7 *(Released Tue, 4 Oct 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.
88 - ``tools/ganeti-listrunner`` allows passing of arguments to executable.
90 Node group improvements
91 ~~~~~~~~~~~~~~~~~~~~~~~
93 - ``gnt-cluster verify`` has been modified to check groups separately,
94 thereby improving performance.
95 - Node group support has been added to ``gnt-cluster verify-disks``,
96 which now operates per node group.
97 - Watcher has been changed to work better with node groups.
99 - One process and state file per node group.
100 - Slow watcher in one group doesn't block other group's watcher.
102 - Added new command, ``gnt-group evacuate``, to move all instances in a
103 node group to other groups.
104 - Added ``gnt-instance change-group`` to move an instance to another
106 - ``gnt-cluster command`` and ``gnt-cluster copyfile`` now support
107 per-group operations.
108 - Node groups can be tagged.
109 - Some operations switch from an exclusive to a shared lock as soon as
111 - Instance's primary and secondary nodes' groups are now available as
112 query fields (``pnode.group``, ``pnode.group.uuid``, ``snodes.group``
113 and ``snodes.group.uuid``).
118 - Numerous updates to documentation and manpages.
120 - :doc:`RAPI <rapi>` documentation now has detailed parameter
123 - A lockset's internal lock is now also visible in lock monitor.
124 - Log messages from job queue workers now contain information about the
125 opcode they're processing.
126 - ``gnt-instance console`` no longer requires the instance lock.
127 - A short delay when waiting for job changes reduces the number of LUXI
128 requests significantly.
129 - DRBD metadata volumes are overwritten with zeros during disk creation.
135 *(Released Wed, 31 Aug 2011)*
137 This was the third beta release of the 2.5 series.
143 *(Released Mon, 22 Aug 2011)*
145 This was the second beta release of the 2.5 series.
151 *(Released Fri, 12 Aug 2011)*
153 This was the first beta release of the 2.5 series.
159 *(Released Tue, 23 Aug 2011)*
163 - Fixed documentation for importing with ``--src-dir`` option
164 - Fixed a bug in ``ensure-dirs`` with queue/archive permissions
165 - Fixed a parsing issue with DRBD 8.3.11 in the Linux kernel
171 *(Released Fri, 5 Aug 2011)*
173 Many bug-fixes and a few small features:
175 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
176 issues when you tried to add an instance with two MAC addresses in one
178 - KVM: fixed per-instance stored UID value
179 - KVM: configure bridged NICs at migration start
180 - KVM: Fix a bug where instance will not start with never KVM versions
182 - Added OS search path to ``gnt-cluster info``
183 - Fixed an issue with ``file_storage_dir`` where you were forced to
184 provide an absolute path, but the documentation states it is a
185 relative path, the documentation was right
186 - Added a new parameter to instance stop/start called ``--no-remember``
187 that will make the state change to not be remembered
188 - Implemented ``no_remember`` at RAPI level
189 - Improved the documentation
190 - Node evacuation: don't call IAllocator if node is already empty
191 - Fixed bug in DRBD8 replace disks on current nodes
192 - Fixed bug in recreate-disks for DRBD instances
193 - Moved assertion checking locks in ``gnt-instance replace-disks``
194 causing it to abort with not owning the right locks for some situation
195 - Job queue: Fixed potential race condition when cancelling queued jobs
196 - Fixed off-by-one bug in job serial generation
197 - ``gnt-node volumes``: Fix instance names
198 - Fixed aliases in bash completion
199 - Fixed a bug in reopening log files after being sent a SIGHUP
200 - Added a flag to burnin to allow specifying VCPU count
201 - Bugfixes to non-root Ganeti configuration
207 *(Released Thu, 12 May 2011)*
209 Many bug-fixes and a few new small features:
211 - Fixed a bug related to log opening failures
212 - Fixed a bug in instance listing with orphan instances
213 - Fixed a bug which prevented resetting the cluster-level node parameter
214 ``oob_program`` to the default
215 - Many fixes related to the ``cluster-merge`` tool
216 - Fixed a race condition in the lock monitor, which caused failures
217 during (at least) creation of many instances in parallel
218 - Improved output for gnt-job info
219 - Removed the quiet flag on some ssh calls which prevented debugging
221 - Improved the N+1 failure messages in cluster verify by actually
222 showing the memory values (needed and available)
223 - Increased lock attempt timeouts so that when executing long operations
224 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
225 too early and thus prevent the use of the 'fair' mechanism
226 - Changed instance query data (``gnt-instance info``) to not acquire
227 locks unless needed, thus allowing its use on locked instance if only
228 static information is asked for
229 - Improved behaviour with filesystems that do not support rename on an
231 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
232 kept locks on all nodes during the wipe, which is unneeded
233 - Fixed ``gnt-watcher`` handling of errors during hooks execution
234 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
235 10GiB) which caused the wipe to fail right at the end in some cases
236 - Fixed master IP activation when doing master failover with no-voting
237 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
238 the master node itself
239 - Fixed potential data-loss in under disk full conditions, where Ganeti
240 wouldn't check correctly the return code and would consider
241 partially-written files 'correct'
242 - Fixed bug related to multiple VGs and DRBD disk replacing
243 - Added new disk parameter ``metavg`` that allows placement of the meta
244 device for DRBD in a different volume group
245 - Fixed error handling in the node daemon when the system libc doesn't
246 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
247 - Fixed lock release during replace-disks, which kept cluster-wide locks
248 when doing disk replaces with an iallocator script
249 - Added check for missing bridges in cluster verify
250 - Handle EPIPE errors while writing to the terminal better, so that
251 piping the output to e.g. ``less`` doesn't cause a backtrace
252 - Fixed rare case where a ^C during Luxi calls could have been
253 interpreted as server errors, instead of simply terminating
254 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
256 - Added a few more parameters to the KVM hypervisor, allowing a second
257 CDROM, custom disk type for CDROMs and a floppy image
258 - Removed redundant message in instance rename when the name is given
260 - Added option to ``gnt-instance recreate-disks`` to allow creating the
261 disks on new nodes, allowing recreation when the original instance
262 nodes are completely gone
263 - Added option when converting disk templates to DRBD to skip waiting
264 for the resync, in order to make the instance available sooner
265 - Added two new variables to the OS scripts environment (containing the
267 - Made the root_path and optional parameter for the xen-pvm hypervisor,
268 to allow use of ``pvgrub`` as bootloader
269 - Changed the instance memory modifications to only check out-of-memory
270 conditions on memory increases, and turned the secondary node warnings
271 into errors (they can still be overridden via ``--force``)
272 - Fixed the handling of a corner case when the Python installation gets
273 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
274 to execute a command that doesn't exist
275 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
276 node of the instance returned failures during instance shutdown; this
277 adds the option ``--ignore-consistency`` to gnt-instance move
279 And as usual, various improvements to the error messages, documentation
285 *(Released Wed, 09 Mar 2011)*
287 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
288 the RAPI users file if run twice (even with ``--dry-run``).
