9 - Deprecated ``admin_up`` field. Instead, ``admin_state`` is introduced,
10 with 3 possible values -- ``up``,``down`` and ``offline``.
15 *(Released Thu, 27 Oct 2011)*
17 Incompatible/important changes and bugfixes
18 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
20 - The default of the ``/2/instances/[instance_name]/rename`` RAPI
21 resource's ``ip_check`` parameter changed from ``True`` to ``False``
22 to match the underlying LUXI interface.
23 - The ``/2/nodes/[node_name]/evacuate`` RAPI resource was changed to use
24 body parameters, see :doc:`RAPI documentation <rapi>`. The server does
25 not maintain backwards-compatibility as the underlying operation
26 changed in an incompatible way. The RAPI client can talk to old
27 servers, but it needs to be told so as the return value changed.
28 - When creating file-based instances via RAPI, the ``file_driver``
29 parameter no longer defaults to ``loop`` and must be specified.
30 - The deprecated ``bridge`` NIC parameter is no longer supported. Use
32 - Support for the undocumented and deprecated RAPI instance creation
33 request format version 0 has been dropped. Use version 1, supported
34 since Ganeti 2.1.3 and :doc:`documented <rapi>`, instead.
35 - Pyparsing 1.4.6 or above is required, see :doc:`installation
36 documentation <install>`.
37 - The "cluster-verify" hooks are now executed per group by the
38 ``OP_CLUSTER_VERIFY_GROUP`` opcode. This maintains the same behavior
39 if you just run ``gnt-cluster verify``, which generates one opcode per
41 - The environment as passed to the OS scripts is cleared, and thus no
42 environment variables defined in the node daemon's environment will be
43 inherited by the scripts.
44 - The :doc:`iallocator <iallocator>` mode ``multi-evacuate`` has been
46 - :doc:`New iallocator modes <design-multi-reloc>` have been added to
47 support operations involving multiple node groups.
48 - Offline nodes are ignored when failing over an instance.
53 - The ganeti-htools project has been merged into the ganeti-core source
54 tree and will be built as part of Ganeti (see :doc:`install-quick`).
55 - Implemented support for :doc:`shared storage <design-shared-storage>`.
56 - Add support for disks larger than 2 TB in ``lvmstrap`` by supporting
57 GPT-style partition tables (requires `parted
58 <http://www.gnu.org/s/parted/>`_).
59 - Added support for floppy drive and 2nd CD-ROM drive in KVM hypervisor.
60 - Allowed adding tags on instance creation.
61 - Export instance tags to hooks (``INSTANCE_TAGS``, see :doc:`hooks`)
62 - Allow instances to be started in a paused state, enabling the user to
63 see the complete console output on boot using the console.
64 - Added new hypervisor flag to control default reboot behaviour
65 (``reboot_behavior``).
66 - Added support for KVM keymaps (hypervisor parameter ``keymap``).
67 - Improved out-of-band management support:
69 - Added ``gnt-node health`` command reporting the health status of
71 - Added ``gnt-node power`` command to manage power status of nodes.
72 - Added command for emergency power-off (EPO), ``gnt-cluster epo``.
74 - Instance migration can fall back to failover if instance is not
76 - Filters can be used when listing nodes, instances, groups and locks;
77 see *ganeti(7)* manpage.
78 - Added post-execution status as variables to :doc:`hooks <hooks>`
80 - Instance tags are exported/imported together with the instance.
81 - When given an explicit job ID, ``gnt-job info`` will work for archived
83 - Jobs can define dependencies on other jobs (not yet supported via
84 RAPI or command line, but used by internal commands and usable via
87 - Lock monitor (``gnt-debug locks``) shows jobs waiting for
90 - Instance failover is now available as a RAPI resource
91 (``/2/instances/[instance_name]/failover``).
92 - ``gnt-instance info`` defaults to static information if primary node
94 - Opcodes have a new ``comment`` attribute.
95 - Added basic SPICE support to KVM hypervisor.
96 - ``tools/ganeti-listrunner`` allows passing of arguments to executable.
98 Node group improvements
99 ~~~~~~~~~~~~~~~~~~~~~~~
101 - ``gnt-cluster verify`` has been modified to check groups separately,
102 thereby improving performance.
103 - Node group support has been added to ``gnt-cluster verify-disks``,
104 which now operates per node group.
105 - Watcher has been changed to work better with node groups.
107 - One process and state file per node group.
108 - Slow watcher in one group doesn't block other group's watcher.
110 - Added new command, ``gnt-group evacuate``, to move all instances in a
111 node group to other groups.
112 - Added ``gnt-instance change-group`` to move an instance to another
114 - ``gnt-cluster command`` and ``gnt-cluster copyfile`` now support
115 per-group operations.
116 - Node groups can be tagged.
117 - Some operations switch from an exclusive to a shared lock as soon as
119 - Instance's primary and secondary nodes' groups are now available as
120 query fields (``pnode.group``, ``pnode.group.uuid``, ``snodes.group``
121 and ``snodes.group.uuid``).
126 - Numerous updates to documentation and manpages.
128 - :doc:`RAPI <rapi>` documentation now has detailed parameter
130 - Some opcode/job results are now also documented, see :doc:`RAPI
133 - A lockset's internal lock is now also visible in lock monitor.
134 - Log messages from job queue workers now contain information about the
135 opcode they're processing.
136 - ``gnt-instance console`` no longer requires the instance lock.
137 - A short delay when waiting for job changes reduces the number of LUXI
138 requests significantly.
139 - DRBD metadata volumes are overwritten with zeros during disk creation.
