converting disk templates to DRBD to skip waiting
247 for the resync, in order to make the instance available sooner
248 - Added two new variables to the OS scripts environment (containing the
250 - Made the root_path and optional parameter for the xen-pvm hypervisor,
251 to allow use of ``pvgrub`` as bootloader
252 - Changed the instance memory modifications to only check out-of-memory
253 conditions on memory increases, and turned the secondary node warnings
254 into errors (they can still be overridden via ``--force``)
255 - Fixed the handling of a corner case when the Python installation gets
256 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
257 to execute a command that doesn't exist
258 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
259 node of the instance returned failures during instance shutdown; this
260 adds the option ``--ignore-consistency`` to gnt-instance move
262 And as usual, various improvements to the error messages, documentation
268 *(Released Wed, 09 Mar 2011)*
270 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
271 the RAPI users file if run twice (even with ``--dry-run``).
273 The release fixes that bug (nothing else changed).
279 *(Released Mon, 07 Mar 2011)*
281 Final 2.4.0 release. Just a few small fixes:
283 - Fixed RAPI node evacuate
284 - Fixed the kvm-ifup script
285 - Fixed internal error handling for special job cases
286 - Updated man page to specify the escaping feature for options
292 *(Released Mon, 28 Feb 2011)*
294 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
295 that this feature wiped the disks of the wrong instance, leading to loss
300 - Fixed title of query field containing instance name
301 - Expanded the glossary in the documentation
302 - Fixed one unittest (internal issue)
308 *(Released Mon, 21 Feb 2011)*
310 A number of bug fixes plus just a couple functionality changes.
312 On the user-visible side, the ``gnt-* list`` command output has changed
313 with respect to "special" field states. The current rc1 style of display
314 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
315 the default output mode special fields states are displayed as follows:
317 - Offline resource: ``*``
318 - Unavailable/not applicable: ``-``
319 - Data missing (RPC failure): ``?``
320 - Unknown field: ``??``
322 Another user-visible change is the addition of ``--force-join`` to
327 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
328 - Fixed regression in RAPI/instance reinstall where all parameters were
329 required (instead of optional)
330 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
331 - Fixed iallocator usage (offline nodes were not considered offline)
332 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
333 - Fixed hypervisor and OS parameter validation with respect to
335 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
337 - Fixed ``tools/listrunner`` with respect to agent-based usage
343 *(Released Fri, 4 Feb 2011)*
345 Many changes and fixes since the beta1 release. While there were some
346 internal changes, the code has been mostly stabilised for the RC
349 Note: the dumb allocator was removed in this release, as it was not kept
350 up-to-date with the IAllocator protocol changes. It is recommended to
351 use the ``hail`` command from the ganeti-htools package.
353 Note: the 2.4 and up versions of Ganeti are not compatible with the
354 0.2.x branch of ganeti-htools. You need to upgrade to
355 ganeti-htools-0.3.0 (or later).
357 Regressions fixed from 2.3
358 ~~~~~~~~~~~~~~~~~~~~~~~~~~
360 - Fixed the ``gnt-cluster verify-disks`` command
361 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
363 - Fixed disk adoption breakage
364 - Fixed wrong headers in instance listing for field aliases
369 - Fixed corner case in KVM handling of NICs
370 - Fixed many cases of wrong handling of non-vm_capable nodes
371 - Fixed a bug where a missing instance symlink was not possible to
372 recreate with any ``gnt-*`` command (now ``gnt-instance
373 activate-disks`` does it)
374 - Fixed the volume group name as reported by ``gnt-cluster
376 - Increased timeouts for the import-export code, hopefully leading to
377 fewer aborts due network or instance timeouts
378 - Fixed bug in ``gnt-node list-storage``
379 - Fixed bug where not all daemons were started on cluster
380 initialisation, but only at the first watcher run
381 - Fixed many bugs in the OOB implementation
382 - Fixed watcher behaviour in presence of instances with offline
384 - Fixed instance list output for instances running on the wrong node
385 - a few fixes to the cluster-merge tool, but it still cannot merge
386 multi-node groups (currently it is not recommended to use this tool)
392 - Improved network configuration for the KVM hypervisor
393 - Added e1000 as a supported NIC for Xen-HVM
394 - Improved the lvmstrap tool to also be able to use partitions, as
395 opposed to full disks
396 - Improved speed of disk wiping (the cluster parameter
397 ``prealloc_wipe_disks``, so that it has a low impact on the total time
398 of instance creations
399 - Added documentation for the OS parameters
400 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
401 hypervisor is not responding
402 - Added display of blacklisted and hidden OS information in
404 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
405 NIC and node parameters, which might create problems with currently
406 invalid (but undetected) configuration files, but prevents validation
407 failures when unrelated parameters are modified
408 - Changed cluster initialisation to wait for the master daemon to become
410 - Expanded the RAPI interface:
412 - Added config redistribution resource
413 - Added activation/deactivation of instance disks
414 - Added export of console information
416 - Implemented log file reopening on SIGHUP, which allows using
417 logrotate(8) for the Ganeti log files
418 - Added a basic OOB helper script as an example
424 *(Released Fri, 14 Jan 2011)*
429 - Fixed timezone issues when formatting timestamps
430 - Added support for node groups, available via ``gnt-group`` and other
432 - Added out-of-band framework and management, see :doc:`design
433 document <design-oob>`
434 - Removed support for roman numbers from ``gnt-node list`` and
435 ``gnt-instance list``.
436 - Allowed modification of master network interface via ``gnt-cluster
437 modify --master-netdev``
438 - Accept offline secondaries while shutting down instance disks
439 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
440 - Added support for multiple LVM volume groups
441 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
443 - Added commands to list available fields:
445 - ``gnt-node list-fields``
446 - ``gnt-group list-fields``
447 - ``gnt-instance list-fields``
449 - Updated documentation and man pages
454 - Moved ``rapi_users`` file into separate directory, now named
455 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
457 - Added new tool for running commands on many machines,
458 ``tools/ganeti-listrunner``
459 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
460 improving the ``gnt-instance console`` output
461 - Allowed customisation of disk index separator at ``configure`` time
462 - Export node group allocation policy to :doc:`iallocator <iallocator>`
463 - Added support for non-partitioned md disks in ``lvmstrap``
464 - Added script to gracefully power off KVM instances
465 - Split ``utils`` module into smaller parts
466 - Changed query operations to return more detailed information, e.g.