290 The release fixes that bug (nothing else changed).
296 *(Released Mon, 07 Mar 2011)*
298 Final 2.4.0 release. Just a few small fixes:
300 - Fixed RAPI node evacuate
301 - Fixed the kvm-ifup script
302 - Fixed internal error handling for special job cases
303 - Updated man page to specify the escaping feature for options
309 *(Released Mon, 28 Feb 2011)*
311 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
312 that this feature wiped the disks of the wrong instance, leading to loss
317 - Fixed title of query field containing instance name
318 - Expanded the glossary in the documentation
319 - Fixed one unittest (internal issue)
325 *(Released Mon, 21 Feb 2011)*
327 A number of bug fixes plus just a couple functionality changes.
329 On the user-visible side, the ``gnt-* list`` command output has changed
330 with respect to "special" field states. The current rc1 style of display
331 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
332 the default output mode special fields states are displayed as follows:
334 - Offline resource: ``*``
335 - Unavailable/not applicable: ``-``
336 - Data missing (RPC failure): ``?``
337 - Unknown field: ``??``
339 Another user-visible change is the addition of ``--force-join`` to
344 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
345 - Fixed regression in RAPI/instance reinstall where all parameters were
346 required (instead of optional)
347 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
348 - Fixed iallocator usage (offline nodes were not considered offline)
349 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
350 - Fixed hypervisor and OS parameter validation with respect to
352 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
354 - Fixed ``tools/listrunner`` with respect to agent-based usage
360 *(Released Fri, 4 Feb 2011)*
362 Many changes and fixes since the beta1 release. While there were some
363 internal changes, the code has been mostly stabilised for the RC
366 Note: the dumb allocator was removed in this release, as it was not kept
367 up-to-date with the IAllocator protocol changes. It is recommended to
368 use the ``hail`` command from the ganeti-htools package.
370 Note: the 2.4 and up versions of Ganeti are not compatible with the
371 0.2.x branch of ganeti-htools. You need to upgrade to
372 ganeti-htools-0.3.0 (or later).
374 Regressions fixed from 2.3
375 ~~~~~~~~~~~~~~~~~~~~~~~~~~
377 - Fixed the ``gnt-cluster verify-disks`` command
378 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
380 - Fixed disk adoption breakage
381 - Fixed wrong headers in instance listing for field aliases
386 - Fixed corner case in KVM handling of NICs
387 - Fixed many cases of wrong handling of non-vm_capable nodes
388 - Fixed a bug where a missing instance symlink was not possible to
389 recreate with any ``gnt-*`` command (now ``gnt-instance
390 activate-disks`` does it)
391 - Fixed the volume group name as reported by ``gnt-cluster
393 - Increased timeouts for the import-export code, hopefully leading to
394 fewer aborts due network or instance timeouts
395 - Fixed bug in ``gnt-node list-storage``
396 - Fixed bug where not all daemons were started on cluster
397 initialisation, but only at the first watcher run
398 - Fixed many bugs in the OOB implementation
399 - Fixed watcher behaviour in presence of instances with offline
401 - Fixed instance list output for instances running on the wrong node
402 - a few fixes to the cluster-merge tool, but it still cannot merge
403 multi-node groups (currently it is not recommended to use this tool)
409 - Improved network configuration for the KVM hypervisor
410 - Added e1000 as a supported NIC for Xen-HVM
411 - Improved the lvmstrap tool to also be able to use partitions, as
412 opposed to full disks
413 - Improved speed of disk wiping (the cluster parameter
414 ``prealloc_wipe_disks``, so that it has a low impact on the total time
415 of instance creations
416 - Added documentation for the OS parameters
417 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
418 hypervisor is not responding
419 - Added display of blacklisted and hidden OS information in
421 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
422 NIC and node parameters, which might create problems with currently
423 invalid (but undetected) configuration files, but prevents validation
424 failures when unrelated parameters are modified
425 - Changed cluster initialisation to wait for the master daemon to become
427 - Expanded the RAPI interface:
429 - Added config redistribution resource
430 - Added activation/deactivation of instance disks
431 - Added export of console information
433 - Implemented log file reopening on SIGHUP, which allows using
434 logrotate(8) for the Ganeti log files
435 - Added a basic OOB helper script as an example
441 *(Released Fri, 14 Jan 2011)*
446 - Fixed timezone issues when formatting timestamps
447 - Added support for node groups, available via ``gnt-group`` and other
449 - Added out-of-band framework and management, see :doc:`design
450 document <design-oob>`
451 - Removed support for roman numbers from ``gnt-node list`` and
452 ``gnt-instance list``.
453 - Allowed modification of master network interface via ``gnt-cluster
454 modify --master-netdev``
455 - Accept offline secondaries while shutting down instance disks
456 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
457 - Added support for multiple LVM volume groups
458 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
460 - Added commands to list available fields:
462 - ``gnt-node list-fields``
463 - ``gnt-group list-fields``
464 - ``gnt-instance list-fields``
466 - Updated documentation and man pages
471 - Moved ``rapi_users`` file into separate directory, now named
472 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
474 - Added new tool for running commands on many machines,
475 ``tools/ganeti-listrunner``
476 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
477 improving the ``gnt-instance console`` output
478 - Allowed customisation of disk index separator at ``configure`` time
479 - Export node group allocation policy to :doc:`iallocator <iallocator>`
480 - Added support for non-partitioned md disks in ``lvmstrap``
481 - Added script to gracefully power off KVM instances
482 - Split ``utils`` module into smaller parts
483 - Changed query operations to return more detailed information, e.g.
484 whether an information is unavailable due to an offline node. To use
485 this new functionality, the LUXI call ``Query`` must be used. Field
486 information is now stored by the master daemon and can be retrieved
487 using ``QueryFields``. Instances, nodes and groups can also be queried
488 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
489 exposed via RAPI). The following commands make use of this
490 infrastructure change:
493 - ``gnt-group list-fields``
495 - ``gnt-node list-fields``
496 - ``gnt-instance list``
497 - ``gnt-instance list-fields``
498 - ``gnt-debug locks``
503 - New RAPI resources (see :doc:`rapi`):
507 - ``/2/groups/[group_name]``
508 - ``/2/groups/[group_name]/assign-nodes``
509 - ``/2/groups/[group_name]/modify``
510 - ``/2/groups/[group_name]/rename``
511 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
515 - Implemented ``no_install`` for instance creation
516 - Implemented OS parameters for instance reinstallation, allowing
517 use of special settings on reinstallation (e.g. for preserving data)