145 *(Released Wed, 26 Oct 2011)*
147 This was the third release candidate of the 2.5 series.
153 *(Released Tue, 18 Oct 2011)*
155 This was the second release candidate of the 2.5 series.
161 *(Released Tue, 4 Oct 2011)*
163 This was the first release candidate of the 2.5 series.
169 *(Released Wed, 31 Aug 2011)*
171 This was the third beta release of the 2.5 series.
177 *(Released Mon, 22 Aug 2011)*
179 This was the second beta release of the 2.5 series.
185 *(Released Fri, 12 Aug 2011)*
187 This was the first beta release of the 2.5 series.
193 *(Released Thu, 27 Oct 2011)*
195 - Fixed bug when parsing command line parameter values ending in
197 - Fixed assertion error after unclean master shutdown
198 - Disable HTTP client pool for RPC, significantly reducing memory usage
200 - Fixed queue archive creation with wrong permissions
206 *(Released Tue, 23 Aug 2011)*
210 - Fixed documentation for importing with ``--src-dir`` option
211 - Fixed a bug in ``ensure-dirs`` with queue/archive permissions
212 - Fixed a parsing issue with DRBD 8.3.11 in the Linux kernel
218 *(Released Fri, 5 Aug 2011)*
220 Many bug-fixes and a few small features:
222 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
223 issues when you tried to add an instance with two MAC addresses in one
225 - KVM: fixed per-instance stored UID value
226 - KVM: configure bridged NICs at migration start
227 - KVM: Fix a bug where instance will not start with never KVM versions
229 - Added OS search path to ``gnt-cluster info``
230 - Fixed an issue with ``file_storage_dir`` where you were forced to
231 provide an absolute path, but the documentation states it is a
232 relative path, the documentation was right
233 - Added a new parameter to instance stop/start called ``--no-remember``
234 that will make the state change to not be remembered
235 - Implemented ``no_remember`` at RAPI level
236 - Improved the documentation
237 - Node evacuation: don't call IAllocator if node is already empty
238 - Fixed bug in DRBD8 replace disks on current nodes
239 - Fixed bug in recreate-disks for DRBD instances
240 - Moved assertion checking locks in ``gnt-instance replace-disks``
241 causing it to abort with not owning the right locks for some situation
242 - Job queue: Fixed potential race condition when cancelling queued jobs
243 - Fixed off-by-one bug in job serial generation
244 - ``gnt-node volumes``: Fix instance names
245 - Fixed aliases in bash completion
246 - Fixed a bug in reopening log files after being sent a SIGHUP
247 - Added a flag to burnin to allow specifying VCPU count
248 - Bugfixes to non-root Ganeti configuration
254 *(Released Thu, 12 May 2011)*
256 Many bug-fixes and a few new small features:
258 - Fixed a bug related to log opening failures
259 - Fixed a bug in instance listing with orphan instances
260 - Fixed a bug which prevented resetting the cluster-level node parameter
261 ``oob_program`` to the default
262 - Many fixes related to the ``cluster-merge`` tool
263 - Fixed a race condition in the lock monitor, which caused failures
264 during (at least) creation of many instances in parallel
265 - Improved output for gnt-job info
266 - Removed the quiet flag on some ssh calls which prevented debugging
268 - Improved the N+1 failure messages in cluster verify by actually
269 showing the memory values (needed and available)
270 - Increased lock attempt timeouts so that when executing long operations
271 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
272 too early and thus prevent the use of the 'fair' mechanism
273 - Changed instance query data (``gnt-instance info``) to not acquire
274 locks unless needed, thus allowing its use on locked instance if only
275 static information is asked for
276 - Improved behaviour with filesystems that do not support rename on an
278 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
279 kept locks on all nodes during the wipe, which is unneeded
280 - Fixed ``gnt-watcher`` handling of errors during hooks execution
281 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
282 10GiB) which caused the wipe to fail right at the end in some cases
283 - Fixed master IP activation when doing master failover with no-voting
284 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
285 the master node itself
286 - Fixed potential data-loss in under disk full conditions, where Ganeti
287 wouldn't check correctly the return code and would consider
288 partially-written files 'correct'
289 - Fixed bug related to multiple VGs and DRBD disk replacing
290 - Added new disk parameter ``metavg`` that allows placement of the meta
291 device for DRBD in a different volume group
292 - Fixed error handling in the node daemon when the system libc doesn't
293 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
294 - Fixed lock release during replace-disks, which kept cluster-wide locks
295 when doing disk replaces with an iallocator script
296 - Added check for missing bridges in cluster verify
297 - Handle EPIPE errors while writing to the terminal better, so that
298 piping the output to e.g. ``less`` doesn't cause a backtrace
299 - Fixed rare case where a ^C during Luxi calls could have been
300 interpreted as server errors, instead of simply terminating
301 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
303 - Added a few more parameters to the KVM hypervisor, allowing a second
304 CDROM, custom disk type for CDROMs and a floppy image
305 - Removed redundant message in instance rename when the name is given
307 - Added option to ``gnt-instance recreate-disks`` to allow creating the
308 disks on new nodes, allowing recreation when the original instance
309 nodes are completely gone
310 - Added option when converting disk templates to DRBD to skip waiting
311 for the resync, in order to make the instance available sooner
312 - Added two new variables to the OS scripts environment (containing the
314 - Made the root_path and optional parameter for the xen-pvm hypervisor,
315 to allow use of ``pvgrub`` as bootloader
316 - Changed the instance memory modifications to only check out-of-memory
317 conditions on memory increases, and turned the secondary node warnings
318 into errors (they can still be overridden via ``--force``)
319 - Fixed the handling of a corner case when the Python installation gets
320 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
321 to execute a command that doesn't exist
322 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
323 node of the instance returned failures during instance shutdown; this
324 adds the option ``--ignore-consistency`` to gnt-instance move
326 And as usual, various improvements to the error messages, documentation
332 *(Released Wed, 09 Mar 2011)*
334 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
335 the RAPI users file if run twice (even with ``--dry-run``).
337 The release fixes that bug (nothing else changed).
343 *(Released Mon, 07 Mar 2011)*
345 Final 2.4.0 release. Just a few small fixes:
347 - Fixed RAPI node evacuate
348 - Fixed the kvm-ifup script
349 - Fixed internal error handling for special job cases
350 - Updated man page to specify the escaping feature for options
356 *(Released Mon, 28 Feb 2011)*
358 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
359 that this feature wiped the disks of the wrong instance, leading to loss
364 - Fixed title of query field containing instance name
365 - Expanded the glossary in the documentation
366 - Fixed one unittest (internal issue)
372 *(Released Mon, 21 Feb 2011)*
374 A number of bug fixes plus just a couple functionality changes.
376 On the user-visible side, the ``gnt-* list`` command output has changed
377 with respect to "special" field states. The current rc1 style of display
378 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
379 the default output mode special fields states are displayed as follows:
381 - Offline resource: ``*``
382 - Unavailable/not applicable: ``-``
383 - Data missing (RPC failure): ``?``
384 - Unknown field: ``??``
386 Another user-visible change is the addition of ``--force-join`` to
391 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
392 - Fixed regression in RAPI/instance reinstall where all parameters were
393 required (instead of optional)
394 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
395 - Fixed iallocator usage (offline nodes were not considered offline)
396 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
397 - Fixed hypervisor and OS parameter validation with respect to
399 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
401 - Fixed ``tools/listrunner`` with respect to agent-based usage
407 *(Released Fri, 4 Feb 2011)*
409 Many changes and fixes since the beta1 release. While there were some
410 internal changes, the code has been mostly stabilised for the RC
413 Note: the dumb allocator was removed in this release, as it was not kept
414 up-to-date with the IAllocator protocol changes. It is recommended to
415 use the ``hail`` command from the ganeti-htools package.
417 Note: the 2.4 and up versions of Ganeti are not compatible with the
418 0.2.x branch of ganeti-htools. You need to upgrade to
419 ganeti-htools-0.3.0 (or later).
421 Regressions fixed from 2.3
422 ~~~~~~~~~~~~~~~~~~~~~~~~~~
424 - Fixed the ``gnt-cluster verify-disks`` command
425 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
427 - Fixed disk adoption breakage
428 - Fixed wrong headers in instance listing for field aliases
433 - Fixed corner case in KVM handling of NICs
434 - Fixed many cases of wrong handling of non-vm_capable nodes
435 - Fixed a bug where a missing instance symlink was not possible to
436 recreate with any ``gnt-*`` command (now ``gnt-instance
437 activate-disks`` does it)
438 - Fixed the volume group name as reported by ``gnt-cluster
440 - Increased timeouts for the import-export code, hopefully leading to
441 fewer aborts due network or instance timeouts
442 - Fixed bug in ``gnt-node list-storage``
443 - Fixed bug where not all daemons were started on cluster
444 initialisation, but only at the first watcher run
445 - Fixed many bugs in the OOB implementation
446 - Fixed watcher behaviour in presence of instances with offline
448 - Fixed instance list output for instances running on the wrong node
449 - a few fixes to the cluster-merge tool, but it still cannot merge
450 multi-node groups (currently it is not recommended to use this tool)
456 - Improved network configuration for the KVM hypervisor
457 - Added e1000 as a supported NIC for Xen-HVM
458 - Improved the lvmstrap tool to also be able to use partitions, as
459 opposed to full disks
460 - Improved speed of disk wiping (the cluster parameter
461 ``prealloc_wipe_disks``, so that it has a low impact on the total time
462 of instance creations
463 - Added documentation for the OS parameters
464 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
465 hypervisor is not responding
466 - Added display of blacklisted and hidden OS information in
468 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
469 NIC and node parameters, which might create problems with currently
470 invalid (but undetected) configuration files, but prevents validation
471 failures when unrelated parameters are modified
472 - Changed cluster initialisation to wait for the master daemon to become
474 - Expanded the RAPI interface:
476 - Added config redistribution resource
477 - Added activation/deactivation of instance disks
478 - Added export of console information
480 - Implemented log file reopening on SIGHUP, which allows using
481 logrotate(8) for the Ganeti log files
482 - Added a basic OOB helper script as an example
488 *(Released Fri, 14 Jan 2011)*
493 - Fixed timezone issues when formatting timestamps
494 - Added support for node groups, available via ``gnt-group`` and other
496 - Added out-of-band framework and management, see :doc:`design
497 document <design-oob>`
498 - Removed support for roman numbers from ``gnt-node list`` and
499 ``gnt-instance list``.