467 whether an information is unavailable due to an offline node. To use
468 this new functionality, the LUXI call ``Query`` must be used. Field
469 information is now stored by the master daemon and can be retrieved
470 using ``QueryFields``. Instances, nodes and groups can also be queried
471 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
472 exposed via RAPI). The following commands make use of this
473 infrastructure change:
476 - ``gnt-group list-fields``
478 - ``gnt-node list-fields``
479 - ``gnt-instance list``
480 - ``gnt-instance list-fields``
481 - ``gnt-debug locks``
486 - New RAPI resources (see :doc:`rapi`):
490 - ``/2/groups/[group_name]``
491 - ``/2/groups/[group_name]/assign-nodes``
492 - ``/2/groups/[group_name]/modify``
493 - ``/2/groups/[group_name]/rename``
494 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
498 - Implemented ``no_install`` for instance creation
499 - Implemented OS parameters for instance reinstallation, allowing
500 use of special settings on reinstallation (e.g. for preserving data)
505 - Added IPv6 support in import/export
506 - Pause DRBD synchronization while wiping disks on instance creation
507 - Updated unittests and QA scripts
508 - Improved network parameters passed to KVM
509 - Converted man pages from docbook to reStructuredText
515 *(Released Mon, 20 Dec 2010)*
517 Released version 2.3.1~rc1 without any changes.
523 *(Released Wed, 1 Dec 2010)*
525 - impexpd: Disable OpenSSL compression in socat if possible (backport
526 from master, commit e90739d625b, see :doc:`installation guide
527 <install-quick>` for details)
528 - Changed unittest coverage report to exclude test scripts
529 - Added script to check version format
535 *(Released Wed, 1 Dec 2010)*
537 Released version 2.3.0~rc1 without any changes.
543 *(Released Fri, 19 Nov 2010)*
545 A number of bugfixes and documentation updates:
547 - Update ganeti-os-interface documentation
548 - Fixed a bug related to duplicate MACs or similar items which should be
550 - Fix breakage in OS state modify
551 - Reinstall instance: disallow offline secondaries (fixes bug related to
552 OS changing but reinstall failing)
553 - plus all the other fixes between 2.2.1 and 2.2.2
559 *(Released Tue, 2 Nov 2010)*
561 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
562 - Fixed master failover race with watcher
563 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
565 - Accept previously stopped instance for export with instance removal
566 - Simplify and extend the environment variables for instance OS scripts
567 - Added new node flags, ``master_capable`` and ``vm_capable``
568 - Added optional instance disk wiping prior during allocation. This is a
569 cluster-wide option and can be set/modified using
570 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
571 - Added IPv6 support, see :doc:`design document <design-2.3>` and
573 - Added a new watcher option (``--ignore-pause``)
574 - Added option to ignore offline node on instance start/stop
575 (``--ignore-offline``)
576 - Allow overriding OS parameters with ``gnt-instance reinstall``
577 - Added ability to change node's secondary IP address using ``gnt-node
579 - Implemented privilege separation for all daemons except
580 ``ganeti-noded``, see ``configure`` options
581 - Complain if an instance's disk is marked faulty in ``gnt-cluster
583 - Implemented job priorities (see ``ganeti(7)`` manpage)
584 - Ignore failures while shutting down instances during failover from
586 - Exit daemon's bootstrap process only once daemon is ready
587 - Export more information via ``LUInstanceQuery``/remote API
588 - Improved documentation, QA and unittests
589 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
590 restart if the file was created or changed
591 - Added LUXI protocol version sent with each request and response,
592 allowing detection of server/client mismatches
593 - Moved the Python scripts among gnt-* and ganeti-* into modules
594 - Moved all code related to setting up SSH to an external script,
596 - Infrastructure changes for node group support in future versions
602 *(Released Fri, 19 Nov 2010)*
604 A few small bugs fixed, and some improvements to the build system:
606 - Fix documentation regarding conversion to drbd
607 - Fix validation of parameters in cluster modify (``gnt-cluster modify
609 - Fix error handling in node modify with multiple changes
610 - Allow remote imports without checked names
616 *(Released Tue, 19 Oct 2010)*
618 - Disable SSL session ID cache in RPC client
624 *(Released Thu, 14 Oct 2010)*
626 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
627 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
629 - Fix problems with interaction between Curl and Python's HTTP server,
630 resulting in increased speed in many RPC calls
631 - Improve our release script to prevent breakage with older aclocal and
638 *(Released Thu, 7 Oct 2010)*
640 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
641 value retrieved from master
642 - Added support for blacklisted or hidden OS definitions
643 - Added simple lock monitor (accessible via (``gnt-debug locks``)
644 - Added support for -mem-path in KVM hypervisor abstraction layer
645 - Allow overriding instance parameters in tool for inter-cluster
646 instance moves (``tools/move-instance``)
647 - Improved opcode summaries (e.g. in ``gnt-job list``)
648 - Improve consistency of OS listing by sorting it
649 - Documentation updates
655 *(Released Fri, 8 Oct 2010)*
657 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
663 *(Released Mon, 4 Oct 2010)*
665 - Fixed regression in ``gnt-instance rename``
671 *(Released Wed, 22 Sep 2010)*
673 - Fixed OS_VARIANT variable for OS scripts
674 - Fixed cluster tag operations via RAPI
675 - Made ``setup-ssh`` exit with non-zero code if an error occurred
676 - Disabled RAPI CA checks in watcher
682 *(Released Mon, 23 Aug 2010)*
684 - Support DRBD versions of the format "a.b.c.d"
686 - Re-introduce support for usage from multiple threads in RAPI client
687 - Instance renames and modify via RAPI
688 - Work around race condition between processing and archival in job
690 - Mark opcodes following failed one as failed, too
691 - Job field ``lock_status`` was removed due to difficulties making it
692 work with the changed job queue in Ganeti 2.2; a better way to monitor
693 locks is expected for a later 2.2.x release
694 - Fixed dry-run behaviour with many commands
695 - Support ``ssh-agent`` again when adding nodes
696 - Many additional bugfixes
702 *(Released Fri, 30 Jul 2010)*
704 Important change: the internal RPC mechanism between Ganeti nodes has
705 changed from using a home-grown http library (based on the Python base
706 libraries) to use the PycURL library. This requires that PycURL is
707 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
708 against GnuTLS by default. cURL's support for GnuTLS had known issues
709 before cURL 7.21.0 and we recommend using the latest cURL release or