522 - Added IPv6 support in import/export
523 - Pause DRBD synchronization while wiping disks on instance creation
524 - Updated unittests and QA scripts
525 - Improved network parameters passed to KVM
526 - Converted man pages from docbook to reStructuredText
532 *(Released Mon, 20 Dec 2010)*
534 Released version 2.3.1~rc1 without any changes.
540 *(Released Wed, 1 Dec 2010)*
542 - impexpd: Disable OpenSSL compression in socat if possible (backport
543 from master, commit e90739d625b, see :doc:`installation guide
544 <install-quick>` for details)
545 - Changed unittest coverage report to exclude test scripts
546 - Added script to check version format
552 *(Released Wed, 1 Dec 2010)*
554 Released version 2.3.0~rc1 without any changes.
560 *(Released Fri, 19 Nov 2010)*
562 A number of bugfixes and documentation updates:
564 - Update ganeti-os-interface documentation
565 - Fixed a bug related to duplicate MACs or similar items which should be
567 - Fix breakage in OS state modify
568 - Reinstall instance: disallow offline secondaries (fixes bug related to
569 OS changing but reinstall failing)
570 - plus all the other fixes between 2.2.1 and 2.2.2
576 *(Released Tue, 2 Nov 2010)*
578 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
579 - Fixed master failover race with watcher
580 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
582 - Accept previously stopped instance for export with instance removal
583 - Simplify and extend the environment variables for instance OS scripts
584 - Added new node flags, ``master_capable`` and ``vm_capable``
585 - Added optional instance disk wiping prior during allocation. This is a
586 cluster-wide option and can be set/modified using
587 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
588 - Added IPv6 support, see :doc:`design document <design-2.3>` and
590 - Added a new watcher option (``--ignore-pause``)
591 - Added option to ignore offline node on instance start/stop
592 (``--ignore-offline``)
593 - Allow overriding OS parameters with ``gnt-instance reinstall``
594 - Added ability to change node's secondary IP address using ``gnt-node
596 - Implemented privilege separation for all daemons except
597 ``ganeti-noded``, see ``configure`` options
598 - Complain if an instance's disk is marked faulty in ``gnt-cluster
600 - Implemented job priorities (see ``ganeti(7)`` manpage)
601 - Ignore failures while shutting down instances during failover from
603 - Exit daemon's bootstrap process only once daemon is ready
604 - Export more information via ``LUInstanceQuery``/remote API
605 - Improved documentation, QA and unittests
606 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
607 restart if the file was created or changed
608 - Added LUXI protocol version sent with each request and response,
609 allowing detection of server/client mismatches
610 - Moved the Python scripts among gnt-* and ganeti-* into modules
611 - Moved all code related to setting up SSH to an external script,
613 - Infrastructure changes for node group support in future versions
619 *(Released Fri, 19 Nov 2010)*
621 A few small bugs fixed, and some improvements to the build system:
623 - Fix documentation regarding conversion to drbd
624 - Fix validation of parameters in cluster modify (``gnt-cluster modify
626 - Fix error handling in node modify with multiple changes
627 - Allow remote imports without checked names
633 *(Released Tue, 19 Oct 2010)*
635 - Disable SSL session ID cache in RPC client
641 *(Released Thu, 14 Oct 2010)*
643 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
644 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
646 - Fix problems with interaction between Curl and Python's HTTP server,
647 resulting in increased speed in many RPC calls
648 - Improve our release script to prevent breakage with older aclocal and
655 *(Released Thu, 7 Oct 2010)*
657 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
658 value retrieved from master
659 - Added support for blacklisted or hidden OS definitions
660 - Added simple lock monitor (accessible via (``gnt-debug locks``)
661 - Added support for -mem-path in KVM hypervisor abstraction layer
662 - Allow overriding instance parameters in tool for inter-cluster
663 instance moves (``tools/move-instance``)
664 - Improved opcode summaries (e.g. in ``gnt-job list``)
665 - Improve consistency of OS listing by sorting it
666 - Documentation updates
672 *(Released Fri, 8 Oct 2010)*
674 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
680 *(Released Mon, 4 Oct 2010)*
682 - Fixed regression in ``gnt-instance rename``
688 *(Released Wed, 22 Sep 2010)*
690 - Fixed OS_VARIANT variable for OS scripts
691 - Fixed cluster tag operations via RAPI
692 - Made ``setup-ssh`` exit with non-zero code if an error occurred
693 - Disabled RAPI CA checks in watcher
699 *(Released Mon, 23 Aug 2010)*
701 - Support DRBD versions of the format "a.b.c.d"
703 - Re-introduce support for usage from multiple threads in RAPI client
704 - Instance renames and modify via RAPI
705 - Work around race condition between processing and archival in job
707 - Mark opcodes following failed one as failed, too
708 - Job field ``lock_status`` was removed due to difficulties making it
709 work with the changed job queue in Ganeti 2.2; a better way to monitor
710 locks is expected for a later 2.2.x release
711 - Fixed dry-run behaviour with many commands
712 - Support ``ssh-agent`` again when adding nodes
713 - Many additional bugfixes
719 *(Released Fri, 30 Jul 2010)*
721 Important change: the internal RPC mechanism between Ganeti nodes has
722 changed from using a home-grown http library (based on the Python base
723 libraries) to use the PycURL library. This requires that PycURL is
724 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
725 against GnuTLS by default. cURL's support for GnuTLS had known issues
726 before cURL 7.21.0 and we recommend using the latest cURL release or
727 linking against OpenSSL. Most other distributions already link PycURL
728 and cURL against OpenSSL. The command::
730 python -c 'import pycurl; print pycurl.version'
732 can be used to determine the libraries PycURL and cURL are linked
735 Other significant changes:
737 - Rewrote much of the internals of the job queue, in order to achieve
738 better parallelism; this decouples job query operations from the job
739 processing, and it should allow much nicer behaviour of the master
740 daemon under load, and it also has uncovered some long-standing bugs
741 related to the job serialisation (now fixed)
742 - Added a default iallocator setting to the cluster parameters,
743 eliminating the need to always pass nodes or an iallocator for
744 operations that require selection of new node(s)
745 - Added experimental support for the LXC virtualization method
746 - Added support for OS parameters, which allows the installation of
747 instances to pass parameter to OS scripts in order to customise the
749 - Added a hypervisor parameter controlling the migration type (live or
750 non-live), since hypervisors have various levels of reliability; this
751 has renamed the 'live' parameter to 'mode'
752 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
753 logical volumes, meaning that cluster verify will ignore them and not
754 flag their presence as errors
755 - The watcher will now reset the error count for failed instances after
756 8 hours, thus allowing self-healing if the problem that caused the
757 instances to be down/fail to start has cleared in the meantime
758 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
759 check for, and warn, if the drbd module parameter ``usermode_helper``
760 is not consistent with the cluster-wide setting; this is needed to
761 make diagnose easier of failed drbd creations
762 - Started adding base IPv6 support, but this is not yet
763 enabled/available for use
764 - Rename operations (cluster, instance) will now return the new name,
765 which is especially useful if a short name was passed in
766 - Added support for instance migration in RAPI
767 - Added a tool to pre-configure nodes for the SSH setup, before joining
768 them to the cluster; this will allow in the future a simplified model
769 for node joining (but not yet fully enabled in 2.2); this needs the
770 paramiko python library
771 - Fixed handling of name-resolving errors
772 - Fixed consistency of job results on the error path
773 - Fixed master-failover race condition when executed multiple times in
775 - Fixed many bugs related to the job queue (mostly introduced during the
776 2.2 development cycle, so not all are impacting 2.1)
777 - Fixed instance migration with missing disk symlinks
778 - Fixed handling of unknown jobs in ``gnt-job archive``
779 - And many other small fixes/improvements
783 - Enhanced both the unittest and the QA coverage
784 - Switched the opcode validation to a generic model, and extended the
785 validation to all opcode parameters
786 - Changed more parts of the code that write shell scripts to use the
788 - Switched the master daemon to use the asyncore library for the Luxi
795 *(Released Thu, 17 Jun 2010)*
797 - Added tool (``move-instance``) and infrastructure to move instances
798 between separate clusters (see :doc:`separate documentation
799 <move-instance>` and :doc:`design document <design-2.2>`)
800 - Added per-request RPC timeout
801 - RAPI now requires a Content-Type header for requests with a body (e.g.