500 - Allowed modification of master network interface via ``gnt-cluster
501 modify --master-netdev``
502 - Accept offline secondaries while shutting down instance disks
503 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
504 - Added support for multiple LVM volume groups
505 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
507 - Added commands to list available fields:
509 - ``gnt-node list-fields``
510 - ``gnt-group list-fields``
511 - ``gnt-instance list-fields``
513 - Updated documentation and man pages
518 - Moved ``rapi_users`` file into separate directory, now named
519 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
521 - Added new tool for running commands on many machines,
522 ``tools/ganeti-listrunner``
523 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
524 improving the ``gnt-instance console`` output
525 - Allowed customisation of disk index separator at ``configure`` time
526 - Export node group allocation policy to :doc:`iallocator <iallocator>`
527 - Added support for non-partitioned md disks in ``lvmstrap``
528 - Added script to gracefully power off KVM instances
529 - Split ``utils`` module into smaller parts
530 - Changed query operations to return more detailed information, e.g.
531 whether an information is unavailable due to an offline node. To use
532 this new functionality, the LUXI call ``Query`` must be used. Field
533 information is now stored by the master daemon and can be retrieved
534 using ``QueryFields``. Instances, nodes and groups can also be queried
535 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
536 exposed via RAPI). The following commands make use of this
537 infrastructure change:
540 - ``gnt-group list-fields``
542 - ``gnt-node list-fields``
543 - ``gnt-instance list``
544 - ``gnt-instance list-fields``
545 - ``gnt-debug locks``
550 - New RAPI resources (see :doc:`rapi`):
554 - ``/2/groups/[group_name]``
555 - ``/2/groups/[group_name]/assign-nodes``
556 - ``/2/groups/[group_name]/modify``
557 - ``/2/groups/[group_name]/rename``
558 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
562 - Implemented ``no_install`` for instance creation
563 - Implemented OS parameters for instance reinstallation, allowing
564 use of special settings on reinstallation (e.g. for preserving data)
569 - Added IPv6 support in import/export
570 - Pause DRBD synchronization while wiping disks on instance creation
571 - Updated unittests and QA scripts
572 - Improved network parameters passed to KVM
573 - Converted man pages from docbook to reStructuredText
579 *(Released Mon, 20 Dec 2010)*
581 Released version 2.3.1~rc1 without any changes.
587 *(Released Wed, 1 Dec 2010)*
589 - impexpd: Disable OpenSSL compression in socat if possible (backport
590 from master, commit e90739d625b, see :doc:`installation guide
591 <install-quick>` for details)
592 - Changed unittest coverage report to exclude test scripts
593 - Added script to check version format
599 *(Released Wed, 1 Dec 2010)*
601 Released version 2.3.0~rc1 without any changes.
607 *(Released Fri, 19 Nov 2010)*
609 A number of bugfixes and documentation updates:
611 - Update ganeti-os-interface documentation
612 - Fixed a bug related to duplicate MACs or similar items which should be
614 - Fix breakage in OS state modify
615 - Reinstall instance: disallow offline secondaries (fixes bug related to
616 OS changing but reinstall failing)
617 - plus all the other fixes between 2.2.1 and 2.2.2
623 *(Released Tue, 2 Nov 2010)*
625 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
626 - Fixed master failover race with watcher
627 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
629 - Accept previously stopped instance for export with instance removal
630 - Simplify and extend the environment variables for instance OS scripts
631 - Added new node flags, ``master_capable`` and ``vm_capable``
632 - Added optional instance disk wiping prior during allocation. This is a
633 cluster-wide option and can be set/modified using
634 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
635 - Added IPv6 support, see :doc:`design document <design-2.3>` and
637 - Added a new watcher option (``--ignore-pause``)
638 - Added option to ignore offline node on instance start/stop
639 (``--ignore-offline``)
640 - Allow overriding OS parameters with ``gnt-instance reinstall``
641 - Added ability to change node's secondary IP address using ``gnt-node
643 - Implemented privilege separation for all daemons except
644 ``ganeti-noded``, see ``configure`` options
645 - Complain if an instance's disk is marked faulty in ``gnt-cluster
647 - Implemented job priorities (see ``ganeti(7)`` manpage)
648 - Ignore failures while shutting down instances during failover from
650 - Exit daemon's bootstrap process only once daemon is ready
651 - Export more information via ``LUInstanceQuery``/remote API
652 - Improved documentation, QA and unittests
653 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
654 restart if the file was created or changed
655 - Added LUXI protocol version sent with each request and response,
656 allowing detection of server/client mismatches
657 - Moved the Python scripts among gnt-* and ganeti-* into modules
658 - Moved all code related to setting up SSH to an external script,
660 - Infrastructure changes for node group support in future versions
666 *(Released Fri, 19 Nov 2010)*
668 A few small bugs fixed, and some improvements to the build system:
670 - Fix documentation regarding conversion to drbd
671 - Fix validation of parameters in cluster modify (``gnt-cluster modify
673 - Fix error handling in node modify with multiple changes
674 - Allow remote imports without checked names
680 *(Released Tue, 19 Oct 2010)*
682 - Disable SSL session ID cache in RPC client
688 *(Released Thu, 14 Oct 2010)*
690 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
691 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
693 - Fix problems with interaction between Curl and Python's HTTP server,
694 resulting in increased speed in many RPC calls
695 - Improve our release script to prevent breakage with older aclocal and
702 *(Released Thu, 7 Oct 2010)*
704 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
705 value retrieved from master
706 - Added support for blacklisted or hidden OS definitions
707 - Added simple lock monitor (accessible via (``gnt-debug locks``)
708 - Added support for -mem-path in KVM hypervisor abstraction layer
709 - Allow overriding instance parameters in tool for inter-cluster
710 instance moves (``tools/move-instance``)
711 - Improved opcode summaries (e.g. in ``gnt-job list``)
712 - Improve consistency of OS listing by sorting it
713 - Documentation updates
719 *(Released Fri, 8 Oct 2010)*
721 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
727 *(Released Mon, 4 Oct 2010)*
729 - Fixed regression in ``gnt-instance rename``
735 *(Released Wed, 22 Sep 2010)*
737 - Fixed OS_VARIANT variable for OS scripts
738 - Fixed cluster tag operations via RAPI
739 - Made ``setup-ssh`` exit with non-zero code if an error occurred
740 - Disabled RAPI CA checks in watcher
746 *(Released Mon, 23 Aug 2010)*
748 - Support DRBD versions of the format "a.b.c.d"
750 - Re-introduce support for usage from multiple threads in RAPI client
751 - Instance renames and modify via RAPI
752 - Work around race condition between processing and archival in job
754 - Mark opcodes following failed one as failed, too
755 - Job field ``lock_status`` was removed due to difficulties making it
756 work with the changed job queue in Ganeti 2.2; a better way to monitor
757 locks is expected for a later 2.2.x release
758 - Fixed dry-run behaviour with many commands
759 - Support ``ssh-agent`` again when adding nodes
760 - Many additional bugfixes
766 *(Released Fri, 30 Jul 2010)*
768 Important change: the internal RPC mechanism between Ganeti nodes has
769 changed from using a home-grown http library (based on the Python base
770 libraries) to use the PycURL library. This requires that PycURL is
771 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