710 linking against OpenSSL. Most other distributions already link PycURL
711 and cURL against OpenSSL. The command::
713 python -c 'import pycurl; print pycurl.version'
715 can be used to determine the libraries PycURL and cURL are linked
718 Other significant changes:
720 - Rewrote much of the internals of the job queue, in order to achieve
721 better parallelism; this decouples job query operations from the job
722 processing, and it should allow much nicer behaviour of the master
723 daemon under load, and it also has uncovered some long-standing bugs
724 related to the job serialisation (now fixed)
725 - Added a default iallocator setting to the cluster parameters,
726 eliminating the need to always pass nodes or an iallocator for
727 operations that require selection of new node(s)
728 - Added experimental support for the LXC virtualization method
729 - Added support for OS parameters, which allows the installation of
730 instances to pass parameter to OS scripts in order to customise the
732 - Added a hypervisor parameter controlling the migration type (live or
733 non-live), since hypervisors have various levels of reliability; this
734 has renamed the 'live' parameter to 'mode'
735 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
736 logical volumes, meaning that cluster verify will ignore them and not
737 flag their presence as errors
738 - The watcher will now reset the error count for failed instances after
739 8 hours, thus allowing self-healing if the problem that caused the
740 instances to be down/fail to start has cleared in the meantime
741 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
742 check for, and warn, if the drbd module parameter ``usermode_helper``
743 is not consistent with the cluster-wide setting; this is needed to
744 make diagnose easier of failed drbd creations
745 - Started adding base IPv6 support, but this is not yet
746 enabled/available for use
747 - Rename operations (cluster, instance) will now return the new name,
748 which is especially useful if a short name was passed in
749 - Added support for instance migration in RAPI
750 - Added a tool to pre-configure nodes for the SSH setup, before joining
751 them to the cluster; this will allow in the future a simplified model
752 for node joining (but not yet fully enabled in 2.2); this needs the
753 paramiko python library
754 - Fixed handling of name-resolving errors
755 - Fixed consistency of job results on the error path
756 - Fixed master-failover race condition when executed multiple times in
758 - Fixed many bugs related to the job queue (mostly introduced during the
759 2.2 development cycle, so not all are impacting 2.1)
760 - Fixed instance migration with missing disk symlinks
761 - Fixed handling of unknown jobs in ``gnt-job archive``
762 - And many other small fixes/improvements
766 - Enhanced both the unittest and the QA coverage
767 - Switched the opcode validation to a generic model, and extended the
768 validation to all opcode parameters
769 - Changed more parts of the code that write shell scripts to use the
771 - Switched the master daemon to use the asyncore library for the Luxi
778 *(Released Thu, 17 Jun 2010)*
780 - Added tool (``move-instance``) and infrastructure to move instances
781 between separate clusters (see :doc:`separate documentation
782 <move-instance>` and :doc:`design document <design-2.2>`)
783 - Added per-request RPC timeout
784 - RAPI now requires a Content-Type header for requests with a body (e.g.
785 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
786 :rfc:`2616` (HTTP/1.1), section 7.2.1)
787 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
789 - Implemented initial support for running Ganeti daemons as separate
790 users, see configure-time flags ``--with-user-prefix`` and
791 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
793 - Instances can be removed after export (``gnt-backup export
795 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
796 key (instead of 1024 bit)
797 - Added new cluster configuration file for cluster domain secret
798 - Import/export now use SSL instead of SSH
799 - Added support for showing estimated time when exporting an instance,
800 see the ``ganeti-os-interface(7)`` manpage and look for
807 *(Released Tue, 16 Nov 2010)*
809 Some more bugfixes. Unless critical bugs occur, this will be the last
812 - Fix case of MAC special-values
813 - Fix mac checker regex
814 - backend: Fix typo causing "out of range" error
815 - Add missing --units in gnt-instance list man page
821 *(Released Tue, 24 Aug 2010)*
824 - Don't ignore secondary node silently on non-mirrored disk templates
826 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
827 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
828 - Properly document the usb_mouse parameter
829 - Fix path in ganeti-rapi(8) (issue 116)
830 - Adjust error message when the ganeti user's .ssh directory is
832 - Add same-node-check when changing the disk template to drbd
838 *(Released Fri, 16 Jul 2010)*
841 - Add an option to only select some reboot types during qa/burnin.
842 (on some hypervisors consequent reboots are not supported)
843 - Fix infrequent race condition in master failover. Sometimes the old
844 master ip address would be still detected as up for a short time
845 after it was removed, causing failover to fail.
846 - Decrease mlockall warnings when the ctypes module is missing. On
847 Python 2.4 we support running even if no ctypes module is installed,
848 but we were too verbose about this issue.
849 - Fix building on old distributions, on which man doesn't have a
851 - Fix RAPI not to ignore the MAC address on instance creation
852 - Implement the old instance creation format in the RAPI client.
858 *(Released Thu, 01 Jul 2010)*
860 A small bugfix release:
861 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
862 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
863 a non-existing option
864 - Fix instance create: the --force-variant option was ignored
865 - Improve pylint 0.21 compatibility and warnings with Python 2.6
866 - Fix modify node storage with non-FQDN arguments
867 - Fix RAPI client to authenticate under Python 2.6 when used
868 for more than 5 requests needing authentication
869 - Fix gnt-instance modify -t (storage) giving a wrong error message
870 when converting a non-shutdown drbd instance to plain
876 *(Released Fri, 18 Jun 2010)*
878 A small bugfix release:
880 - Fix live migration of KVM instances started with older Ganeti
881 versions which had fewer hypervisor parameters
882 - Fix gnt-instance grow-disk on down instances
883 - Fix an error-reporting bug during instance migration
884 - Better checking of the ``--net`` and ``--disk`` values, to avoid
885 silently ignoring broken ones
886 - Fix an RPC error reporting bug affecting, for example, RAPI client
888 - Fix bug triggered by different API version os-es on different nodes
889 - Fix a bug in instance startup with custom hvparams: OS level
890 parameters would fail to be applied.