802 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
803 :rfc:`2616` (HTTP/1.1), section 7.2.1)
804 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
806 - Implemented initial support for running Ganeti daemons as separate
807 users, see configure-time flags ``--with-user-prefix`` and
808 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
810 - Instances can be removed after export (``gnt-backup export
812 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
813 key (instead of 1024 bit)
814 - Added new cluster configuration file for cluster domain secret
815 - Import/export now use SSL instead of SSH
816 - Added support for showing estimated time when exporting an instance,
817 see the ``ganeti-os-interface(7)`` manpage and look for
824 *(Released Tue, 16 Nov 2010)*
826 Some more bugfixes. Unless critical bugs occur, this will be the last
829 - Fix case of MAC special-values
830 - Fix mac checker regex
831 - backend: Fix typo causing "out of range" error
832 - Add missing --units in gnt-instance list man page
838 *(Released Tue, 24 Aug 2010)*
841 - Don't ignore secondary node silently on non-mirrored disk templates
843 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
844 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
845 - Properly document the usb_mouse parameter
846 - Fix path in ganeti-rapi(8) (issue 116)
847 - Adjust error message when the ganeti user's .ssh directory is
849 - Add same-node-check when changing the disk template to drbd
855 *(Released Fri, 16 Jul 2010)*
858 - Add an option to only select some reboot types during qa/burnin.
859 (on some hypervisors consequent reboots are not supported)
860 - Fix infrequent race condition in master failover. Sometimes the old
861 master ip address would be still detected as up for a short time
862 after it was removed, causing failover to fail.
863 - Decrease mlockall warnings when the ctypes module is missing. On
864 Python 2.4 we support running even if no ctypes module is installed,
865 but we were too verbose about this issue.
866 - Fix building on old distributions, on which man doesn't have a
868 - Fix RAPI not to ignore the MAC address on instance creation
869 - Implement the old instance creation format in the RAPI client.
875 *(Released Thu, 01 Jul 2010)*
877 A small bugfix release:
878 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
879 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
880 a non-existing option
881 - Fix instance create: the --force-variant option was ignored
882 - Improve pylint 0.21 compatibility and warnings with Python 2.6
883 - Fix modify node storage with non-FQDN arguments
884 - Fix RAPI client to authenticate under Python 2.6 when used
885 for more than 5 requests needing authentication
886 - Fix gnt-instance modify -t (storage) giving a wrong error message
887 when converting a non-shutdown drbd instance to plain
893 *(Released Fri, 18 Jun 2010)*
895 A small bugfix release:
897 - Fix live migration of KVM instances started with older Ganeti
898 versions which had fewer hypervisor parameters
899 - Fix gnt-instance grow-disk on down instances
900 - Fix an error-reporting bug during instance migration
901 - Better checking of the ``--net`` and ``--disk`` values, to avoid
902 silently ignoring broken ones
903 - Fix an RPC error reporting bug affecting, for example, RAPI client
905 - Fix bug triggered by different API version os-es on different nodes
906 - Fix a bug in instance startup with custom hvparams: OS level
907 parameters would fail to be applied.
908 - Fix the RAPI client under Python 2.6 (but more work is needed to
909 make it work completely well with OpenSSL)
910 - Fix handling of errors when resolving names from DNS
916 *(Released Thu, 3 Jun 2010)*
918 A medium sized development cycle. Some new features, and some
919 fixes/small improvements/cleanups.
924 The node deamon now tries to mlock itself into memory, unless the
925 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
926 its logs, and falls back to console logging. This allows emergency
927 features such as ``gnt-node powercycle`` to work even in the event of a
928 broken node disk (tested offlining the disk hosting the node's
929 filesystem and dropping its memory caches; don't try this at home)
931 KVM: add vhost-net acceleration support. It can be tested with a new
932 enough version of the kernel and of qemu-kvm.
934 KVM: Add instance chrooting feature. If you use privilege dropping for
935 your VMs you can also now force them to chroot to an empty directory,
936 before starting the emulated guest.
938 KVM: Add maximum migration bandwith and maximum downtime tweaking
939 support (requires a new-enough version of qemu-kvm).
941 Cluster verify will now warn if the master node doesn't have the master
944 Add a new (incompatible) instance creation request format to RAPI which
945 supports all parameters (previously only a subset was supported, and it
946 wasn't possible to extend the old format to accomodate all the new
947 features. The old format is still supported, and a client can check for
948 this feature, before using it, by checking for its presence in the
949 ``features`` RAPI resource.
951 Now with ancient latin support. Try it passing the ``--roman`` option to
952 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
953 (requires the python-roman module to be installed, in order to work).
958 As usual many internal code refactorings, documentation updates, and
961 - Lots of improvements and cleanups to the experimental Remote API
962 (RAPI) client library.
963 - A new unit test suite for the core daemon libraries.
964 - A fix to creating missing directories makes sure the umask is not
965 applied anymore. This enforces the same directory permissions
967 - Better handling terminating daemons with ctrl+c (used when running
968 them in debugging mode).
969 - Fix a race condition in live migrating a KVM instance, when stat()
970 on the old proc status file returned EINVAL, which is an unexpected
972 - Fixed manpage checking with newer man and utf-8 charachters. But now
973 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
979 *(Released Fri, 7 May 2010)*
981 Fix a bug which prevented untagged KVM instances from starting.
987 *(Released Fri, 7 May 2010)*
989 Another release with a long development cycle, during which many
990 different features were added.
995 The KVM hypervisor now can run the individual instances as non-root, to
996 reduce the impact of a VM being hijacked due to bugs in the
997 hypervisor. It is possible to run all instances as a single (non-root)
998 user, to manually specify a user for each instance, or to dynamically
999 allocate a user out of a cluster-wide pool to each instance, with the
1000 guarantee that no two instances will run under the same user ID on any
1003 An experimental RAPI client library, that can be used standalone
1004 (without the other Ganeti libraries), is provided in the source tree as
1005 ``lib/rapi/client.py``. Note this client might change its interface in
1006 the future, as we iterate on its capabilities.