772 against GnuTLS by default. cURL's support for GnuTLS had known issues
773 before cURL 7.21.0 and we recommend using the latest cURL release or
774 linking against OpenSSL. Most other distributions already link PycURL
775 and cURL against OpenSSL. The command::
777 python -c 'import pycurl; print pycurl.version'
779 can be used to determine the libraries PycURL and cURL are linked
782 Other significant changes:
784 - Rewrote much of the internals of the job queue, in order to achieve
785 better parallelism; this decouples job query operations from the job
786 processing, and it should allow much nicer behaviour of the master
787 daemon under load, and it also has uncovered some long-standing bugs
788 related to the job serialisation (now fixed)
789 - Added a default iallocator setting to the cluster parameters,
790 eliminating the need to always pass nodes or an iallocator for
791 operations that require selection of new node(s)
792 - Added experimental support for the LXC virtualization method
793 - Added support for OS parameters, which allows the installation of
794 instances to pass parameter to OS scripts in order to customise the
796 - Added a hypervisor parameter controlling the migration type (live or
797 non-live), since hypervisors have various levels of reliability; this
798 has renamed the 'live' parameter to 'mode'
799 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
800 logical volumes, meaning that cluster verify will ignore them and not
801 flag their presence as errors
802 - The watcher will now reset the error count for failed instances after
803 8 hours, thus allowing self-healing if the problem that caused the
804 instances to be down/fail to start has cleared in the meantime
805 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
806 check for, and warn, if the drbd module parameter ``usermode_helper``
807 is not consistent with the cluster-wide setting; this is needed to
808 make diagnose easier of failed drbd creations
809 - Started adding base IPv6 support, but this is not yet
810 enabled/available for use
811 - Rename operations (cluster, instance) will now return the new name,
812 which is especially useful if a short name was passed in
813 - Added support for instance migration in RAPI
814 - Added a tool to pre-configure nodes for the SSH setup, before joining
815 them to the cluster; this will allow in the future a simplified model
816 for node joining (but not yet fully enabled in 2.2); this needs the
817 paramiko python library
818 - Fixed handling of name-resolving errors
819 - Fixed consistency of job results on the error path
820 - Fixed master-failover race condition when executed multiple times in
822 - Fixed many bugs related to the job queue (mostly introduced during the
823 2.2 development cycle, so not all are impacting 2.1)
824 - Fixed instance migration with missing disk symlinks
825 - Fixed handling of unknown jobs in ``gnt-job archive``
826 - And many other small fixes/improvements
830 - Enhanced both the unittest and the QA coverage
831 - Switched the opcode validation to a generic model, and extended the
832 validation to all opcode parameters
833 - Changed more parts of the code that write shell scripts to use the
835 - Switched the master daemon to use the asyncore library for the Luxi
842 *(Released Thu, 17 Jun 2010)*
844 - Added tool (``move-instance``) and infrastructure to move instances
845 between separate clusters (see :doc:`separate documentation
846 <move-instance>` and :doc:`design document <design-2.2>`)
847 - Added per-request RPC timeout
848 - RAPI now requires a Content-Type header for requests with a body (e.g.
849 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
850 :rfc:`2616` (HTTP/1.1), section 7.2.1)
851 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
853 - Implemented initial support for running Ganeti daemons as separate
854 users, see configure-time flags ``--with-user-prefix`` and
855 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
857 - Instances can be removed after export (``gnt-backup export
859 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
860 key (instead of 1024 bit)
861 - Added new cluster configuration file for cluster domain secret
862 - Import/export now use SSL instead of SSH
863 - Added support for showing estimated time when exporting an instance,
864 see the ``ganeti-os-interface(7)`` manpage and look for
871 *(Released Tue, 16 Nov 2010)*
873 Some more bugfixes. Unless critical bugs occur, this will be the last
876 - Fix case of MAC special-values
877 - Fix mac checker regex
878 - backend: Fix typo causing "out of range" error
879 - Add missing --units in gnt-instance list man page
885 *(Released Tue, 24 Aug 2010)*
888 - Don't ignore secondary node silently on non-mirrored disk templates
890 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
891 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
892 - Properly document the usb_mouse parameter
893 - Fix path in ganeti-rapi(8) (issue 116)
894 - Adjust error message when the ganeti user's .ssh directory is
896 - Add same-node-check when changing the disk template to drbd
902 *(Released Fri, 16 Jul 2010)*
905 - Add an option to only select some reboot types during qa/burnin.
906 (on some hypervisors consequent reboots are not supported)
907 - Fix infrequent race condition in master failover. Sometimes the old
908 master ip address would be still detected as up for a short time
909 after it was removed, causing failover to fail.
910 - Decrease mlockall warnings when the ctypes module is missing. On
911 Python 2.4 we support running even if no ctypes module is installed,
912 but we were too verbose about this issue.
913 - Fix building on old distributions, on which man doesn't have a
915 - Fix RAPI not to ignore the MAC address on instance creation
916 - Implement the old instance creation format in the RAPI client.
922 *(Released Thu, 01 Jul 2010)*
924 A small bugfix release:
925 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
926 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
927 a non-existing option
928 - Fix instance create: the --force-variant option was ignored
929 - Improve pylint 0.21 compatibility and warnings with Python 2.6
930 - Fix modify node storage with non-FQDN arguments
931 - Fix RAPI client to authenticate under Python 2.6 when used
932 for more than 5 requests needing authentication
933 - Fix gnt-instance modify -t (storage) giving a wrong error message
934 when converting a non-shutdown drbd instance to plain
940 *(Released Fri, 18 Jun 2010)*
942 A small bugfix release:
944 - Fix live migration of KVM instances started with older Ganeti
945 versions which had fewer hypervisor parameters
946 - Fix gnt-instance grow-disk on down instances
947 - Fix an error-reporting bug during instance migration
948 - Better checking of the ``--net`` and ``--disk`` values, to avoid
949 silently ignoring broken ones
950 - Fix an RPC error reporting bug affecting, for example, RAPI client
952 - Fix bug triggered by different API version os-es on different nodes
953 - Fix a bug in instance startup with custom hvparams: OS level
954 parameters would fail to be applied.
955 - Fix the RAPI client under Python 2.6 (but more work is needed to
956 make it work completely well with OpenSSL)
957 - Fix handling of errors when resolving names from DNS
963 *(Released Thu, 3 Jun 2010)*
965 A medium sized development cycle. Some new features, and some
966 fixes/small improvements/cleanups.
971 The node deamon now tries to mlock itself into memory, unless the
972 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
973 its logs, and falls back to console logging. This allows emergency
974 features such as ``gnt-node powercycle`` to work even in the event of a
975 broken node disk (tested offlining the disk hosting the node's
976 filesystem and dropping its memory caches; don't try this at home)
978 KVM: add vhost-net acceleration support. It can be tested with a new
979 enough version of the kernel and of qemu-kvm.