891 - Fix the RAPI client under Python 2.6 (but more work is needed to
892 make it work completely well with OpenSSL)
893 - Fix handling of errors when resolving names from DNS
899 *(Released Thu, 3 Jun 2010)*
901 A medium sized development cycle. Some new features, and some
902 fixes/small improvements/cleanups.
907 The node deamon now tries to mlock itself into memory, unless the
908 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
909 its logs, and falls back to console logging. This allows emergency
910 features such as ``gnt-node powercycle`` to work even in the event of a
911 broken node disk (tested offlining the disk hosting the node's
912 filesystem and dropping its memory caches; don't try this at home)
914 KVM: add vhost-net acceleration support. It can be tested with a new
915 enough version of the kernel and of qemu-kvm.
917 KVM: Add instance chrooting feature. If you use privilege dropping for
918 your VMs you can also now force them to chroot to an empty directory,
919 before starting the emulated guest.
921 KVM: Add maximum migration bandwith and maximum downtime tweaking
922 support (requires a new-enough version of qemu-kvm).
924 Cluster verify will now warn if the master node doesn't have the master
927 Add a new (incompatible) instance creation request format to RAPI which
928 supports all parameters (previously only a subset was supported, and it
929 wasn't possible to extend the old format to accomodate all the new
930 features. The old format is still supported, and a client can check for
931 this feature, before using it, by checking for its presence in the
932 ``features`` RAPI resource.
934 Now with ancient latin support. Try it passing the ``--roman`` option to
935 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
936 (requires the python-roman module to be installed, in order to work).
941 As usual many internal code refactorings, documentation updates, and
944 - Lots of improvements and cleanups to the experimental Remote API
945 (RAPI) client library.
946 - A new unit test suite for the core daemon libraries.
947 - A fix to creating missing directories makes sure the umask is not
948 applied anymore. This enforces the same directory permissions
950 - Better handling terminating daemons with ctrl+c (used when running
951 them in debugging mode).
952 - Fix a race condition in live migrating a KVM instance, when stat()
953 on the old proc status file returned EINVAL, which is an unexpected
955 - Fixed manpage checking with newer man and utf-8 charachters. But now
956 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
962 *(Released Fri, 7 May 2010)*
964 Fix a bug which prevented untagged KVM instances from starting.
970 *(Released Fri, 7 May 2010)*
972 Another release with a long development cycle, during which many
973 different features were added.
978 The KVM hypervisor now can run the individual instances as non-root, to
979 reduce the impact of a VM being hijacked due to bugs in the
980 hypervisor. It is possible to run all instances as a single (non-root)
981 user, to manually specify a user for each instance, or to dynamically
982 allocate a user out of a cluster-wide pool to each instance, with the
983 guarantee that no two instances will run under the same user ID on any
986 An experimental RAPI client library, that can be used standalone
987 (without the other Ganeti libraries), is provided in the source tree as
988 ``lib/rapi/client.py``. Note this client might change its interface in
989 the future, as we iterate on its capabilities.
991 A new command, ``gnt-cluster renew-crypto`` has been added to easily
992 replace the cluster's certificates and crypto keys. This might help in
993 case they have been compromised, or have simply expired.
995 A new disk option for instance creation has been added that allows one
996 to "adopt" currently existing logical volumes, with data
997 preservation. This should allow easier migration to Ganeti from
998 unmanaged (or managed via other software) instances.
1000 Another disk improvement is the possibility to convert between redundant
1001 (DRBD) and plain (LVM) disk configuration for an instance. This should
1002 allow better scalability (starting with one node and growing the
1003 cluster, or shrinking a two-node cluster to one node).
1005 A new feature that could help with automated node failovers has been
1006 implemented: if a node sees itself as offline (by querying the master
1007 candidates), it will try to shutdown (hard) all instances and any active
1008 DRBD devices. This reduces the risk of duplicate instances if an
1009 external script automatically failovers the instances on such nodes. To
1010 enable this, the cluster parameter ``maintain_node_health`` should be
1011 enabled; in the future this option (per the name) will enable other
1012 automatic maintenance features.
1014 Instance export/import now will reuse the original instance
1015 specifications for all parameters; that means exporting an instance,
1016 deleting it and the importing it back should give an almost identical
1017 instance. Note that the default import behaviour has changed from
1018 before, where it created only one NIC; now it recreates the original
1021 Cluster verify has added a few new checks: SSL certificates validity,
1022 /etc/hosts consistency across the cluster, etc.
1027 As usual, many internal changes were done, documentation fixes,
1030 - Fixed cluster initialization with disabled cluster storage (regression
1031 introduced in 2.1.1)
1032 - File-based storage supports growing the disks
1033 - Fixed behaviour of node role changes
1034 - Fixed cluster verify for some corner cases, plus a general rewrite of
1035 cluster verify to allow future extension with more checks
1036 - Fixed log spamming by watcher and node daemon (regression introduced
1038 - Fixed possible validation issues when changing the list of enabled
1040 - Fixed cleanup of /etc/hosts during node removal
1041 - Fixed RAPI response for invalid methods
1042 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
1043 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
1044 from ide disks, etc.)
1045 - Allow OS changes without re-installation (to record a changed OS
1046 outside of Ganeti, or to allow OS renames)
1047 - Allow instance creation without OS installation (useful for example if
1048 the OS will be installed manually, or restored from a backup not in
1050 - Implemented option to make cluster ``copyfile`` use the replication
1052 - Added list of enabled hypervisors to ssconf (possibly useful for
1054 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
1056 - A partial form of node re-IP is possible via node readd, which now
1057 allows changed node primary IP
1058 - Command line utilities now show an informational message if the job is
1060 - The logs of the master daemon now show the PID/UID/GID of the
1067 *(Released Fri, 12 Mar 2010)*
1069 During the 2.1.0 long release candidate cycle, a lot of improvements and
1070 changes have accumulated with were released later as 2.1.1.