1008 A new command, ``gnt-cluster renew-crypto`` has been added to easily
1009 replace the cluster's certificates and crypto keys. This might help in
1010 case they have been compromised, or have simply expired.
1012 A new disk option for instance creation has been added that allows one
1013 to "adopt" currently existing logical volumes, with data
1014 preservation. This should allow easier migration to Ganeti from
1015 unmanaged (or managed via other software) instances.
1017 Another disk improvement is the possibility to convert between redundant
1018 (DRBD) and plain (LVM) disk configuration for an instance. This should
1019 allow better scalability (starting with one node and growing the
1020 cluster, or shrinking a two-node cluster to one node).
1022 A new feature that could help with automated node failovers has been
1023 implemented: if a node sees itself as offline (by querying the master
1024 candidates), it will try to shutdown (hard) all instances and any active
1025 DRBD devices. This reduces the risk of duplicate instances if an
1026 external script automatically failovers the instances on such nodes. To
1027 enable this, the cluster parameter ``maintain_node_health`` should be
1028 enabled; in the future this option (per the name) will enable other
1029 automatic maintenance features.
1031 Instance export/import now will reuse the original instance
1032 specifications for all parameters; that means exporting an instance,
1033 deleting it and the importing it back should give an almost identical
1034 instance. Note that the default import behaviour has changed from
1035 before, where it created only one NIC; now it recreates the original
1038 Cluster verify has added a few new checks: SSL certificates validity,
1039 /etc/hosts consistency across the cluster, etc.
1044 As usual, many internal changes were done, documentation fixes,
1047 - Fixed cluster initialization with disabled cluster storage (regression
1048 introduced in 2.1.1)
1049 - File-based storage supports growing the disks
1050 - Fixed behaviour of node role changes
1051 - Fixed cluster verify for some corner cases, plus a general rewrite of
1052 cluster verify to allow future extension with more checks
1053 - Fixed log spamming by watcher and node daemon (regression introduced
1055 - Fixed possible validation issues when changing the list of enabled
1057 - Fixed cleanup of /etc/hosts during node removal
1058 - Fixed RAPI response for invalid methods
1059 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
1060 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
1061 from ide disks, etc.)
1062 - Allow OS changes without re-installation (to record a changed OS
1063 outside of Ganeti, or to allow OS renames)
1064 - Allow instance creation without OS installation (useful for example if
1065 the OS will be installed manually, or restored from a backup not in
1067 - Implemented option to make cluster ``copyfile`` use the replication
1069 - Added list of enabled hypervisors to ssconf (possibly useful for
1071 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
1073 - A partial form of node re-IP is possible via node readd, which now
1074 allows changed node primary IP
1075 - Command line utilities now show an informational message if the job is
1077 - The logs of the master daemon now show the PID/UID/GID of the
1084 *(Released Fri, 12 Mar 2010)*
1086 During the 2.1.0 long release candidate cycle, a lot of improvements and
1087 changes have accumulated with were released later as 2.1.1.
1092 The node evacuate command (``gnt-node evacuate``) was significantly
1093 rewritten, and as such the IAllocator protocol was changed - a new
1094 request type has been added. This unfortunate change during a stable
1095 series is designed to improve performance of node evacuations; on
1096 clusters with more than about five nodes and which are well-balanced,
1097 evacuation should proceed in parallel for all instances of the node
1098 being evacuated. As such, any existing IAllocator scripts need to be
1099 updated, otherwise the above command will fail due to the unknown
1100 request. The provided "dumb" allocator has not been updated; but the
1101 ganeti-htools package supports the new protocol since version 0.2.4.
1103 Another important change is increased validation of node and instance
1104 names. This might create problems in special cases, if invalid host
1105 names are being used.
1107 Also, a new layer of hypervisor parameters has been added, that sits at
1108 OS level between the cluster defaults and the instance ones. This allows
1109 customisation of virtualization parameters depending on the installed
1110 OS. For example instances with OS 'X' may have a different KVM kernel
1111 (or any other parameter) than the cluster defaults. This is intended to
1112 help managing a multiple OSes on the same cluster, without manual
1113 modification of each instance's parameters.
1115 A tool for merging clusters, ``cluster-merge``, has been added in the
1116 tools sub-directory.
1121 - Improved the int/float conversions that should make the code more
1122 robust in face of errors from the node daemons
1123 - Fixed the remove node code in case of internal configuration errors
1124 - Fixed the node daemon behaviour in face of inconsistent queue
1125 directory (e.g. read-only file-system where we can't open the files
1127 - Fixed the behaviour of gnt-node modify for master candidate demotion;
1128 now it either aborts cleanly or, if given the new "auto_promote"
1129 parameter, will automatically promote other nodes as needed
1130 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
1131 completely prevent Ganeti from working
1132 - Fixed bug for instance export when not all disks were successfully
1134 - Fixed behaviour of node add when the new node is slow in starting up
1136 - Fixed handling of signals in the LUXI client, which should improve
1137 behaviour of command-line scripts
1138 - Added checks for invalid node/instance names in the configuration (now
1139 flagged during cluster verify)
1140 - Fixed watcher behaviour for disk activation errors
1141 - Fixed two potentially endless loops in http library, which led to the
1142 RAPI daemon hanging and consuming 100% CPU in some cases
1143 - Fixed bug in RAPI daemon related to hashed passwords
1144 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
1146 - Enhanced compatibility with non-Debian OSes, but not using absolute
1147 path in some commands and allowing customisation of the ssh
1148 configuration directory
1149 - Fixed possible future issue with new Python versions by abiding to the
1150 proper use of ``__slots__`` attribute on classes
1151 - Added checks that should prevent directory traversal attacks
1152 - Many documentation fixes based on feedback from users
1157 - Added an "early_release" more for instance replace disks and node
1158 evacuate, where we release locks earlier and thus allow higher
1159 parallelism within the cluster
1160 - Added watcher hooks, intended to allow the watcher to restart other
1161 daemons (e.g. from the ganeti-nbma project), but they can be used of
1162 course for any other purpose
1163 - Added a compile-time disable for DRBD barriers, to increase
1164 performance if the administrator trusts the power supply or the
1165 storage system to not lose writes
1166 - Added the option of using syslog for logging instead of, or in
1167 addition to, Ganeti's own log files
1168 - Removed boot restriction for paravirtual NICs for KVM, recent versions
1169 can indeed boot from a paravirtual NIC
1170 - Added a generic debug level for many operations; while this is not
1171 used widely yet, it allows one to pass the debug value all the way to
1173 - Enhanced the hooks environment for instance moves (failovers,
1174 migrations) where the primary/secondary nodes changed during the
1175 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
1176 - Enhanced data validations for many user-supplied values; one important