981 KVM: Add instance chrooting feature. If you use privilege dropping for
982 your VMs you can also now force them to chroot to an empty directory,
983 before starting the emulated guest.
985 KVM: Add maximum migration bandwith and maximum downtime tweaking
986 support (requires a new-enough version of qemu-kvm).
988 Cluster verify will now warn if the master node doesn't have the master
991 Add a new (incompatible) instance creation request format to RAPI which
992 supports all parameters (previously only a subset was supported, and it
993 wasn't possible to extend the old format to accomodate all the new
994 features. The old format is still supported, and a client can check for
995 this feature, before using it, by checking for its presence in the
996 ``features`` RAPI resource.
998 Now with ancient latin support. Try it passing the ``--roman`` option to
999 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
1000 (requires the python-roman module to be installed, in order to work).
1005 As usual many internal code refactorings, documentation updates, and
1008 - Lots of improvements and cleanups to the experimental Remote API
1009 (RAPI) client library.
1010 - A new unit test suite for the core daemon libraries.
1011 - A fix to creating missing directories makes sure the umask is not
1012 applied anymore. This enforces the same directory permissions
1014 - Better handling terminating daemons with ctrl+c (used when running
1015 them in debugging mode).
1016 - Fix a race condition in live migrating a KVM instance, when stat()
1017 on the old proc status file returned EINVAL, which is an unexpected
1019 - Fixed manpage checking with newer man and utf-8 charachters. But now
1020 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
1026 *(Released Fri, 7 May 2010)*
1028 Fix a bug which prevented untagged KVM instances from starting.
1034 *(Released Fri, 7 May 2010)*
1036 Another release with a long development cycle, during which many
1037 different features were added.
1039 Significant features
1040 ~~~~~~~~~~~~~~~~~~~~
1042 The KVM hypervisor now can run the individual instances as non-root, to
1043 reduce the impact of a VM being hijacked due to bugs in the
1044 hypervisor. It is possible to run all instances as a single (non-root)
1045 user, to manually specify a user for each instance, or to dynamically
1046 allocate a user out of a cluster-wide pool to each instance, with the
1047 guarantee that no two instances will run under the same user ID on any
1050 An experimental RAPI client library, that can be used standalone
1051 (without the other Ganeti libraries), is provided in the source tree as
1052 ``lib/rapi/client.py``. Note this client might change its interface in
1053 the future, as we iterate on its capabilities.
1055 A new command, ``gnt-cluster renew-crypto`` has been added to easily
1056 replace the cluster's certificates and crypto keys. This might help in
1057 case they have been compromised, or have simply expired.
1059 A new disk option for instance creation has been added that allows one
1060 to "adopt" currently existing logical volumes, with data
1061 preservation. This should allow easier migration to Ganeti from
1062 unmanaged (or managed via other software) instances.
1064 Another disk improvement is the possibility to convert between redundant
1065 (DRBD) and plain (LVM) disk configuration for an instance. This should
1066 allow better scalability (starting with one node and growing the
1067 cluster, or shrinking a two-node cluster to one node).
1069 A new feature that could help with automated node failovers has been
1070 implemented: if a node sees itself as offline (by querying the master
1071 candidates), it will try to shutdown (hard) all instances and any active
1072 DRBD devices. This reduces the risk of duplicate instances if an
1073 external script automatically failovers the instances on such nodes. To
1074 enable this, the cluster parameter ``maintain_node_health`` should be
1075 enabled; in the future this option (per the name) will enable other
1076 automatic maintenance features.
1078 Instance export/import now will reuse the original instance
1079 specifications for all parameters; that means exporting an instance,
1080 deleting it and the importing it back should give an almost identical
1081 instance. Note that the default import behaviour has changed from
1082 before, where it created only one NIC; now it recreates the original
1085 Cluster verify has added a few new checks: SSL certificates validity,
1086 /etc/hosts consistency across the cluster, etc.
1091 As usual, many internal changes were done, documentation fixes,
1094 - Fixed cluster initialization with disabled cluster storage (regression
1095 introduced in 2.1.1)
1096 - File-based storage supports growing the disks
1097 - Fixed behaviour of node role changes
1098 - Fixed cluster verify for some corner cases, plus a general rewrite of
1099 cluster verify to allow future extension with more checks
1100 - Fixed log spamming by watcher and node daemon (regression introduced
1102 - Fixed possible validation issues when changing the list of enabled
1104 - Fixed cleanup of /etc/hosts during node removal
1105 - Fixed RAPI response for invalid methods
1106 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
1107 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
1108 from ide disks, etc.)
1109 - Allow OS changes without re-installation (to record a changed OS
1110 outside of Ganeti, or to allow OS renames)
1111 - Allow instance creation without OS installation (useful for example if
1112 the OS will be installed manually, or restored from a backup not in
1114 - Implemented option to make cluster ``copyfile`` use the replication
1116 - Added list of enabled hypervisors to ssconf (possibly useful for
1118 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
1120 - A partial form of node re-IP is possible via node readd, which now
1121 allows changed node primary IP
1122 - Command line utilities now show an informational message if the job is
1124 - The logs of the master daemon now show the PID/UID/GID of the
1131 *(Released Fri, 12 Mar 2010)*
1133 During the 2.1.0 long release candidate cycle, a lot of improvements and
1134 changes have accumulated with were released later as 2.1.1.
1139 The node evacuate command (``gnt-node evacuate``) was significantly
1140 rewritten, and as such the IAllocator protocol was changed - a new
1141 request type has been added. This unfortunate change during a stable
1142 series is designed to improve performance of node evacuations; on
1143 clusters with more than about five nodes and which are well-balanced,
1144 evacuation should proceed in parallel for all instances of the node
1145 being evacuated. As such, any existing IAllocator scripts need to be
1146 updated, otherwise the above command will fail due to the unknown
1147 request. The provided "dumb" allocator has not been updated; but the
1148 ganeti-htools package supports the new protocol since version 0.2.4.
1150 Another important change is increased validation of node and instance
1151 names. This might create problems in special cases, if invalid host
1152 names are being used.
1154 Also, a new layer of hypervisor parameters has been added, that sits at
1155 OS level between the cluster defaults and the instance ones. This allows
1156 customisation of virtualization parameters depending on the installed
1157 OS. For example instances with OS 'X' may have a different KVM kernel
1158 (or any other parameter) than the cluster defaults. This is intended to
1159 help managing a multiple OSes on the same cluster, without manual
1160 modification of each instance's parameters.
1162 A tool for merging clusters, ``cluster-merge``, has been added in the
1163 tools sub-directory.