1075 The node evacuate command (``gnt-node evacuate``) was significantly
1076 rewritten, and as such the IAllocator protocol was changed - a new
1077 request type has been added. This unfortunate change during a stable
1078 series is designed to improve performance of node evacuations; on
1079 clusters with more than about five nodes and which are well-balanced,
1080 evacuation should proceed in parallel for all instances of the node
1081 being evacuated. As such, any existing IAllocator scripts need to be
1082 updated, otherwise the above command will fail due to the unknown
1083 request. The provided "dumb" allocator has not been updated; but the
1084 ganeti-htools package supports the new protocol since version 0.2.4.
1086 Another important change is increased validation of node and instance
1087 names. This might create problems in special cases, if invalid host
1088 names are being used.
1090 Also, a new layer of hypervisor parameters has been added, that sits at
1091 OS level between the cluster defaults and the instance ones. This allows
1092 customisation of virtualization parameters depending on the installed
1093 OS. For example instances with OS 'X' may have a different KVM kernel
1094 (or any other parameter) than the cluster defaults. This is intended to
1095 help managing a multiple OSes on the same cluster, without manual
1096 modification of each instance's parameters.
1098 A tool for merging clusters, ``cluster-merge``, has been added in the
1099 tools sub-directory.
1104 - Improved the int/float conversions that should make the code more
1105 robust in face of errors from the node daemons
1106 - Fixed the remove node code in case of internal configuration errors
1107 - Fixed the node daemon behaviour in face of inconsistent queue
1108 directory (e.g. read-only file-system where we can't open the files
1110 - Fixed the behaviour of gnt-node modify for master candidate demotion;
1111 now it either aborts cleanly or, if given the new "auto_promote"
1112 parameter, will automatically promote other nodes as needed
1113 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
1114 completely prevent Ganeti from working
1115 - Fixed bug for instance export when not all disks were successfully
1117 - Fixed behaviour of node add when the new node is slow in starting up
1119 - Fixed handling of signals in the LUXI client, which should improve
1120 behaviour of command-line scripts
1121 - Added checks for invalid node/instance names in the configuration (now
1122 flagged during cluster verify)
1123 - Fixed watcher behaviour for disk activation errors
1124 - Fixed two potentially endless loops in http library, which led to the
1125 RAPI daemon hanging and consuming 100% CPU in some cases
1126 - Fixed bug in RAPI daemon related to hashed passwords
1127 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
1129 - Enhanced compatibility with non-Debian OSes, but not using absolute
1130 path in some commands and allowing customisation of the ssh
1131 configuration directory
1132 - Fixed possible future issue with new Python versions by abiding to the
1133 proper use of ``__slots__`` attribute on classes
1134 - Added checks that should prevent directory traversal attacks
1135 - Many documentation fixes based on feedback from users
1140 - Added an "early_release" more for instance replace disks and node
1141 evacuate, where we release locks earlier and thus allow higher
1142 parallelism within the cluster
1143 - Added watcher hooks, intended to allow the watcher to restart other
1144 daemons (e.g. from the ganeti-nbma project), but they can be used of
1145 course for any other purpose
1146 - Added a compile-time disable for DRBD barriers, to increase
1147 performance if the administrator trusts the power supply or the
1148 storage system to not lose writes
1149 - Added the option of using syslog for logging instead of, or in
1150 addition to, Ganeti's own log files
1151 - Removed boot restriction for paravirtual NICs for KVM, recent versions
1152 can indeed boot from a paravirtual NIC
1153 - Added a generic debug level for many operations; while this is not
1154 used widely yet, it allows one to pass the debug value all the way to
1156 - Enhanced the hooks environment for instance moves (failovers,
1157 migrations) where the primary/secondary nodes changed during the
1158 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
1159 - Enhanced data validations for many user-supplied values; one important
1160 item is the restrictions imposed on instance and node names, which
1161 might reject some (invalid) host names
1162 - Add a configure-time option to disable file-based storage, if it's not
1163 needed; this allows greater security separation between the master
1164 node and the other nodes from the point of view of the inter-node RPC
1166 - Added user notification in interactive tools if job is waiting in the
1167 job queue or trying to acquire locks
1168 - Added log messages when a job is waiting for locks
1169 - Added filtering by node tags in instance operations which admit
1170 multiple instances (start, stop, reboot, reinstall)
1171 - Added a new tool for cluster mergers, ``cluster-merge``
1172 - Parameters from command line which are of the form ``a=b,c=d`` can now
1173 use backslash escapes to pass in values which contain commas,
1174 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
1176 - For KVM, the instance name is the first parameter passed to KVM, so
1177 that it's more visible in the process list
1183 *(Released Tue, 2 Mar 2010)*
1185 Ganeti 2.1 brings many improvements with it. Major changes:
1187 - Added infrastructure to ease automated disk repairs
1188 - Added new daemon to export configuration data in a cheaper way than
1189 using the remote API
1190 - Instance NICs can now be routed instead of being associated with a
1192 - Improved job locking logic to reduce impact of jobs acquiring multiple
1193 locks waiting for other long-running jobs
1195 In-depth implementation details can be found in the Ganeti 2.1 design
1201 - Added chroot hypervisor
1202 - Added more options to xen-hvm hypervisor (``kernel_path`` and
1204 - Added more options to xen-pvm hypervisor (``use_bootloader``,
1205 ``bootloader_path`` and ``bootloader_args``)
1206 - Added the ``use_localtime`` option for the xen-hvm and kvm
1207 hypervisors, and the default value for this has changed to false (in
1208 2.0 xen-hvm always enabled it)
1209 - Added luxi call to submit multiple jobs in one go
1210 - Added cluster initialization option to not modify ``/etc/hosts``
1212 - Added network interface parameters
1213 - Added dry run mode to some LUs
1214 - Added RAPI resources:
1216 - ``/2/instances/[instance_name]/info``
1217 - ``/2/instances/[instance_name]/replace-disks``
1218 - ``/2/nodes/[node_name]/evacuate``
1219 - ``/2/nodes/[node_name]/migrate``
1220 - ``/2/nodes/[node_name]/role``
1221 - ``/2/nodes/[node_name]/storage``
1222 - ``/2/nodes/[node_name]/storage/modify``
1223 - ``/2/nodes/[node_name]/storage/repair``
1225 - Added OpCodes to evacuate or migrate all instances on a node
1226 - Added new command to list storage elements on nodes (``gnt-node