1177 item is the restrictions imposed on instance and node names, which
1178 might reject some (invalid) host names
1179 - Add a configure-time option to disable file-based storage, if it's not
1180 needed; this allows greater security separation between the master
1181 node and the other nodes from the point of view of the inter-node RPC
1183 - Added user notification in interactive tools if job is waiting in the
1184 job queue or trying to acquire locks
1185 - Added log messages when a job is waiting for locks
1186 - Added filtering by node tags in instance operations which admit
1187 multiple instances (start, stop, reboot, reinstall)
1188 - Added a new tool for cluster mergers, ``cluster-merge``
1189 - Parameters from command line which are of the form ``a=b,c=d`` can now
1190 use backslash escapes to pass in values which contain commas,
1191 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
1193 - For KVM, the instance name is the first parameter passed to KVM, so
1194 that it's more visible in the process list
1200 *(Released Tue, 2 Mar 2010)*
1202 Ganeti 2.1 brings many improvements with it. Major changes:
1204 - Added infrastructure to ease automated disk repairs
1205 - Added new daemon to export configuration data in a cheaper way than
1206 using the remote API
1207 - Instance NICs can now be routed instead of being associated with a
1209 - Improved job locking logic to reduce impact of jobs acquiring multiple
1210 locks waiting for other long-running jobs
1212 In-depth implementation details can be found in the Ganeti 2.1 design
1218 - Added chroot hypervisor
1219 - Added more options to xen-hvm hypervisor (``kernel_path`` and
1221 - Added more options to xen-pvm hypervisor (``use_bootloader``,
1222 ``bootloader_path`` and ``bootloader_args``)
1223 - Added the ``use_localtime`` option for the xen-hvm and kvm
1224 hypervisors, and the default value for this has changed to false (in
1225 2.0 xen-hvm always enabled it)
1226 - Added luxi call to submit multiple jobs in one go
1227 - Added cluster initialization option to not modify ``/etc/hosts``
1229 - Added network interface parameters
1230 - Added dry run mode to some LUs
1231 - Added RAPI resources:
1233 - ``/2/instances/[instance_name]/info``
1234 - ``/2/instances/[instance_name]/replace-disks``
1235 - ``/2/nodes/[node_name]/evacuate``
1236 - ``/2/nodes/[node_name]/migrate``
1237 - ``/2/nodes/[node_name]/role``
1238 - ``/2/nodes/[node_name]/storage``
1239 - ``/2/nodes/[node_name]/storage/modify``
1240 - ``/2/nodes/[node_name]/storage/repair``
1242 - Added OpCodes to evacuate or migrate all instances on a node
1243 - Added new command to list storage elements on nodes (``gnt-node
1244 list-storage``) and modify them (``gnt-node modify-storage``)
1245 - Added new ssconf files with master candidate IP address
1246 (``ssconf_master_candidates_ips``), node primary IP address
1247 (``ssconf_node_primary_ips``) and node secondary IP address
1248 (``ssconf_node_secondary_ips``)
1249 - Added ``ganeti-confd`` and a client library to query the Ganeti
1250 configuration via UDP
1251 - Added ability to run hooks after cluster initialization and before
1253 - Added automatic mode for disk replace (``gnt-instance replace-disks
1255 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
1256 after catastrophic data-loss
1257 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
1259 - Added ``gnt-instance move`` command to move instances
1260 - Added ``gnt-cluster watcher`` command to control watcher
1261 - Added ``gnt-node powercycle`` command to powercycle nodes
1262 - Added new job status field ``lock_status``
1263 - Added parseable error codes to cluster verification (``gnt-cluster
1264 verify --error-codes``) and made output less verbose (use
1265 ``--verbose`` to restore previous behaviour)
1266 - Added UUIDs to the main config entities (cluster, nodes, instances)
1267 - Added support for OS variants
1268 - Added support for hashed passwords in the Ganeti remote API users file
1270 - Added option to specify maximum timeout on instance shutdown
1271 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
1272 - Added new helper script to start and stop Ganeti daemons
1273 (``daemon-util``), with the intent to reduce the work necessary to
1274 adjust Ganeti for non-Debian distributions and to start/stop daemons
1276 - Added more unittests
1277 - Fixed critical bug in ganeti-masterd startup
1278 - Removed the configure-time ``kvm-migration-port`` parameter, this is
1279 now customisable at the cluster level for both the KVM and Xen
1280 hypervisors using the new ``migration_port`` parameter
1281 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
1282 reinstalling an instance
1283 - Allowed ``@`` in tag names
1284 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
1285 - Many documentation updates
1286 - Distribute hypervisor files on ``gnt-cluster redist-conf``
1287 - ``gnt-instance reinstall`` can now reinstall multiple instances
1288 - Updated many command line parameters
1289 - Introduced new OS API version 15
1290 - No longer support a default hypervisor
1291 - Treat virtual LVs as inexistent
1292 - Improved job locking logic to reduce lock contention
1293 - Match instance and node names case insensitively
1294 - Reimplemented bash completion script to be more complete
1301 *(Released Thu, 4 Feb 2010)*
1303 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
1304 - Fix a string formatting bug
1305 - Improve safety of the code in some error paths
1306 - Improve data validation in the master of values returned from nodes
1312 *(Released Thu, 17 Dec 2009)*
1314 - Fix security issue due to missing validation of iallocator names; this
1315 allows local and remote execution of arbitrary executables
1316 - Fix failure of gnt-node list during instance removal
1317 - Ship the RAPI documentation in the archive
1323 *(Released Wed, 30 Sep 2009)*
1325 - Fixed many wrong messages
1326 - Fixed a few bugs related to the locking library
1327 - Fixed MAC checking at instance creation time
1328 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
1329 - Fixed a few issues related to signal handling in both daemons and
1331 - Fixed the example startup script provided
1332 - Fixed insserv dependencies in the example startup script (patch from
1334 - Fixed handling of drained nodes in the iallocator framework
1335 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
1337 - Fixed error related to invalid job IDs in job polling
1338 - Fixed job/opcode persistence on unclean master shutdown
1339 - Fixed handling of partial job processing after unclean master
1341 - Fixed error reporting from LUs, previously all errors were converted
1342 into execution errors
1343 - Fixed error reporting from burnin
1344 - Decreased significantly the memory usage of the job queue
1345 - Optimised slightly multi-job submission
1346 - Optimised slightly opcode loading
1347 - Backported the multi-job submit framework from the development
1348 branch; multi-instance start and stop should be faster
1349 - Added script to clean archived jobs after 21 days; this will reduce
1350 the size of the queue directory
1351 - Added some extra checks in disk size tracking
1352 - Added an example ethers hook script
1353 - Added a cluster parameter that prevents Ganeti from modifying of
1355 - Added more node information to RAPI responses
1356 - Added a ``gnt-job watch`` command that allows following the ouput of a
1358 - Added a bind-address option to ganeti-rapi
1359 - Added more checks to the configuration verify
1360 - Enhanced the burnin script such that some operations can be retried
1362 - Converted instance reinstall to multi-instance model
1368 *(Released Fri, 7 Aug 2009)*
1370 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
1371 to allow using the pre-2.0.2 behaviour in activation, if any existing
1372 instances have mismatched disk sizes in the configuration
1373 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
1374 any configuration mismatches for disk sizes