1168 - Improved the int/float conversions that should make the code more
1169 robust in face of errors from the node daemons
1170 - Fixed the remove node code in case of internal configuration errors
1171 - Fixed the node daemon behaviour in face of inconsistent queue
1172 directory (e.g. read-only file-system where we can't open the files
1174 - Fixed the behaviour of gnt-node modify for master candidate demotion;
1175 now it either aborts cleanly or, if given the new "auto_promote"
1176 parameter, will automatically promote other nodes as needed
1177 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
1178 completely prevent Ganeti from working
1179 - Fixed bug for instance export when not all disks were successfully
1181 - Fixed behaviour of node add when the new node is slow in starting up
1183 - Fixed handling of signals in the LUXI client, which should improve
1184 behaviour of command-line scripts
1185 - Added checks for invalid node/instance names in the configuration (now
1186 flagged during cluster verify)
1187 - Fixed watcher behaviour for disk activation errors
1188 - Fixed two potentially endless loops in http library, which led to the
1189 RAPI daemon hanging and consuming 100% CPU in some cases
1190 - Fixed bug in RAPI daemon related to hashed passwords
1191 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
1193 - Enhanced compatibility with non-Debian OSes, but not using absolute
1194 path in some commands and allowing customisation of the ssh
1195 configuration directory
1196 - Fixed possible future issue with new Python versions by abiding to the
1197 proper use of ``__slots__`` attribute on classes
1198 - Added checks that should prevent directory traversal attacks
1199 - Many documentation fixes based on feedback from users
1204 - Added an "early_release" more for instance replace disks and node
1205 evacuate, where we release locks earlier and thus allow higher
1206 parallelism within the cluster
1207 - Added watcher hooks, intended to allow the watcher to restart other
1208 daemons (e.g. from the ganeti-nbma project), but they can be used of
1209 course for any other purpose
1210 - Added a compile-time disable for DRBD barriers, to increase
1211 performance if the administrator trusts the power supply or the
1212 storage system to not lose writes
1213 - Added the option of using syslog for logging instead of, or in
1214 addition to, Ganeti's own log files
1215 - Removed boot restriction for paravirtual NICs for KVM, recent versions
1216 can indeed boot from a paravirtual NIC
1217 - Added a generic debug level for many operations; while this is not
1218 used widely yet, it allows one to pass the debug value all the way to
1220 - Enhanced the hooks environment for instance moves (failovers,
1221 migrations) where the primary/secondary nodes changed during the
1222 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
1223 - Enhanced data validations for many user-supplied values; one important
1224 item is the restrictions imposed on instance and node names, which
1225 might reject some (invalid) host names
1226 - Add a configure-time option to disable file-based storage, if it's not
1227 needed; this allows greater security separation between the master
1228 node and the other nodes from the point of view of the inter-node RPC
1230 - Added user notification in interactive tools if job is waiting in the
1231 job queue or trying to acquire locks
1232 - Added log messages when a job is waiting for locks
1233 - Added filtering by node tags in instance operations which admit
1234 multiple instances (start, stop, reboot, reinstall)
1235 - Added a new tool for cluster mergers, ``cluster-merge``
1236 - Parameters from command line which are of the form ``a=b,c=d`` can now
1237 use backslash escapes to pass in values which contain commas,
1238 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
1240 - For KVM, the instance name is the first parameter passed to KVM, so
1241 that it's more visible in the process list
1247 *(Released Tue, 2 Mar 2010)*
1249 Ganeti 2.1 brings many improvements with it. Major changes:
1251 - Added infrastructure to ease automated disk repairs
1252 - Added new daemon to export configuration data in a cheaper way than
1253 using the remote API
1254 - Instance NICs can now be routed instead of being associated with a
1256 - Improved job locking logic to reduce impact of jobs acquiring multiple
1257 locks waiting for other long-running jobs
1259 In-depth implementation details can be found in the Ganeti 2.1 design
1265 - Added chroot hypervisor
1266 - Added more options to xen-hvm hypervisor (``kernel_path`` and
1268 - Added more options to xen-pvm hypervisor (``use_bootloader``,
1269 ``bootloader_path`` and ``bootloader_args``)
1270 - Added the ``use_localtime`` option for the xen-hvm and kvm
1271 hypervisors, and the default value for this has changed to false (in
1272 2.0 xen-hvm always enabled it)
1273 - Added luxi call to submit multiple jobs in one go
1274 - Added cluster initialization option to not modify ``/etc/hosts``
1276 - Added network interface parameters
1277 - Added dry run mode to some LUs
1278 - Added RAPI resources:
1280 - ``/2/instances/[instance_name]/info``
1281 - ``/2/instances/[instance_name]/replace-disks``
1282 - ``/2/nodes/[node_name]/evacuate``
1283 - ``/2/nodes/[node_name]/migrate``
1284 - ``/2/nodes/[node_name]/role``
1285 - ``/2/nodes/[node_name]/storage``
1286 - ``/2/nodes/[node_name]/storage/modify``
1287 - ``/2/nodes/[node_name]/storage/repair``
1289 - Added OpCodes to evacuate or migrate all instances on a node
1290 - Added new command to list storage elements on nodes (``gnt-node
1291 list-storage``) and modify them (``gnt-node modify-storage``)
1292 - Added new ssconf files with master candidate IP address
1293 (``ssconf_master_candidates_ips``), node primary IP address
1294 (``ssconf_node_primary_ips``) and node secondary IP address
1295 (``ssconf_node_secondary_ips``)
1296 - Added ``ganeti-confd`` and a client library to query the Ganeti
1297 configuration via UDP
1298 - Added ability to run hooks after cluster initialization and before
1300 - Added automatic mode for disk replace (``gnt-instance replace-disks
1302 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
1303 after catastrophic data-loss
1304 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
1306 - Added ``gnt-instance move`` command to move instances
1307 - Added ``gnt-cluster watcher`` command to control watcher
1308 - Added ``gnt-node powercycle`` command to powercycle nodes
1309 - Added new job status field ``lock_status``
1310 - Added parseable error codes to cluster verification (``gnt-cluster
1311 verify --error-codes``) and made output less verbose (use
1312 ``--verbose`` to restore previous behaviour)
1313 - Added UUIDs to the main config entities (cluster, nodes, instances)
1314 - Added support for OS variants
1315 - Added support for hashed passwords in the Ganeti remote API users file
1317 - Added option to specify maximum timeout on instance shutdown
1318 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
1319 - Added new helper script to start and stop Ganeti daemons
1320 (``daemon-util``), with the intent to reduce the work necessary to
1321 adjust Ganeti for non-Debian distributions and to start/stop daemons
1323 - Added more unittests
1324 - Fixed critical bug in ganeti-masterd startup
1325 - Removed the configure-time ``kvm-migration-port`` parameter, this is
1326 now customisable at the cluster level for both the KVM and Xen
1327 hypervisors using the new ``migration_port`` parameter
1328 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
1329 reinstalling an instance
1330 - Allowed ``@`` in tag names
1331 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
1332 - Many documentation updates
1333 - Distribute hypervisor files on ``gnt-cluster redist-conf``
1334 - ``gnt-instance reinstall`` can now reinstall multiple instances
1335 - Updated many command line parameters
1336 - Introduced new OS API version 15
1337 - No longer support a default hypervisor
1338 - Treat virtual LVs as inexistent
1339 - Improved job locking logic to reduce lock contention
1340 - Match instance and node names case insensitively
1341 - Reimplemented bash completion script to be more complete
1348 *(Released Thu, 4 Feb 2010)*
1350 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
1351 - Fix a string formatting bug
1352 - Improve safety of the code in some error paths
1353 - Improve data validation in the master of values returned from nodes
1359 *(Released Thu, 17 Dec 2009)*
1361 - Fix security issue due to missing validation of iallocator names; this
1362 allows local and remote execution of arbitrary executables
1363 - Fix failure of gnt-node list during instance removal
1364 - Ship the RAPI documentation in the archive
1370 *(Released Wed, 30 Sep 2009)*
1372 - Fixed many wrong messages
1373 - Fixed a few bugs related to the locking library
1374 - Fixed MAC checking at instance creation time
1375 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
1376 - Fixed a few issues related to signal handling in both daemons and
1378 - Fixed the example startup script provided
1379 - Fixed insserv dependencies in the example startup script (patch from
1381 - Fixed handling of drained nodes in the iallocator framework
1382 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
1384 - Fixed error related to invalid job IDs in job polling
1385 - Fixed job/opcode persistence on unclean master shutdown