1227 list-storage``) and modify them (``gnt-node modify-storage``)
1228 - Added new ssconf files with master candidate IP address
1229 (``ssconf_master_candidates_ips``), node primary IP address
1230 (``ssconf_node_primary_ips``) and node secondary IP address
1231 (``ssconf_node_secondary_ips``)
1232 - Added ``ganeti-confd`` and a client library to query the Ganeti
1233 configuration via UDP
1234 - Added ability to run hooks after cluster initialization and before
1236 - Added automatic mode for disk replace (``gnt-instance replace-disks
1238 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
1239 after catastrophic data-loss
1240 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
1242 - Added ``gnt-instance move`` command to move instances
1243 - Added ``gnt-cluster watcher`` command to control watcher
1244 - Added ``gnt-node powercycle`` command to powercycle nodes
1245 - Added new job status field ``lock_status``
1246 - Added parseable error codes to cluster verification (``gnt-cluster
1247 verify --error-codes``) and made output less verbose (use
1248 ``--verbose`` to restore previous behaviour)
1249 - Added UUIDs to the main config entities (cluster, nodes, instances)
1250 - Added support for OS variants
1251 - Added support for hashed passwords in the Ganeti remote API users file
1253 - Added option to specify maximum timeout on instance shutdown
1254 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
1255 - Added new helper script to start and stop Ganeti daemons
1256 (``daemon-util``), with the intent to reduce the work necessary to
1257 adjust Ganeti for non-Debian distributions and to start/stop daemons
1259 - Added more unittests
1260 - Fixed critical bug in ganeti-masterd startup
1261 - Removed the configure-time ``kvm-migration-port`` parameter, this is
1262 now customisable at the cluster level for both the KVM and Xen
1263 hypervisors using the new ``migration_port`` parameter
1264 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
1265 reinstalling an instance
1266 - Allowed ``@`` in tag names
1267 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
1268 - Many documentation updates
1269 - Distribute hypervisor files on ``gnt-cluster redist-conf``
1270 - ``gnt-instance reinstall`` can now reinstall multiple instances
1271 - Updated many command line parameters
1272 - Introduced new OS API version 15
1273 - No longer support a default hypervisor
1274 - Treat virtual LVs as inexistent
1275 - Improved job locking logic to reduce lock contention
1276 - Match instance and node names case insensitively
1277 - Reimplemented bash completion script to be more complete
1284 *(Released Thu, 4 Feb 2010)*
1286 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
1287 - Fix a string formatting bug
1288 - Improve safety of the code in some error paths
1289 - Improve data validation in the master of values returned from nodes
1295 *(Released Thu, 17 Dec 2009)*
1297 - Fix security issue due to missing validation of iallocator names; this
1298 allows local and remote execution of arbitrary executables
1299 - Fix failure of gnt-node list during instance removal
1300 - Ship the RAPI documentation in the archive
1306 *(Released Wed, 30 Sep 2009)*
1308 - Fixed many wrong messages
1309 - Fixed a few bugs related to the locking library
1310 - Fixed MAC checking at instance creation time
1311 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
1312 - Fixed a few issues related to signal handling in both daemons and
1314 - Fixed the example startup script provided
1315 - Fixed insserv dependencies in the example startup script (patch from
1317 - Fixed handling of drained nodes in the iallocator framework
1318 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
1320 - Fixed error related to invalid job IDs in job polling
1321 - Fixed job/opcode persistence on unclean master shutdown
1322 - Fixed handling of partial job processing after unclean master
1324 - Fixed error reporting from LUs, previously all errors were converted
1325 into execution errors
1326 - Fixed error reporting from burnin
1327 - Decreased significantly the memory usage of the job queue
1328 - Optimised slightly multi-job submission
1329 - Optimised slightly opcode loading
1330 - Backported the multi-job submit framework from the development
1331 branch; multi-instance start and stop should be faster
1332 - Added script to clean archived jobs after 21 days; this will reduce
1333 the size of the queue directory
1334 - Added some extra checks in disk size tracking
1335 - Added an example ethers hook script
1336 - Added a cluster parameter that prevents Ganeti from modifying of
1338 - Added more node information to RAPI responses
1339 - Added a ``gnt-job watch`` command that allows following the ouput of a
1341 - Added a bind-address option to ganeti-rapi
1342 - Added more checks to the configuration verify
1343 - Enhanced the burnin script such that some operations can be retried
1345 - Converted instance reinstall to multi-instance model
1351 *(Released Fri, 7 Aug 2009)*
1353 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
1354 to allow using the pre-2.0.2 behaviour in activation, if any existing
1355 instances have mismatched disk sizes in the configuration
1356 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
1357 any configuration mismatches for disk sizes
1358 - Added ``gnt-master cluste-failover --no-voting`` to allow master
1359 failover to work on two-node clusters
1360 - Fixed the ``--net`` option of ``gnt-backup import``, which was
1362 - Fixed detection of OS script errors in ``gnt-backup export``
1363 - Fixed exit code of ``gnt-backup export``
1369 *(Released Fri, 17 Jul 2009)*
1371 - Added experimental support for stripped logical volumes; this should
1372 enhance performance but comes with a higher complexity in the block
1373 device handling; stripping is only enabled when passing
1374 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
1375 affected even in the non-stripped mode
1376 - Improved resiliency against transient failures at the end of DRBD
1377 resyncs, and in general of DRBD resync checks
1378 - Fixed a couple of issues with exports and snapshot errors
1379 - Fixed a couple of issues in instance listing
1380 - Added display of the disk size in ``gnt-instance info``
1381 - Fixed checking for valid OSes in instance creation
1382 - Fixed handling of the "vcpus" parameter in instance listing and in
1383 general of invalid parameters
1384 - Fixed http server library, and thus RAPI, to handle invalid
1385 username/password combinations correctly; this means that now they
1386 report unauthorized for queries too, not only for modifications,
1387 allowing earlier detect of configuration problems
1388 - Added a new "role" node list field, equivalent to the master/master
1389 candidate/drained/offline flags combinations
1390 - Fixed cluster modify and changes of candidate pool size
1391 - Fixed cluster verify error messages for wrong files on regular nodes
1392 - Fixed a couple of issues with node demotion from master candidate role
1393 - Fixed node readd issues
1394 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
1395 - Added a new ``--no-voting`` option for masterfailover to fix failover