1375 - Added ``gnt-master cluste-failover --no-voting`` to allow master
1376 failover to work on two-node clusters
1377 - Fixed the ``--net`` option of ``gnt-backup import``, which was
1379 - Fixed detection of OS script errors in ``gnt-backup export``
1380 - Fixed exit code of ``gnt-backup export``
1386 *(Released Fri, 17 Jul 2009)*
1388 - Added experimental support for stripped logical volumes; this should
1389 enhance performance but comes with a higher complexity in the block
1390 device handling; stripping is only enabled when passing
1391 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
1392 affected even in the non-stripped mode
1393 - Improved resiliency against transient failures at the end of DRBD
1394 resyncs, and in general of DRBD resync checks
1395 - Fixed a couple of issues with exports and snapshot errors
1396 - Fixed a couple of issues in instance listing
1397 - Added display of the disk size in ``gnt-instance info``
1398 - Fixed checking for valid OSes in instance creation
1399 - Fixed handling of the "vcpus" parameter in instance listing and in
1400 general of invalid parameters
1401 - Fixed http server library, and thus RAPI, to handle invalid
1402 username/password combinations correctly; this means that now they
1403 report unauthorized for queries too, not only for modifications,
1404 allowing earlier detect of configuration problems
1405 - Added a new "role" node list field, equivalent to the master/master
1406 candidate/drained/offline flags combinations
1407 - Fixed cluster modify and changes of candidate pool size
1408 - Fixed cluster verify error messages for wrong files on regular nodes
1409 - Fixed a couple of issues with node demotion from master candidate role
1410 - Fixed node readd issues
1411 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
1412 - Added a new ``--no-voting`` option for masterfailover to fix failover
1413 on two-nodes clusters when the former master node is unreachable
1414 - Added instance reinstall over RAPI
1420 *(Released Tue, 16 Jun 2009)*
1422 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
1423 allow re-adding the start in single-user option (regression from 1.2)
1424 - the watcher writes the instance status to a file, to allow monitoring
1425 to report the instance status (from the master) based on cached
1426 results of the watcher's queries; while this can get stale if the
1427 watcher is being locked due to other work on the cluster, this is
1428 still an improvement
1429 - the watcher now also restarts the node daemon and the rapi daemon if
1431 - fixed the watcher to handle full and drained queue cases
1432 - hooks export more instance data in the environment, which helps if
1433 hook scripts need to take action based on the instance's properties
1434 (no longer need to query back into ganeti)
1435 - instance failovers when the instance is stopped do not check for free
1436 RAM, so that failing over a stopped instance is possible in low memory
1438 - rapi uses queries for tags instead of jobs (for less job traffic), and
1439 for cluster tags it won't talk to masterd at all but read them from
1441 - a couple of error handling fixes in RAPI
1442 - drbd handling: improved the error handling of inconsistent disks after
1443 resync to reduce the frequency of "there are some degraded disks for
1444 this instance" messages
1445 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
1446 error handling path called a wrong function name)
1452 *(Released Wed, 27 May 2009)*
1454 - no changes from rc5
1457 Version 2.0 release candidate 5
1458 -------------------------------
1460 *(Released Wed, 20 May 2009)*
1462 - fix a couple of bugs (validation, argument checks)
1463 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
1464 - some small improvements to RAPI and IAllocator
1465 - make watcher automatically start the master daemon if down
1468 Version 2.0 release candidate 4
1469 -------------------------------
1471 *(Released Mon, 27 Apr 2009)*
1473 - change the OS list to not require locks; this helps with big clusters
1474 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
1475 volume group is broken
1476 - ``gnt-instance info``, without any arguments, doesn't run for all
1477 instances anymore; either pass ``--all`` or pass the desired
1478 instances; this helps against mistakes on big clusters where listing
1479 the information for all instances takes a long time
1480 - miscellaneous doc and man pages fixes
1483 Version 2.0 release candidate 3
1484 -------------------------------
1486 *(Released Wed, 8 Apr 2009)*
1488 - Change the internal locking model of some ``gnt-node`` commands, in
1489 order to reduce contention (and blocking of master daemon) when
1490 batching many creation/reinstall jobs
1491 - Fixes to Xen soft reboot
1492 - No longer build documentation at build time, instead distribute it in
1493 the archive, in order to reduce the need for the whole docbook/rst
1497 Version 2.0 release candidate 2
1498 -------------------------------
1500 *(Released Fri, 27 Mar 2009)*
1502 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
1503 - Fix watcher startup sequence, improves the behaviour of busy clusters
1504 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
1505 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
1506 bind address and other places
1507 - Some documentation fixes and updates
1510 Version 2.0 release candidate 1
1511 -------------------------------
1513 *(Released Mon, 2 Mar 2009)*
1515 - More documentation updates, now all docs should be more-or-less
1517 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
1519 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
1521 - Fix an issue related to $libdir/run/ganeti and cluster creation
1527 *(Released Thu, 19 Feb 2009)*
1529 - Xen PVM and KVM have switched the default value for the instance root
1530 disk to the first partition on the first drive, instead of the whole
1531 drive; this means that the OS installation scripts must be changed
1533 - Man pages have been updated
1534 - RAPI has been switched by default to HTTPS, and the exported functions
1535 should all work correctly
1536 - RAPI v1 has been removed
1537 - Many improvements to the KVM hypervisor
1538 - Block device errors are now better reported
1539 - Many other bugfixes and small improvements
1545 *(Released Mon, 26 Jan 2009)*
1547 - Version 2 is a general rewrite of the code and therefore the
1548 differences are too many to list, see the design document for 2.0 in
1549 the ``doc/`` subdirectory for more details
1550 - In this beta version there is not yet a migration path from 1.2 (there
1551 will be one in the final 2.0 release)
1552 - A few significant changes are:
1554 - all commands are executed by a daemon (``ganeti-masterd``) and the
1555 various ``gnt-*`` commands are just front-ends to it
1556 - all the commands are entered into, and executed from a job queue,
1557 see the ``gnt-job(8)`` manpage
1558 - the RAPI daemon supports read-write operations, secured by basic
1559 HTTP authentication on top of HTTPS
1560 - DRBD version 0.7 support has been removed, DRBD 8 is the only
1561 supported version (when migrating from Ganeti 1.2 to 2.0, you need
1562 to migrate to DRBD 8 first while still running Ganeti 1.2)
1563 - DRBD devices are using statically allocated minor numbers, which
1564 will be assigned to existing instances during the migration process
1565 - there is support for both Xen PVM and Xen HVM instances running on
1567 - KVM virtualization is supported too
1568 - file-based storage has been implemented, which means that it is
1569 possible to run the cluster without LVM and DRBD storage, for
1570 example using a shared filesystem exported from shared storage (and
1571 still have live migration)
1577 *(Released Tue, 13 Jan 2009)*
1579 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
1580 - Reuse the old instance mac address by default on instance import, if
1581 the instance name is the same.