1386 - Fixed handling of partial job processing after unclean master
1388 - Fixed error reporting from LUs, previously all errors were converted
1389 into execution errors
1390 - Fixed error reporting from burnin
1391 - Decreased significantly the memory usage of the job queue
1392 - Optimised slightly multi-job submission
1393 - Optimised slightly opcode loading
1394 - Backported the multi-job submit framework from the development
1395 branch; multi-instance start and stop should be faster
1396 - Added script to clean archived jobs after 21 days; this will reduce
1397 the size of the queue directory
1398 - Added some extra checks in disk size tracking
1399 - Added an example ethers hook script
1400 - Added a cluster parameter that prevents Ganeti from modifying of
1402 - Added more node information to RAPI responses
1403 - Added a ``gnt-job watch`` command that allows following the ouput of a
1405 - Added a bind-address option to ganeti-rapi
1406 - Added more checks to the configuration verify
1407 - Enhanced the burnin script such that some operations can be retried
1409 - Converted instance reinstall to multi-instance model
1415 *(Released Fri, 7 Aug 2009)*
1417 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
1418 to allow using the pre-2.0.2 behaviour in activation, if any existing
1419 instances have mismatched disk sizes in the configuration
1420 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
1421 any configuration mismatches for disk sizes
1422 - Added ``gnt-master cluste-failover --no-voting`` to allow master
1423 failover to work on two-node clusters
1424 - Fixed the ``--net`` option of ``gnt-backup import``, which was
1426 - Fixed detection of OS script errors in ``gnt-backup export``
1427 - Fixed exit code of ``gnt-backup export``
1433 *(Released Fri, 17 Jul 2009)*
1435 - Added experimental support for stripped logical volumes; this should
1436 enhance performance but comes with a higher complexity in the block
1437 device handling; stripping is only enabled when passing
1438 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
1439 affected even in the non-stripped mode
1440 - Improved resiliency against transient failures at the end of DRBD
1441 resyncs, and in general of DRBD resync checks
1442 - Fixed a couple of issues with exports and snapshot errors
1443 - Fixed a couple of issues in instance listing
1444 - Added display of the disk size in ``gnt-instance info``
1445 - Fixed checking for valid OSes in instance creation
1446 - Fixed handling of the "vcpus" parameter in instance listing and in
1447 general of invalid parameters
1448 - Fixed http server library, and thus RAPI, to handle invalid
1449 username/password combinations correctly; this means that now they
1450 report unauthorized for queries too, not only for modifications,
1451 allowing earlier detect of configuration problems
1452 - Added a new "role" node list field, equivalent to the master/master
1453 candidate/drained/offline flags combinations
1454 - Fixed cluster modify and changes of candidate pool size
1455 - Fixed cluster verify error messages for wrong files on regular nodes
1456 - Fixed a couple of issues with node demotion from master candidate role
1457 - Fixed node readd issues
1458 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
1459 - Added a new ``--no-voting`` option for masterfailover to fix failover
1460 on two-nodes clusters when the former master node is unreachable
1461 - Added instance reinstall over RAPI
1467 *(Released Tue, 16 Jun 2009)*
1469 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
1470 allow re-adding the start in single-user option (regression from 1.2)
1471 - the watcher writes the instance status to a file, to allow monitoring
1472 to report the instance status (from the master) based on cached
1473 results of the watcher's queries; while this can get stale if the
1474 watcher is being locked due to other work on the cluster, this is
1475 still an improvement
1476 - the watcher now also restarts the node daemon and the rapi daemon if
1478 - fixed the watcher to handle full and drained queue cases
1479 - hooks export more instance data in the environment, which helps if
1480 hook scripts need to take action based on the instance's properties
1481 (no longer need to query back into ganeti)
1482 - instance failovers when the instance is stopped do not check for free
1483 RAM, so that failing over a stopped instance is possible in low memory
1485 - rapi uses queries for tags instead of jobs (for less job traffic), and
1486 for cluster tags it won't talk to masterd at all but read them from
1488 - a couple of error handling fixes in RAPI
1489 - drbd handling: improved the error handling of inconsistent disks after
1490 resync to reduce the frequency of "there are some degraded disks for
1491 this instance" messages
1492 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
1493 error handling path called a wrong function name)
1499 *(Released Wed, 27 May 2009)*
1501 - no changes from rc5
1504 Version 2.0 release candidate 5
1505 -------------------------------
1507 *(Released Wed, 20 May 2009)*
1509 - fix a couple of bugs (validation, argument checks)
1510 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
1511 - some small improvements to RAPI and IAllocator
1512 - make watcher automatically start the master daemon if down
1515 Version 2.0 release candidate 4
1516 -------------------------------
1518 *(Released Mon, 27 Apr 2009)*
1520 - change the OS list to not require locks; this helps with big clusters
1521 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
1522 volume group is broken
1523 - ``gnt-instance info``, without any arguments, doesn't run for all
1524 instances anymore; either pass ``--all`` or pass the desired
1525 instances; this helps against mistakes on big clusters where listing
1526 the information for all instances takes a long time
1527 - miscellaneous doc and man pages fixes
1530 Version 2.0 release candidate 3
1531 -------------------------------
1533 *(Released Wed, 8 Apr 2009)*
1535 - Change the internal locking model of some ``gnt-node`` commands, in
1536 order to reduce contention (and blocking of master daemon) when
1537 batching many creation/reinstall jobs
1538 - Fixes to Xen soft reboot
1539 - No longer build documentation at build time, instead distribute it in
1540 the archive, in order to reduce the need for the whole docbook/rst
1544 Version 2.0 release candidate 2
1545 -------------------------------
1547 *(Released Fri, 27 Mar 2009)*
1549 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
1550 - Fix watcher startup sequence, improves the behaviour of busy clusters
1551 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
1552 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
1553 bind address and other places
1554 - Some documentation fixes and updates
1557 Version 2.0 release candidate 1
1558 -------------------------------
1560 *(Released Mon, 2 Mar 2009)*
1562 - More documentation updates, now all docs should be more-or-less
1564 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
1566 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
1568 - Fix an issue related to $libdir/run/ganeti and cluster creation
1574 *(Released Thu, 19 Feb 2009)*
1576 - Xen PVM and KVM have switched the default value for the instance root
1577 disk to the first partition on the first drive, instead of the whole
1578 drive; this means that the OS installation scripts must be changed
1580 - Man pages have been updated
1581 - RAPI has been switched by default to HTTPS, and the exported functions
1582 should all work correctly
1583 - RAPI v1 has been removed
1584 - Many improvements to the KVM hypervisor
1585 - Block device errors are now better reported
1586 - Many other bugfixes and small improvements
1592 *(Released Mon, 26 Jan 2009)*
1594 - Version 2 is a general rewrite of the code and therefore the
1595 differences are too many to list, see the design document for 2.0 in
1596 the ``doc/`` subdirectory for more details
1597 - In this beta version there is not yet a migration path from 1.2 (there
1598 will be one in the final 2.0 release)
1599 - A few significant changes are:
1601 - all commands are executed by a daemon (``ganeti-masterd``) and the
1602 various ``gnt-*`` commands are just front-ends to it
1603 - all the commands are entered into, and executed from a job queue,
1604 see the ``gnt-job(8)`` manpage
1605 - the RAPI daemon supports read-write operations, secured by basic
1606 HTTP authentication on top of HTTPS
1607 - DRBD version 0.7 support has been removed, DRBD 8 is the only
1608 supported version (when migrating from Ganeti 1.2 to 2.0, you need
1609 to migrate to DRBD 8 first while still running Ganeti 1.2)
1610 - DRBD devices are using statically allocated minor numbers, which
1611 will be assigned to existing instances during the migration process
1612 - there is support for both Xen PVM and Xen HVM instances running on
1614 - KVM virtualization is supported too
1615 - file-based storage has been implemented, which means that it is
1616 possible to run the cluster without LVM and DRBD storage, for
1617 example using a shared filesystem exported from shared storage (and
1618 still have live migration)
1624 *(Released Tue, 13 Jan 2009)*
1626 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
1627 - Reuse the old instance mac address by default on instance import, if
1628 the instance name is the same.