1396 on two-nodes clusters when the former master node is unreachable
1397 - Added instance reinstall over RAPI
1403 *(Released Tue, 16 Jun 2009)*
1405 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
1406 allow re-adding the start in single-user option (regression from 1.2)
1407 - the watcher writes the instance status to a file, to allow monitoring
1408 to report the instance status (from the master) based on cached
1409 results of the watcher's queries; while this can get stale if the
1410 watcher is being locked due to other work on the cluster, this is
1411 still an improvement
1412 - the watcher now also restarts the node daemon and the rapi daemon if
1414 - fixed the watcher to handle full and drained queue cases
1415 - hooks export more instance data in the environment, which helps if
1416 hook scripts need to take action based on the instance's properties
1417 (no longer need to query back into ganeti)
1418 - instance failovers when the instance is stopped do not check for free
1419 RAM, so that failing over a stopped instance is possible in low memory
1421 - rapi uses queries for tags instead of jobs (for less job traffic), and
1422 for cluster tags it won't talk to masterd at all but read them from
1424 - a couple of error handling fixes in RAPI
1425 - drbd handling: improved the error handling of inconsistent disks after
1426 resync to reduce the frequency of "there are some degraded disks for
1427 this instance" messages
1428 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
1429 error handling path called a wrong function name)
1435 *(Released Wed, 27 May 2009)*
1437 - no changes from rc5
1440 Version 2.0 release candidate 5
1441 -------------------------------
1443 *(Released Wed, 20 May 2009)*
1445 - fix a couple of bugs (validation, argument checks)
1446 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
1447 - some small improvements to RAPI and IAllocator
1448 - make watcher automatically start the master daemon if down
1451 Version 2.0 release candidate 4
1452 -------------------------------
1454 *(Released Mon, 27 Apr 2009)*
1456 - change the OS list to not require locks; this helps with big clusters
1457 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
1458 volume group is broken
1459 - ``gnt-instance info``, without any arguments, doesn't run for all
1460 instances anymore; either pass ``--all`` or pass the desired
1461 instances; this helps against mistakes on big clusters where listing
1462 the information for all instances takes a long time
1463 - miscellaneous doc and man pages fixes
1466 Version 2.0 release candidate 3
1467 -------------------------------
1469 *(Released Wed, 8 Apr 2009)*
1471 - Change the internal locking model of some ``gnt-node`` commands, in
1472 order to reduce contention (and blocking of master daemon) when
1473 batching many creation/reinstall jobs
1474 - Fixes to Xen soft reboot
1475 - No longer build documentation at build time, instead distribute it in
1476 the archive, in order to reduce the need for the whole docbook/rst
1480 Version 2.0 release candidate 2
1481 -------------------------------
1483 *(Released Fri, 27 Mar 2009)*
1485 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
1486 - Fix watcher startup sequence, improves the behaviour of busy clusters
1487 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
1488 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
1489 bind address and other places
1490 - Some documentation fixes and updates
1493 Version 2.0 release candidate 1
1494 -------------------------------
1496 *(Released Mon, 2 Mar 2009)*
1498 - More documentation updates, now all docs should be more-or-less
1500 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
1502 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
1504 - Fix an issue related to $libdir/run/ganeti and cluster creation
1510 *(Released Thu, 19 Feb 2009)*
1512 - Xen PVM and KVM have switched the default value for the instance root
1513 disk to the first partition on the first drive, instead of the whole
1514 drive; this means that the OS installation scripts must be changed
1516 - Man pages have been updated
1517 - RAPI has been switched by default to HTTPS, and the exported functions
1518 should all work correctly
1519 - RAPI v1 has been removed
1520 - Many improvements to the KVM hypervisor
1521 - Block device errors are now better reported
1522 - Many other bugfixes and small improvements
1528 *(Released Mon, 26 Jan 2009)*
1530 - Version 2 is a general rewrite of the code and therefore the
1531 differences are too many to list, see the design document for 2.0 in
1532 the ``doc/`` subdirectory for more details
1533 - In this beta version there is not yet a migration path from 1.2 (there
1534 will be one in the final 2.0 release)
1535 - A few significant changes are:
1537 - all commands are executed by a daemon (``ganeti-masterd``) and the
1538 various ``gnt-*`` commands are just front-ends to it
1539 - all the commands are entered into, and executed from a job queue,
1540 see the ``gnt-job(8)`` manpage
1541 - the RAPI daemon supports read-write operations, secured by basic
1542 HTTP authentication on top of HTTPS
1543 - DRBD version 0.7 support has been removed, DRBD 8 is the only
1544 supported version (when migrating from Ganeti 1.2 to 2.0, you need
1545 to migrate to DRBD 8 first while still running Ganeti 1.2)
1546 - DRBD devices are using statically allocated minor numbers, which
1547 will be assigned to existing instances during the migration process
1548 - there is support for both Xen PVM and Xen HVM instances running on
1550 - KVM virtualization is supported too
1551 - file-based storage has been implemented, which means that it is
1552 possible to run the cluster without LVM and DRBD storage, for
1553 example using a shared filesystem exported from shared storage (and
1554 still have live migration)
1560 *(Released Tue, 13 Jan 2009)*
1562 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
1563 - Reuse the old instance mac address by default on instance import, if
1564 the instance name is the same.
1565 - Handle situations in which the node info rpc returns incomplete
1567 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
1568 - Improved version of batcher:
1570 - state file support
1571 - instance mac address support
1572 - support for HVM clusters/instances
1574 - Add an option to show the number of cpu sockets and nodes in
1576 - Support OSes that handle more than one version of the OS api (but do
1577 not change the current API in any other way)
1578 - Fix ``gnt-node migrate``
1579 - ``gnt-debug`` man page
1580 - Fixes various more typos and small issues
1581 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
1587 *(Released Wed, 24 Sep 2008)*
1589 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
1590 type of disk exported to fully virtualized instances.
1591 - provide access to the serial console of HVM instances
1592 - instance auto_balance flag, set by default. If turned off it will
1593 avoid warnings on cluster verify if there is not enough memory to fail
1594 over an instance. in the future it will prevent automatically failing
1595 it over when we will support that.