1582 - Handle situations in which the node info rpc returns incomplete
1584 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
1585 - Improved version of batcher:
1587 - state file support
1588 - instance mac address support
1589 - support for HVM clusters/instances
1591 - Add an option to show the number of cpu sockets and nodes in
1593 - Support OSes that handle more than one version of the OS api (but do
1594 not change the current API in any other way)
1595 - Fix ``gnt-node migrate``
1596 - ``gnt-debug`` man page
1597 - Fixes various more typos and small issues
1598 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
1604 *(Released Wed, 24 Sep 2008)*
1606 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
1607 type of disk exported to fully virtualized instances.
1608 - provide access to the serial console of HVM instances
1609 - instance auto_balance flag, set by default. If turned off it will
1610 avoid warnings on cluster verify if there is not enough memory to fail
1611 over an instance. in the future it will prevent automatically failing
1612 it over when we will support that.
1613 - batcher tool for instance creation, see ``tools/README.batcher``
1614 - ``gnt-instance reinstall --select-os`` to interactively select a new
1615 operating system when reinstalling an instance.
1616 - when changing the memory amount on instance modify a check has been
1617 added that the instance will be able to start. also warnings are
1618 emitted if the instance will not be able to fail over, if auto_balance
1620 - documentation fixes
1621 - sync fields between ``gnt-instance list/modify/add/import``
1622 - fix a race condition in drbd when the sync speed was set after giving
1623 the device a remote peer.
1629 *(Released Tue, 22 Jul 2008)*
1631 - note: the allowed size and number of tags per object were reduced
1632 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
1633 - fixed twisted 8.x compatibility
1634 - fixed ``gnt-instance replace-disks`` with iallocator
1635 - add TCP keepalives on twisted connections to detect restarted nodes
1636 - disk increase support, see ``gnt-instance grow-disk``
1637 - implement bulk node/instance query for RAPI
1638 - add tags in node/instance listing (optional)
1639 - experimental migration (and live migration) support, read the man page
1640 for ``gnt-instance migrate``
1641 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
1642 has some small improvements in handling its state file
1648 *(Released Fri, 13 Jun 2008)*
1650 - Experimental readonly, REST-based remote API implementation;
1651 automatically started on master node, TCP port 5080, if enabled by
1652 ``--enable-rapi`` parameter to configure script.
1653 - Instance allocator support. Add and import instance accept a
1654 ``--iallocator`` parameter, and call that instance allocator to decide
1655 which node to use for the instance. The iallocator document describes
1656 what's expected from an allocator script.
1657 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
1658 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
1659 node is lost there is still enough memory to fail over the instances
1661 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
1662 ``gnt-cluster verify``, to check for site-specific compliance. All the
1663 hooks will run, and their output, if any, will be displayed. Any
1664 failing hook will make the verification return an error value.
1665 - ``gnt-cluster verify`` now checks that its peers are reachable on the
1666 primary and secondary interfaces
1667 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
1668 that is still declared as part of the cluster and has failed.
1669 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
1670 specifying output fields, that just adds the chosen fields to the
1672 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
1673 export from the filesystem.
1674 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
1675 have been added. Using them you can enable/disable acpi and pae
1676 support, and specify a path for a cd image to be exported to the
1677 instance. These parameters as the name suggest only work on HVM
1679 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
1680 PAE support will be set to the previously hardcoded values, but the
1681 (previously hardcoded) path to the CDROM ISO image will be unset and
1682 if required, needs to be set manually with ``gnt-instance modify``
1684 - The address to which an instance's VNC console is bound is now
1685 selectable per-instance, rather than being cluster wide. Of course
1686 this only applies to instances controlled via VNC, so currently just
1687 applies to HVM clusters.
1693 *(Released Mon, 18 Feb 2008)*
1695 - more tweaks to the disk activation code (especially helpful for DRBD)
1696 - change the default ``gnt-instance list`` output format, now there is
1697 one combined status field (see the manpage for the exact values this
1699 - some more fixes for the mac export to hooks change
1700 - make Ganeti not break with DRBD 8.2.x (which changed the version
1701 format in ``/proc/drbd``) (issue 24)
1702 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
1703 template, allowing migration from DRBD0.7+MD to DRBD8
1709 *(Released Wed, 30 Jan 2008)*
1711 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
1713 - add command aliases infrastructure and a few aliases
1714 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
1715 man pages and the ``--help`` option of ``gnt-node
1716 list``/``gnt-instance list``
1717 - fix ``gnt-backup list`` with down nodes (issue 21)
1718 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
1719 - fix the dist archive and add a check for including svn/git files in
1721 - some developer-related changes: improve the burnin and the QA suite,
1722 add an upload script for testing during development
1728 *(Released Wed, 16 Jan 2008)*
1730 - experimental HVM support, read the install document, section
1731 "Initializing the cluster"
1732 - allow for the PVM hypervisor per-instance kernel and initrd paths
1733 - add a new command ``gnt-cluster verify-disks`` which uses a new
1734 algorithm to improve the reconnection of the DRBD pairs if the device
1735 on the secondary node has gone away
1736 - make logical volume code auto-activate LVs at disk activation time
1737 - slightly improve the speed of activating disks
1738 - allow specification of the MAC address at instance creation time, and
1739 changing it later via ``gnt-instance modify``
1740 - fix handling of external commands that generate lots of output on
1742 - update documentation with regard to minimum version of DRBD8 supported
1748 *(Released Tue, 4 Dec 2007)*
1750 - Log the ``xm create`` output to the node daemon log on failure (to
1751 help diagnosing the error)
1752 - In debug mode, log all external commands output if failed to the logs
1753 - Change parsing of lvm commands to ignore stderr
1759 *(Released Wed, 28 Nov 2007)*
1761 - Another round of updates to the DRBD 8 code to deal with more failures
1762 in the replace secondary node operation
1763 - Some more logging of failures in disk operations (lvm, drbd)
1764 - A few documentation updates
1771 *(Released Tue, 13 Nov 2007)*
1773 - Change configuration file format from Python's Pickle to JSON.
1774 Upgrading is possible using the cfgupgrade utility.
1775 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
1776 faster replace disks and is more stable (DRBD 8 has many improvements
1777 compared to DRBD 0.7)
1778 - Added command line tags support (see man pages for ``gnt-instance``,
1779 ``gnt-node``, ``gnt-cluster``)
1780 - Added instance rename support
1781 - Added multi-instance startup/shutdown
1782 - Added cluster rename support
1783 - Added ``gnt-node evacuate`` to simplify some node operations
1784 - Added instance reboot operation that can speedup reboot as compared to
1786 - Soften the requirement that hostnames are in FQDN format
1787 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
1789 - Removed dependency on debian's patched fping that uses the
1790 non-standard ``-S`` option
1791 - Now the OS definitions are searched for in multiple, configurable
1792 paths (easier for distros to package)
1793 - Some changes to the hooks infrastructure (especially the new
1794 post-configuration update hook)
1795 - Other small bugfixes
1797 .. vim: set textwidth=72 syntax=rst :