1629 - Handle situations in which the node info rpc returns incomplete
1631 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
1632 - Improved version of batcher:
1634 - state file support
1635 - instance mac address support
1636 - support for HVM clusters/instances
1638 - Add an option to show the number of cpu sockets and nodes in
1640 - Support OSes that handle more than one version of the OS api (but do
1641 not change the current API in any other way)
1642 - Fix ``gnt-node migrate``
1643 - ``gnt-debug`` man page
1644 - Fixes various more typos and small issues
1645 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
1651 *(Released Wed, 24 Sep 2008)*
1653 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
1654 type of disk exported to fully virtualized instances.
1655 - provide access to the serial console of HVM instances
1656 - instance auto_balance flag, set by default. If turned off it will
1657 avoid warnings on cluster verify if there is not enough memory to fail
1658 over an instance. in the future it will prevent automatically failing
1659 it over when we will support that.
1660 - batcher tool for instance creation, see ``tools/README.batcher``
1661 - ``gnt-instance reinstall --select-os`` to interactively select a new
1662 operating system when reinstalling an instance.
1663 - when changing the memory amount on instance modify a check has been
1664 added that the instance will be able to start. also warnings are
1665 emitted if the instance will not be able to fail over, if auto_balance
1667 - documentation fixes
1668 - sync fields between ``gnt-instance list/modify/add/import``
1669 - fix a race condition in drbd when the sync speed was set after giving
1670 the device a remote peer.
1676 *(Released Tue, 22 Jul 2008)*
1678 - note: the allowed size and number of tags per object were reduced
1679 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
1680 - fixed twisted 8.x compatibility
1681 - fixed ``gnt-instance replace-disks`` with iallocator
1682 - add TCP keepalives on twisted connections to detect restarted nodes
1683 - disk increase support, see ``gnt-instance grow-disk``
1684 - implement bulk node/instance query for RAPI
1685 - add tags in node/instance listing (optional)
1686 - experimental migration (and live migration) support, read the man page
1687 for ``gnt-instance migrate``
1688 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
1689 has some small improvements in handling its state file
1695 *(Released Fri, 13 Jun 2008)*
1697 - Experimental readonly, REST-based remote API implementation;
1698 automatically started on master node, TCP port 5080, if enabled by
1699 ``--enable-rapi`` parameter to configure script.
1700 - Instance allocator support. Add and import instance accept a
1701 ``--iallocator`` parameter, and call that instance allocator to decide
1702 which node to use for the instance. The iallocator document describes
1703 what's expected from an allocator script.
1704 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
1705 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
1706 node is lost there is still enough memory to fail over the instances
1708 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
1709 ``gnt-cluster verify``, to check for site-specific compliance. All the
1710 hooks will run, and their output, if any, will be displayed. Any
1711 failing hook will make the verification return an error value.
1712 - ``gnt-cluster verify`` now checks that its peers are reachable on the
1713 primary and secondary interfaces
1714 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
1715 that is still declared as part of the cluster and has failed.
1716 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
1717 specifying output fields, that just adds the chosen fields to the
1719 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
1720 export from the filesystem.
1721 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
1722 have been added. Using them you can enable/disable acpi and pae
1723 support, and specify a path for a cd image to be exported to the
1724 instance. These parameters as the name suggest only work on HVM
1726 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
1727 PAE support will be set to the previously hardcoded values, but the
1728 (previously hardcoded) path to the CDROM ISO image will be unset and
1729 if required, needs to be set manually with ``gnt-instance modify``
1731 - The address to which an instance's VNC console is bound is now
1732 selectable per-instance, rather than being cluster wide. Of course
1733 this only applies to instances controlled via VNC, so currently just
1734 applies to HVM clusters.
1740 *(Released Mon, 18 Feb 2008)*
1742 - more tweaks to the disk activation code (especially helpful for DRBD)
1743 - change the default ``gnt-instance list`` output format, now there is
1744 one combined status field (see the manpage for the exact values this
1746 - some more fixes for the mac export to hooks change
1747 - make Ganeti not break with DRBD 8.2.x (which changed the version
1748 format in ``/proc/drbd``) (issue 24)
1749 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
1750 template, allowing migration from DRBD0.7+MD to DRBD8
1756 *(Released Wed, 30 Jan 2008)*
1758 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
1760 - add command aliases infrastructure and a few aliases
1761 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
1762 man pages and the ``--help`` option of ``gnt-node
1763 list``/``gnt-instance list``
1764 - fix ``gnt-backup list`` with down nodes (issue 21)
1765 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
1766 - fix the dist archive and add a check for including svn/git files in
1768 - some developer-related changes: improve the burnin and the QA suite,
1769 add an upload script for testing during development
1775 *(Released Wed, 16 Jan 2008)*
1777 - experimental HVM support, read the install document, section
1778 "Initializing the cluster"
1779 - allow for the PVM hypervisor per-instance kernel and initrd paths
1780 - add a new command ``gnt-cluster verify-disks`` which uses a new
1781 algorithm to improve the reconnection of the DRBD pairs if the device
1782 on the secondary node has gone away
1783 - make logical volume code auto-activate LVs at disk activation time
1784 - slightly improve the speed of activating disks
1785 - allow specification of the MAC address at instance creation time, and
1786 changing it later via ``gnt-instance modify``
1787 - fix handling of external commands that generate lots of output on
1789 - update documentation with regard to minimum version of DRBD8 supported
1795 *(Released Tue, 4 Dec 2007)*
1797 - Log the ``xm create`` output to the node daemon log on failure (to
1798 help diagnosing the error)
1799 - In debug mode, log all external commands output if failed to the logs
1800 - Change parsing of lvm commands to ignore stderr
1806 *(Released Wed, 28 Nov 2007)*
1808 - Another round of updates to the DRBD 8 code to deal with more failures
1809 in the replace secondary node operation
1810 - Some more logging of failures in disk operations (lvm, drbd)
1811 - A few documentation updates
1818 *(Released Tue, 13 Nov 2007)*
1820 - Change configuration file format from Python's Pickle to JSON.
1821 Upgrading is possible using the cfgupgrade utility.
1822 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
1823 faster replace disks and is more stable (DRBD 8 has many improvements
1824 compared to DRBD 0.7)
1825 - Added command line tags support (see man pages for ``gnt-instance``,
1826 ``gnt-node``, ``gnt-cluster``)
1827 - Added instance rename support
1828 - Added multi-instance startup/shutdown
1829 - Added cluster rename support
1830 - Added ``gnt-node evacuate`` to simplify some node operations
1831 - Added instance reboot operation that can speedup reboot as compared to
1833 - Soften the requirement that hostnames are in FQDN format
1834 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
1836 - Removed dependency on debian's patched fping that uses the
1837 non-standard ``-S`` option
1838 - Now the OS definitions are searched for in multiple, configurable
1839 paths (easier for distros to package)
1840 - Some changes to the hooks infrastructure (especially the new
1841 post-configuration update hook)
1842 - Other small bugfixes
1844 .. vim: set textwidth=72 syntax=rst :