1596 - batcher tool for instance creation, see ``tools/README.batcher``
1597 - ``gnt-instance reinstall --select-os`` to interactively select a new
1598 operating system when reinstalling an instance.
1599 - when changing the memory amount on instance modify a check has been
1600 added that the instance will be able to start. also warnings are
1601 emitted if the instance will not be able to fail over, if auto_balance
1603 - documentation fixes
1604 - sync fields between ``gnt-instance list/modify/add/import``
1605 - fix a race condition in drbd when the sync speed was set after giving
1606 the device a remote peer.
1612 *(Released Tue, 22 Jul 2008)*
1614 - note: the allowed size and number of tags per object were reduced
1615 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
1616 - fixed twisted 8.x compatibility
1617 - fixed ``gnt-instance replace-disks`` with iallocator
1618 - add TCP keepalives on twisted connections to detect restarted nodes
1619 - disk increase support, see ``gnt-instance grow-disk``
1620 - implement bulk node/instance query for RAPI
1621 - add tags in node/instance listing (optional)
1622 - experimental migration (and live migration) support, read the man page
1623 for ``gnt-instance migrate``
1624 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
1625 has some small improvements in handling its state file
1631 *(Released Fri, 13 Jun 2008)*
1633 - Experimental readonly, REST-based remote API implementation;
1634 automatically started on master node, TCP port 5080, if enabled by
1635 ``--enable-rapi`` parameter to configure script.
1636 - Instance allocator support. Add and import instance accept a
1637 ``--iallocator`` parameter, and call that instance allocator to decide
1638 which node to use for the instance. The iallocator document describes
1639 what's expected from an allocator script.
1640 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
1641 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
1642 node is lost there is still enough memory to fail over the instances
1644 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
1645 ``gnt-cluster verify``, to check for site-specific compliance. All the
1646 hooks will run, and their output, if any, will be displayed. Any
1647 failing hook will make the verification return an error value.
1648 - ``gnt-cluster verify`` now checks that its peers are reachable on the
1649 primary and secondary interfaces
1650 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
1651 that is still declared as part of the cluster and has failed.
1652 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
1653 specifying output fields, that just adds the chosen fields to the
1655 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
1656 export from the filesystem.
1657 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
1658 have been added. Using them you can enable/disable acpi and pae
1659 support, and specify a path for a cd image to be exported to the
1660 instance. These parameters as the name suggest only work on HVM
1662 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
1663 PAE support will be set to the previously hardcoded values, but the
1664 (previously hardcoded) path to the CDROM ISO image will be unset and
1665 if required, needs to be set manually with ``gnt-instance modify``
1667 - The address to which an instance's VNC console is bound is now
1668 selectable per-instance, rather than being cluster wide. Of course
1669 this only applies to instances controlled via VNC, so currently just
1670 applies to HVM clusters.
1676 *(Released Mon, 18 Feb 2008)*
1678 - more tweaks to the disk activation code (especially helpful for DRBD)
1679 - change the default ``gnt-instance list`` output format, now there is
1680 one combined status field (see the manpage for the exact values this
1682 - some more fixes for the mac export to hooks change
1683 - make Ganeti not break with DRBD 8.2.x (which changed the version
1684 format in ``/proc/drbd``) (issue 24)
1685 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
1686 template, allowing migration from DRBD0.7+MD to DRBD8
1692 *(Released Wed, 30 Jan 2008)*
1694 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
1696 - add command aliases infrastructure and a few aliases
1697 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
1698 man pages and the ``--help`` option of ``gnt-node
1699 list``/``gnt-instance list``
1700 - fix ``gnt-backup list`` with down nodes (issue 21)
1701 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
1702 - fix the dist archive and add a check for including svn/git files in
1704 - some developer-related changes: improve the burnin and the QA suite,
1705 add an upload script for testing during development
1711 *(Released Wed, 16 Jan 2008)*
1713 - experimental HVM support, read the install document, section
1714 "Initializing the cluster"
1715 - allow for the PVM hypervisor per-instance kernel and initrd paths
1716 - add a new command ``gnt-cluster verify-disks`` which uses a new
1717 algorithm to improve the reconnection of the DRBD pairs if the device
1718 on the secondary node has gone away
1719 - make logical volume code auto-activate LVs at disk activation time
1720 - slightly improve the speed of activating disks
1721 - allow specification of the MAC address at instance creation time, and
1722 changing it later via ``gnt-instance modify``
1723 - fix handling of external commands that generate lots of output on
1725 - update documentation with regard to minimum version of DRBD8 supported
1731 *(Released Tue, 4 Dec 2007)*
1733 - Log the ``xm create`` output to the node daemon log on failure (to
1734 help diagnosing the error)
1735 - In debug mode, log all external commands output if failed to the logs
1736 - Change parsing of lvm commands to ignore stderr
1742 *(Released Wed, 28 Nov 2007)*
1744 - Another round of updates to the DRBD 8 code to deal with more failures
1745 in the replace secondary node operation
1746 - Some more logging of failures in disk operations (lvm, drbd)
1747 - A few documentation updates
1754 *(Released Tue, 13 Nov 2007)*
1756 - Change configuration file format from Python's Pickle to JSON.
1757 Upgrading is possible using the cfgupgrade utility.
1758 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
1759 faster replace disks and is more stable (DRBD 8 has many improvements
1760 compared to DRBD 0.7)
1761 - Added command line tags support (see man pages for ``gnt-instance``,
1762 ``gnt-node``, ``gnt-cluster``)
1763 - Added instance rename support
1764 - Added multi-instance startup/shutdown
1765 - Added cluster rename support
1766 - Added ``gnt-node evacuate`` to simplify some node operations
1767 - Added instance reboot operation that can speedup reboot as compared to
1769 - Soften the requirement that hostnames are in FQDN format
1770 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
1772 - Removed dependency on debian's patched fping that uses the
1773 non-standard ``-S`` option
1774 - Now the OS definitions are searched for in multiple, configurable
1775 paths (easier for distros to package)
1776 - Some changes to the hooks infrastructure (especially the new
1777 post-configuration update hook)
1778 - Other small bugfixes
1780 .. vim: set textwidth=72 syntax=rst :